Linus Drops BitKeeper

ravenII wrote in to mention a story running on CNet, which discusses Linus Torvald's decision to no longer use BitKeeper. From the article: "Linus Torvalds is looking for a new SCM for his project's source code after a conflict involving the current management system, BitKeeper. 'I've decided to not use BK (BitKeeper) mainly because I need to figure out the alternatives,' Torvalds said in a posting. 'Rather than continuing things as normal, I decided to bite the bullet and just see what life without BK looks like.' Coverage on the BitKeeper announcement from earlier this week is also available. Update: 04/10 16:36 GMT by Z : Updated to reflect the story's origin.

Comments

He should use ClearCase.

[Anonymous Coward]

It's what the professionals use.

I believe even Microsoft don't always 'eat their own dogfood' on this one.

Re:He should use ClearCase.

[Archeopteryx]

Clearcase is great! But requires a full time administrator.

Re:He should use ClearCase.

[CypherOz]

It's what the professionals use.
We use IBM Rational products at work.

Unless IBM donate (to ALL kernel developers) or open source ClearCase then it is way too expensive.

What's more you need the top end version of ClearCase for distributed repositories.

Re:He should use ClearCase.

[Anonymous Coward]

Have you actually used Clearcase over anything but a 100Mbps networks? It's slow. And I mean SLOW as in getting bored out of your mind, before it manages to update a tree of a few thousand files that you updated a week ago over the 10Mb radio-lan link between this house and the next. I'll take Subversion or CVS any day.

Re:He should use ClearCase.

[Nasarius]

Unless IBM donate (to ALL kernel developers) or open source ClearCase then it is way too expensive. What's more you need the top end version of ClearCase for distributed repositories.

That's not so unlikely, if you think about it. It costs nothing for IBM to give out a few dozen licenses, and it would buy them even further goodwill and publicity from the Linux community.

Personally, I hope Linus would reject any such offer (in favor of encouraging fellow open-source developers to make something really good), having learned his lesson about using proprietary software.

Re:He should use ClearCase.

[Anonymous Coward]

Its not slow, its expensive to run it on the recommended hardware, so people don't. If your software is mission critical, clearcase is the way forward.

Re:He should use ClearCase.

[Anonymous Coward]

I'm not sure how urban planning works where you're from. Perhaps anyone is allowed to dig up the pavement and install their own network lines. It doesn't work that way where I'm from, and in many many larger project the point is moot since the different parties are in different pars of the country or even the world. How much exactly did you feel one should spend on dedicated satellite links so that one could use the "way forward" in the form of Clearcase's ridiculously inefficient network protocol?

Re:He should use ClearCase.

[Waffle Iron]

Its not slow, its expensive to run it on the recommended hardware, so people don't.

Of course, expensive hardware has to be recommended to compensate for the fact that it's slow.

Re:He should use ClearCase.

[mrbooze]

I assume the official ClearCase solution to that is using distributed servers in each region that are synchronized/replicated with each other.

I've never worked with ClearCase though, so maybe not. We still use Continuus where I work, which has a distributed server option for different regional development offices.

Re:He should use ClearCase.

[Anonymous Coward]

That would be the party line as far as I know. Unfortunately, my experience is that clearcase is unacceptably slow over a 10MB radio-lan uplink (to the next building shared by 20 or so people. Internal network 100Mbps switched). I don't have exakt numbers, but if I start an update of a view and then go take a coffey break, the update is almost guaranteed to still be running when i return 10 or 15 minutes later. This with perhaps 6-7000 files in the view/VOB. Changing the tree is even worse. Checking in(Committing in CVS-ian) is apt to be a 10 minute affair if i have 30 or so files to check in. And I don't mean administrative time, such figuring out which lables to apply, I mean waiting for the clearcase client and nothing else. When you add administrative tasks such as applying lables and custom attributes you'll probably be talking a total of an hour, and 80% of that is nothing but waiting for clearcase. Over a 100Mbps LAN these problems are almost gone. It's still far slower than CVS or Subversion, but acceptable.

Re:He should use ClearCase.

[Anonymous Coward]

ClearCase's support for distributed repositories requires frequent merges. My company just started using it for my product. Development is done in three sites. Two of the sites have to merge *every day* to stay in sync with the master site. In some cases, I think they even have to merge more than once per day. Our ClearCase admin has been doing CC administration full-time for the last 6 years or so, so I don't think he's incompetent. There has to be a better way, but we haven't found it. I predict that we'll try this and see how it works, then go shopping for a new SCM in about 6 months.

Re:He should use ClearCase.

[filterchild]

I don't think he's going to switch from one commercial SCM to another.

Re:He should use ClearCase.

[Rosyna]

I believe even Microsoft don't always 'eat their own dogfood' on this one.

That's what I've heard as well. Internally MS wouldn't touch Visual SourceSafe with a 10 foot donkey poking pole.

Re:He should use ClearCase.

[cahiha]

If you are going to put a full time administrator on your source code control system, just about any reasonable system is going to work like a charm.

Re:He should use ClearCase.

[smithmc]

I believe even Microsoft don't always 'eat their own dogfood' on this one.

AIUI Microsoft uses a customized version of Perforce, called Source Depot.

Re:He should use ClearCase.

[stor]

If your software is mission critical, clearcase is the way forward.

What a convincing argument. Which brochure did you copy that line from?

Cheers
Stor

Re:He should use ClearCase.

[DrXym]

Heh I know :) Sadly Clearcase is an attractive nuisance, a moneypit and a piece of shit all rolled into one. It's appallingly expensive, needs high end servers to support even a modest number of developers and is very admin intensive.

Unless you are on the same LAN as a Clearcase server, the thing is treacle slow because it sends dozens of packets flying back and forth just to figure out what items to put on a context menu. If you're using a VPN then creating a snapshot view can take many, many hours and even simple things like checkouts / checkins / diffs take minutes.

And because it works so badly over the WAN, if you have multiple sites you must replicate - more expensive servers, more admins and more licences.

It's not even a good source control system. It doesn't do anything aside from a dynamic view that can't be done by most other systems. Dynamic views are more trouble than they're worth anyway.

I truly pity companies who "bet the farm" on this junk. I pity IBM who had a perfectly usable source control system in CMVC who had to switch to this Rational junk.

For all its faults even CVS would be better. And with Subversion being available and UI frontends like TortoiseSVN, Subclipse etc. there really is no reason to be stuck with Clearcase.

Re:He should use ClearCase.

[DrXym]

Yes it is slow. Horribly slow. It's not just the hardware, but the amount of traffic it sends back and forth even for simple operations. It's almost impossible to use Clearcase if you are not physically situated on the same network as the server you're trying to talk to. I should know because I have to do work through a VPN frequently enough.

The issue isn't the traffic or the bandwidth but the latency caused by not being on the same network. Since it sends off a flurry of packets, the effect of even a modest latency (e.g. 100ms) is culmulative and the thing simply sucks. It is just a horribly designed system.

Rational's "solution" to this is to sell a multi-site version that replicates the same repository for each LAN you wish to use the source control from. But this does nothing for the VPN situation and it also means you must burden the enormous expense of replicating your your hardware, and administration for everywhere you want to use it.

It's not mission critical anything - it's an overengineered piece of crap. I truly expect that a sane source control system could get ten fold performance out of the same hardware simply because it's smart enough to figure out if a file is checked out without flooding the network with traffic.

I truly believe that Clearcase could be junked and replaced with Bitkeeper or even Subversion in 99% of cases and it would be a better fit.

Re:He should use ClearCase.

[bheading]

Microsoft allegedly use Perforce. VSS is an absolute turd.

Clearcase is nice in many respects, but to get that niceness you have to pay top dollar. Not only do you pay for the licenses, but then there's the training (Clearcase usage - particularly administration - is highly specialized), the Clearcase admin ($$$$) and the big machines you need to run the view and VOB servers. But what you get is ideal - the revision control database is completely transparent to userspace applications, and it's impossible of users to delete or corrupt the metadata. The Multisite extension, while expensive, allows a lot of powerful options when dealing with a distributed source base. The nicest feature of all is the build avoidance capability.

However there are many aspects about it which are thoroughly unsatisfying. It's the little details that cause all the time to be wasted. One particularly irritating thing that bugs me is that if you edit your configuration spec and make a typo in a label or branch, Clearcase doesn't tell you, and you'll be working away merrily, blissfully unaware that the baseline you're working with is not the one you intended. Another irritating aspect is that some of the commands are long and have all kinds of detailed options. What you end up doing is writing a whole bunch of wrapper scripts, and the upshot of that of course is that no two Clearcase sites are the same so your experience does not transfer so easily from one place to another. For example in base Clearcase the command to find out what changes are in another branch the command is :

ct findmerge -avobs -fversion .../ -print

in Bitkeeper it's

bk changes -R

in Clearcase to list all the labels in a VOB :

ct lstype -kind lbtype

in BK or CVS or similar it's

bk tags

etc.

Re:He should use ClearCase.

[swamp boy]

ClearCase does indeed have an alternative to dynamic views -- snapshot views. I worked with snapshot views back in 2000. Perhaps you should know more about the product before you decide to disparage it. ClearCase does have bandwidth issues on a WAN (easily solved with replication), but IMHO it can't be beat as a good SCM repository.

Re:He should use ClearCase.

[DrXym]

Excuse me, but I do know it. Indeed if you'd bothered to read what I said you would have seen I mentioned both snapshot and dynamic views. It is impossible to use dynamic views over a WAN and snapshots are just barely tolerable and horribly slow.

Re:He should use ClearCase.

[omb]

Oh NO he shouldn't, ClearCase is the most counter productive tool a developer could ever use, it encourages locked serial commits, is complicated, buggy, slow, poorly documented and poorly supported even for large enterprise customers.

All of the objections Linus has made to SVN also apply, and worse, to ClearCase.

Why, anyway, would it be better to buy a ClearCase licence than a BitKeeper licence.

What we need is something new here, and 'git' may be it.

Let me say that I use SVN, internally, used to use CVS, and both work well, and much better than the likes of ClearCase and PVCS; which both bought the configuration management lie.

SVN works very well with a small group of tightly co-ordinated developers, 1-10 people, but this is NOT linux, there are xx,000 out there.

Re:He should use ClearCase.

[mandolin]

I truly pity companies who "bet the farm" on this junk.

For a company that does almost all of its development on a single LAN (like mine), Clearcase's changeset-oriented streams and semi-intelligent merge features kick the crap out of CVS.

You're correct that dynamic views are too much trouble; I use snapshot views so I can control when I sync with the rest of the codebase.

I'm not saying Linus should use Clearcase, but it does have its place.

Re:He should use ClearCase.

[Anonymous Coward]

We used to use CMVC, and we switched to Rational because it has no work-in-place model as ClearCase views do.

Re:He should use ClearCase.

[retinaburn]

From what I have seen of CMVC it would be good for large scale development projects. However configuration looked to be a pain in the butt to get it set up right and define everything correctly. For small projects it just had way too much process overhead to make it efficent, Defects-Tracks-Levels-Releases.

Although it does handle file renames etc much easier than cvs.

If IBM did release the source (which I can only believe would be ugly), or make it freely available to download, there is a ton of documentation already out there for it (god bless redbooks).

So cmvc is not a perfect solution, but it might be better than some of the other alternatives.

Re:He should use ClearCase.

[Anonymous Coward]

ClearCase is useless for real development with large teams distributed around the globe.

The system is too slow, too expensive and requires a full time administrator to hold the thing up by its ears. For large (distributed) projects you need something which is properly robust. Somthing that doesn't require an admin on call 24 hours per day. This is important.

It is really infuriating to have to call someone on the other side of the globe to get out of bed and try to unfuckify the SCM-system at 0300 on a sunday morning.

ClearCase is for small teams that work regular hours in the same office. I have tried 5 different SCM systems the past 15 years and ClearCase is not even on my top 3 lists of SCM systems to consider. Seriously.

Re:He should use ClearCase.

[tsotha]

We've used dynamic views in clearcase for a decade, and we love it. I'm not sure why you're having so much trouble. Our lead developer might spend an hour a month administering the server, which is an obsolete SunOS box.

The biggest problem we've had recently is being stuck on the 2.4 kernal until IBM supports 2.6.

Re:He should use ClearCase.

[DrXym]

I'm having so much trouble because CC is so packet-happy that it's impossible to get acceptable performance when you have a built-in latency such as for VPN. It's not the bandwidth, but the lead-in time that kills performance. I have no idea what data its sending, but on a VPN it takes a good 3-4 seconds before it even knows if a file is checked out or not. If you gripe to the admins you point to you a IBM CC knowledge base article that says VPN / WAN connections are not supported because performance is so horrendous.

The result is you have to be on the same LAN or it is worthless. I believe they're producing some kind of remote client software but its too little too late.

Irrespective of bandwidth, dynamic views are evil simply because you have no idea what you're building against. You couldn be working away and everything is fine one moment and the next everything is hosed. Why is it hosed? It could be becaused your bug is intermittant or someone has gone and changed some other files that affect what you're looking at You might even discover your view is thoroughly broken because someone has checked in something on their view and gone home. I'd only use static views even if I were right next to the server - it's the only way to keep some control of your development environment.

Re:He should use ClearCase.

[Anonymous Coward]

You don't understand SCM. You still have to communicate with your other developers. You can completely isolate yourself from broken builds if you know what you are doing. Its all to do with the config spec.

A dynamic view should always use a branch off a labelled version if you want stability. Its very obvious to me that you were not trained in ClearCase best practices. I would say RTFM, but I doubt it would change your views.

Re:He should use ClearCase.

[Anonymous Coward]

Every top investment bank in the USA uses ClearCase. Are you telling me these guys don't understand distributed development? On trading systems that are worked on in London, New York and Tokyo, by some of the smartest developers in the world.

Perhaps you didn't install ClearCase correctly. I have never experienced any issues with it at all, and this is with Multi-Site and concurrent distributed development.

Re:He should use ClearCase.

[Anonymous Coward]

it encourages locked serial commits, is complicated, buggy, slow, poorly documented and poorly supported even for large enterprise customers.

This is all untrue. ClearCase encourages parallel development via branching and merging, and is the only CM tool that is transparent to the OS and supports true directory versioning. You obviously don't know what you are talking about.

Re:He should use ClearCase.

[tsotha]

The result is you have to be on the same LAN or it is worthless

Oh, it's true you can't deal with much latency in dynamic views. When I have to use a WAN connection I'll normally telnet into a machine at work and use gvim instead of my normal IDE. But I'm not out of the office much, so it might be more of an inconvenience for you.

Why is it hosed? It could be becaused your bug is intermittant or someone has gone and changed some other files that affect what you're looking at You might even discover your view is thoroughly broken because someone has checked in something on their view and gone home.

It doesn't sound like your team is taking advantage of the strengths of the tool. You really shouldn't be having this problem if you've set your branches up properly. I think IBM ships a "best practices" book which you might want to take a look at.

Re:He should use ClearCase.

[Anonymous Coward]

and you know this how?

Re:He should use ClearCase.

[DrXym]

Actually I understand source control just fine. Now explain to me what the point of a dynamic view is if you say I have to work off an independent branch? If that's the case why not do a static view?

Re:He should use ClearCase.

[DrXym]

The latency for static views is awful too but fortunately you pick a time when to update the view so it's not quite so bad. It doesn't sound like your team is taking advantage of the strengths of the tool. You really shouldn't be having this problem if you've set your branches up properly. I think IBM ships a "best practices" book which you might want to take a look at.

I'm talking about a small team working on the same dynamic view. Working without treading on each others toes is impossible and the problems expand with each new member. Now I know what you mean by best practice - we could all go off and create individual branches, but that would introduce complications of its own such as the extra inconvenience of constantly merging. Besides, if everyone has a personal branch, there is no point to using a dynamic view unless you make a ruling that everything which is merged should be picked up by everyone. And that breaks as much as if you just let people check in to the same view since people often forget to checkin directories or new files.

It's simpler for everyone to use a snapshot and checkin / checkout in a disciplined fashion.

Re:He should use ClearCase.

[Anonymous Coward]

Now explain to me what the point of a dynamic view is if you say I have to work off an independent branch? If that's the case why not do a static view?

ClearCase is partly about maximizing the potential parallelism of development (something that is not critical for many projects). Dynamic views are not an all-or-nothing thing. You can choose to pick (say) version 1.2.4 of one subdirectory, plus any dynamic updates to a shared branch on another directory, plus a specific labelled version of one file, and the latest version of another file so long as it has a certain attribute. The ClearCase config spec is totally flexible in that respect.

Perhaps you've never worked on a 200+ developer project, but this is pretty useful stuff when you know that some parts are stable and some are not, and you want to have 10 concurrent versions of a software project in development, merging change sets backwards and forwards between release candidates.

Point is, If I base my dynamic view of an integration branch, or better yet a 'release candidate branch' I can ensure that my configuration item includes the latest working/tested version of any subcomponent.

When your product includes 15000+ C++ classes, believe me, this is important.

Re:He should use ClearCase.

[Anonymous Coward]

I'm talking about a small team working on the same dynamic view.

Why on Earth would you want to do this? Doing this is to miss the whole point of ClearCase!

Branches are cheap, labelling is cheap, merging is semi-automatic. What are you gaining by using ClearCase at all? You might as well just have a tarfile with the current source, and extract it in directory version x.y.z for each new build. Your project does not need SCM if you are using ClearCase in this brain-dead and retarded fashion!!!

Re:He should use ClearCase.

[Anonymous Coward]

You do realise, that labels and branching are all you need to isolate yourself from changes on /main/LATEST ?

Re:He should use ClearCase.

[DrXym]

Sigh, I give up. I tell you dynamic views suck, you say they don't if we have our branches set up (we do), and then you ask me why on earth I'd do that. The answer is I wouldn't. Dynamic views suck. Idiot.

Re:He should use ClearCase.

[Anonymous Coward]

The point is you should not be sharing the view. Dynamic views do not suck, or all the top Wall St firms would not be using ClearCase and loving it. For certain classes of global enterprise development, it is simply the only game in town. We can agree to differ, that's OK. You are still wrong.

Re:He should use ClearCase.

[DrXym]

Heh, I work for one of those "top Wall St firms" and it sucks bigtime.

Hmm.

[millennial]

Well, there's always CVS... *sigh*...
Or you could teach your code to two hundred trained squirrels, a la Tim Burton. Then every time you changed some code, you could train another squirrel. Not only would you have an army of Code Squirrels at your command, but... eh, you'd probably be locked up...

Re:Hmm.

[lewp]

Ah. Concurrent Version Squirrels.

Re:Hmm.

[Anonymous Coward]

In soviet russia the bitkeeper dumps you... er wait.

Re:Hmm.

[node 3]

a la Tim Burton

I'm confused...

Linus is Willy Wonka, and the rest of the kernel developers are Oompa Loompas?

What has me confused isn't so much that I think that that's true, but that it just seems so right somehow.

Oompa Loompa doompadee doo
I've got another puzzle for you.
Oompa Loompa doompadah dee
If you are wise you will listen to me.

Who do you blame when your software goes closed,
The source is kept secret, so nobody knows?
Blaming the user is such a shame
You know exactly who's to blame:
The closed software proprietor!

Oompa Loompa doompadee dah
If you're use GNU then you will go far.
You will live in freedom too
Like the Oompa Loompa doompadee do.

Re:Hmm.

[Frank T. Lofaro Jr.]

Well yeah, he could go to CVS and get some headache medicine, he'll probably need it.

Re:Hmm.

[iplayfast]

Damn! Now I've got the oompa Loompa brain virus! It will take me a week to get this out of my head!

Re:Hmm.

[geminidomino]

*wipes Coke off his LCD monitor*

You, sir, are a sick, twisted bastard.

Re:Hmm.

[Anonymous Coward]

I still use RCS, you insensitive clod!

Re:How about...

[Anonymous Coward]

Are you trying to subvert Linux? You must work for Microsoft!

Ba-dum-tish! Thank you, I'll be here all week. Try the veal, tip your waitress!

Re:DUPE

[millennial]

Not a dupe at all. The previous article was about how KernelTrap was no longer supporting the free version of Bitkeeper. This is about how Linus is no longer using it.
To compare: Article A talks about George Bush Sr. ending his term. Article B talks about Bill Clinton becoming president. Dupe? I think not.

Re:How about...

[Scaba]

Here's why not (read the PS).

Re:How about...

[Speare]

How about Subversion

How about reading Subversion's writeup on why that's not a good idea?

Better headline might be:

[Anonymous Coward]

Bitkeeper loses only customer

Re:Better headline might be:

[Cramer]

More accurately: BitMover loses all of it's advertising.

Seriously, have you ever head of BitKeeper outside of Linux discussions? Have you ever seen an ad for BitKeeper anywhere? Even if the quoted figures are right ($500,000/year), that's pretty cheap advertising for the amount of coverage.

(I've known about BK/BitMover for a long time simply because I know of Larry McVoy. But aside from that association, I've never heard of either.)

Re:Better headline might be:

[SunFan]

Anyone who is searching for version control software will find BitKeeper pretty quickly. It isn't like they really need to advertise--their target market will find them.

Re:Better headline might be:

[Anonymous Coward]

> Bitkeeper loses only customer

Perhaps not the only one.

Re:Better headline might be:

[Breakfast Pants]

Yes you find it now pretty easily. But is that because it was used for Linux for a long while? I can't say for sure but I know how page rank works and if you search for any word half the time you get something dealing with linux in the results.

Re:Better headline might be:

[SunFan]

Wow, moderated into the gutter. What's the agenda, here?

Re:DUPE

[Anonymous Coward]

Um. Yes, there is something new in this story.

The previous story was "BitKeeper drops free version" with a mention that Linux used the free version of BitKeeper, and would probably drop it in favor of something else. This story is that Linux HAS dropped BitKeeper support, confirming suspisions, and links to the previous story as to why he did.

It's like 'continuing coverage'.

Re:DUPE

[millennial]

My God, what have I done? I'm feeding trolls! *cries pitiously*

Breaking News!

[Qbertino]

The Pope died. ...
Honestly, didn't we have this two days ago?
BTW, he said n ot to bug him with Subversion. He's not going to use it because it's not the ideal tool for the job. So say the Subeversion people themselves. He's currently checking out Monotone. But as I said: We had that 2 days ago.

Re:Breaking News!

[Nimrangul]

Meh, I'd say he should try going with OpenCVS, though it won't be ready for major use for months at least unless some more developers get interested and join in the work load.

Re:Breaking News!

[Anonymous Coward]

Ironically enough, he mentioned in TFA he might be willing to use SVN as a last resort, if a month goes by and as a private repository for the central Politburo of maintainers, rather than force all patches to go through him again.

Re:Breaking News!

[katana]

There's an idea. Try something that doesn't work in the hopes that someone else will fix it because it doesn't work for you. Then again, I think this is the entire premise of Queer Eye for the Straight Guy.

Re:Breaking News!

[Nimrangul]

I know what you were trying to say, but you said it poorly enough that I had to reread your post to understand you.

I was more saying that they could take OpenCVS and add what is wanted to it, at first it will only be a basic CVS system, thus it won't have all that added good stuff Linus pines for.

GnuCVS is buggy and unmaintainable, thus OpenCVS was born; but it isn't being made to just be the same old CVS, they want to update it and tighten it down for security purposes. If people got into it and started contributing they could add some of the stuff people want.

Theoretically you could get OpenCVS to contain CVS, RCS and a few other little bits; just as OpenSSH isn't just a daemon and client but a suite.

Anyways, at this time the daemon isn't ready yet, only the client. So as I said, unless someone wanted it done soon and had the skill or money to back that it would be a while before it is ready even if Linus did want it.

Re:Breaking News!

[katana]

In your world, Linus could manage the kernel in vi if only enough people got into it and started contributing to it. So meh, he should switch to vi even though it won't be ready for months unless developers start working on it. Does this form of argument actually make sense to you?

Proposing that Linus should switch to a tool that you freely admit will not work for him is patently ideological and completely unhelpful. Stop it. You're hurting America.

Re:Breaking News!

[Nimrangul]

No, but he should be able to write it in vi, that's what vi is for. CVS is for version management, so he should be able to manage it in CVS. I think you were thinking of emacs.

I don't get your "you're hurting America." bullcrap, I am pointing out that he has a chance to get something made that works the way he wants it to, cause he's been really whiney and wants something that works a special way.

Hell, if he wanted he could try supporting one of the BitKeeper clones too, I was proposing another option.

Honestly I think Linus should have not say in the matter, that the developers should have a vote for what it is put on.

Re:Breaking News!

[katana]

Since the illustrations aren't working, let me break it down for you. You proposed using a piece of software that you freely admit will not work for what Linus needs. Your justification was that it *could* be what he needs if only people would work on it and make it what he needs. Following this line of reasoning, *any* software could be proposed as a solution, because any software *could* be what he needs if enough people work on it and make it into what he needs. Therefore your proposal for a specific piece of software as a superior option to any other is not supported by your argument. The other opinions you're tossing around are irrelevant to this point, though you get points for being honest about not understanding things.

Re:Breaking News!

[Nimrangul]

Well, it's not like CVS is that far from what he wants, you were using vi as an example and that's completely off base.

If you'd said Arch or something like that we'd have three less posts on this thread.

I specifically said OpenCVS because I trust what those guys put out to be secure and I'd rather Linus use something safe than something done by just any random group.

All the work the OpenBSD guys go through to tighen the system and tools I just would rather that other systems take the tools they work on more seriously I guess.

Re:Breaking News!

[katana]

At the risk of violating the Law of Minimum Efficient Threads, I will point out one last time that the argument that you made (that Linus should use OpenCVS, even though it doesn't work, because people could make it work) is incredibly weak. Your auxiliary hypotheses about distance from the ideal Linus tool aren't much better, given that if you don't know what Linus needs, you don't know whether OpenCVS is closer or not. But since you're retreating to weak pragmatism rather than making an argument, there's really nothing more to say here.

Re:Breaking News!

[Anonymous Coward]

How is it that Subversion would not work but CVS would? Do you think it may have to do with a centralized versus distributed model?

Re:Breaking News!

[Nimrangul]

I said nothing about subversion good nor bad, that was because I think highly of the people working on OpenCVS's programming skills, I trust them to make something secure and functional.

I was simply pointing to something I would trust, feel free to read into my lack of mentioning SVK too.

Re:Breaking News!

[Nimrangul]

I do know what Linus needs, or at least what he says he needs, as he's mentioned his wish list a couple times online.

My weak pragmatism was the basis for the entire post in the first place, I never gave some grand proposal nor listed pros and cons of using different systems, I simply said he should try OpenCVS and that it would not be ready if he wanted it now.

There was no massive point in how he could make it something special in my originating post, any additional opinion was added based on posts made in response. My added opinions were added as counter-arguments, not parts of the base post and thus were not the basis for my posting.

You really do seem riled up about this, I just sorta wonder why.

Re:Breaking News!

[katana]

Remember, every time you engage in sloppy thinking, Linus kills a kitten.

Not an option.

[Makzu]

The Subversion devs have already stated that Subversion is a poor choice for kernel development. In fact, the title of the page I just linked is "Please Stop Bugging Linus Torvalds About Subversion." Plus, Linus himself said "Don't bother telling me about Subversion" in his e-mail.

Re:Not an option.

[CypherOz]

Humor asside... I think Linus needs to think about how the kernel source is managed (i.e. the control model) and maybe change that... Then select an appropriate tool.

Get your requirements and architecture right first, then implement. This may mean a change to how the kernel development process works.

Re:Not an option.

[gonk]

Um, you don't think he has thought about it?

robert

Re:Not an option.

[SpaceLifeForm]

The kernel development process *has* already changed. The 2.6 series (vs 2.4) really was able to use BK effectively in conjunction with those 'control model' changes. Linus has already said that he doesn't scale. So, the way it's being handled now is the way it will likely remain. He's not likely to change his behaviour because of a tool problem. He will find a new tool.

Re:DUPE

[millennial]

We're talking about Linus, not Linux. Remember, Linus created Linux, much like Bill Gates created AIDS.
/using Windows right now, doesn't really care either way about MS

Three Words

[MyLongNickName]

Visual Source Safe!

Re:Three Words

[DaHat]

As yet there is no good Linux client for it (V6 at least)... just wait for VSTS (Team System)... it will have 3rd party clients for just about every platform when it ships.

Re:Three Words

[Anonymous Coward]

How about Linus does the following:

1) Find something that is open source/free software that already does what he wants - which is what he's currently trying to do.

2) Start a project that creates a new source code management system that does everything that he wants - basically scratch his own itch. He might do this if he can't find something that fits the bill.

and last but not least:

3) Not use another closed source code management system because he's already gotten his ass kicked by this one.

Oh yeah, and just to join the fray:

Linus:
We told you so.

Re:Three Words

[Duck1123]

I'm sure that'll work great when he ports Linux over to .Net

Re:Three Words

[MyLongNickName]

My thoughts exactly. I heard he is waiting for Longhorn though.

Re:Three Words

[vadim_t]

I hope you're joking. There's no need for Linux client for that monstrosity.

I've tried it myself and it's far, FAR worse than anything else. CVS is wonderful in comparison.

Re:Three Words

[CypherOz]

2) Start a project that creates a new source code management system that does everything that he wants - basically scratch his own itch. He might do this if he can't find something that fits the bill.

You want Linus to waste time on a YASCM? There are many more importrant things to do in the Kernel!

Re:Three Words

[Seumas]

Fortunately, keeping bits will be a much simpler task now that they are standing perpendicular.

Re:Three Words

[Anonymous Coward]

You want Linus to waste time on a YASCM? There are many more importrant things to do in the Kernel!

Easy solution - just integrate SCM into the kernel. Duh.

Re:Three Words

[stuuf]

Maybe, but wouldn't they be falling over all the time?

I hope my dumb off-topic post gets a funny mod too

Re:Three Words

[dj245]

Visual Source Safe!

Not safe enough. Its only a visual safe after all. What this situation calls for is a box that once closed, will remain shut and untampered with. A lockbox, you might say...

Re:Three Words

[Greyfox]

Maybe he could put the hurts on IBM and get them to open up CMVC. I don't believe they're still selling it or anything, so it doesn't really benefit them to keep it closed. And since IBM is nominally our friend and since it would probably be good marketing, I could see them considering it.

The Java CMVC client is actually pretty nice. Sucks less than VSS by a good bit.

Re:Three Words

[DaHat]

Not at all. A source repository is not unlike an e-mail server or web server.

The platform you are running on should not affect your ability to access it.

Imagine a world where you could only access a server if you were running the same client software? That would be a horrible world without interoperability.

Re:Three Words

[Excelsior]

Fortunately, keeping bits will be a much simpler task now that they are standing perpendicular.

You would think, but when bits get exposed to disco all bets are off.

Re:Three Words

[Frank T. Lofaro Jr.]

Heck they did it with a web server.
kHTTPd.

(I better duck)

Re:Three Words

[Cramer]

Go find a copy of Source Off-Site. Then you can hit VSS crap from just about anywhere. Yeap, it sucks as much as you'd think it would... crap ontop of crap is just a bigger pile of ...

I even tried using SS.EXE under linux (with a sutable emulator.) Just don't use VSS. Even CVS is better -- and that's just sad.

Re:Three Words

[Anonymous Coward]

I'm sure VSS is a plot by Microsoft to hold back other software developers. VSS defies common engineering principles that you normally need to make tradeoffs between different software design goals. But VSS manages to combine limited functionality, inefficiency, unreliability, and flakiness all in the same product. It must have take a lot of skilled software engineering to achieve all four of those goals simultaneously.

Re:Three Words

[grozzie2]

Take a look at the license involved with BK for the kernel folks using it. Then you'll know why 2 is an unviable solution. Using the 'free' bk precludes working on any form of version control system, and continues to do so for some time after he stops using it.

Re:Three Words

[Seumas]

Are you insinuating that Linus listens to disco while he codes?

Re:Three Words

[DaHat]

A couple of co-workers use Source Off-Site... and have no end of problems with it, but then it is their only solution for hitting our VSS server without having a tunnel in... which we still have never bothered to build a capability for despite the use.

Re:Three Words

[LoveTheIRS]

haha, that flash was great.

Re:Three Words

[Anonymous Coward]

Visual Source Safe is broken.

Period.

It doesn't work.

If you remove an obsolete file it goes completely, meaning you can't rebuild old releases

You can't branch - you have to create *new* copies of your files in new directories (this idiocy just leaves me speachless)

There are a multitude of bugs where the GUI gives different results to the same actions when performed in scripts

And Microsoft don't use it internally. They use PerForce.

That should really tell you something

*NOTE TO MODERATORS*

[Anonymous Coward]

People who observe that stories concerning bitkeeper and linux have existed before should be moderated "redundant" so that the rest of us don't have to look at them. This is a new story on a new development, and it is difficult to intelligently discuss it in an article crapflooded by people complaining that this is a dupe when it is not.

Re:*NOTE TO MODERATORS*

[Anonymous Coward]

Nonesense. This was done and dusted on LKML 2 days ago.

Re:*NOTE TO MODERATORS*

[Anonymous Coward]

This was posted 3 days ago to the LKML, it's old news.

Re:*NOTE TO MODERATORS*

[Anonymous Coward]

What idiots modded this up four times? Oh well, prepare to get toasted in meta-moderation.

Re:*NOTE TO MODERATORS*

[dsanfte]

Do you want to take the meta-moderation hit for modding them down, then? Any negative mods with Redundant or Offtopic get voted unfair. No matter what.

Re:*NOTE TO MODERATORS*

[A beautiful mind]

Thing is, this "new story" brought nothing new. We knew that he DID drop bitkeeper, from the three 5,informative moderated posts which linked to LKML in the previous story. Now we can read that very LKML announcement in this slashdot story aswell. We knew that he won't pick subversion from the previous story and from the subversion developers aswell.

What is new, that Linus wrote his own SCM (README here)

Maybe it will appear in 1-2 days as another slashdot story?

Re:*NOTE TO MODERATORS*

[MartinG]

What appears in slashdot stories is and should be orthogonal from what appears in the comments.

Not everyone even reads the comments. If there is more news that has not appeared in an article, then there should be an article regardless of whether it has already been discussed in some comments.

Re:How about...

[capoccia]

>How about Subversion?

if you read the article, you would see that linus adresses this.

NOTE! I detest the centralized SCM model, but if push comes to shove, and we just _can't_ get a reasonable parallell merge thing going in the short timeframe (ie month or two), I'll use something like SVN on a trusted site with just a few committers, and at least try to distribute the merging out over a few people rather than making _me_ be the throttle.

The reason I don't really want to do that is once we start doing it that way, I suspect we'll have a _really_ hard time stopping. I think it's a broken model. So I'd much rather try to have some pain in the short run and get a better model running, but I just wanted to let people know that I'm pragmatic enough that I realize that we may not have much choice.

What are the odds...

[Anonymous Coward]

What are the odds those uber-l33t Linux devs will make an open source GPL'd "alternative" to BitKeeper? I read his statement, about it being "the best" and that there will be some migration problems to a new program. I, for one, would think Linus programming for a certain program would, like, totally up its "reliability" 10 fold.

Re:What are the odds...

[bheading]

Great idea. Linus should stop writing kernel code (what everyone seems to think he's good at) and create *yet another* GPL revision control system to add to the 20-odd that already exist.

I wish I could shoot anyone who uses the cliched catchphrase "I for one".

Q & A SCM?

[Anonymous Coward]

Lets start with an easy question.

What qualities does a replacement have to have, and what are the present alternatives missing?

Re:Q & A SCM?

[erikharrison]

The replacement has to be roughly as performant as BK, if possible (so far, not so much), offer the distributed SCM model (several available tools) and hopefully have a stable release (less so).

Re:Q & A SCM?

[tokyopimpdaddy]

As a non-programmer, but a Linux user, and generally intrested in 'the way things are done', I wish the above post had a higher rating, and a few more answers.

Re:Q & A SCM?

[arose]

Linus has to like it.

Re:Q & A SCM?

[Bloater]

Linus needs it to be fast, scriptable, and let him pull changesets. It needs to be a distributed system where everybody has their own version controlled repositories, and the SCM recognises when Linux pulls a patch that he already has and doesn't reapply it. To name just a few things.

Re:Q & A SCM?

[Anonymous Coward]

Sounds like it needs nearest-neighbours code repositories, coupled with a high-level, bi-directional WAN channel.

Re:Q & A SCM?

[SA Stevens]

Linux needs somebody with a Donald Knuth mentality. Someone who will dedicate years on a side-trip to produce the best, most excellent, version controlling system possible. (Similar to Knuth's side-trip into Typesetting that produced the Wonder called TeX).

Re:Q & A SCM?

[Jeremiah Cornelius]

Oh, yeah. We got extra Knuths, just dropping out of the trees.

Re:Q & A SCM?

[tomhudson]

Grandparent post:

Lets start with an easy question.

What qualities does a replacement have to have, and what are the present alternatives missing?

Parent post:

As a non-programmer, but a Linux user, and generally intrested in 'the way things are done', I wish the above post had a higher rating, and a few more answers.

Me: ROFLMAO - "Let's start with an easy question ..."

As a programmer, I'm going to let you in on the BIG secret: getting the initial requirements right is NOT an easy question. It's the difference between success and failure, between on time and late, between on the money and over-budget.

Ask any coder how much time they spent convincing people that what they thought they required was

1. stupid
2. a waste of resources
3. going to cost too much
4. better put off until a later iteration.
5. not practical
6. at cross-purposes to the projects' goal
7. all the above

There's a reason we have terms like "feature creep" and "kill your first baby" (to kill your first baby means "don't get too attached to the code you wrote, because you're probably going to have to throw it all out at some point". Peopel can actually get attached to a chunk of code they wrote, that seemed elegant at the time, but just doesn't cut it, and they won't throw it out because of the psychological investment they have in it)....

Being able to accurately spec out any non-trivial project is not easy.

Re:Q & A SCM?

[Anonymous Coward]

Linus has to like it.

Ah ha! Tove can be the new SCM.

Re:Q & A SCM?

[Frank T. Lofaro Jr.]

Actually you do (if you're in the US)

The USA has a lot of people laid off due to outsourcing and the dot-com bust.

These people literally having nothing better (or at all) to do.

* San Jose and the rest of the Bay Area in particular.

Re:Q & A SCM?

[Minwee]

"Are you _the_ Donald Knuth?"

"No, I'm _a_ Donald Knuth. We come in six packs now."

Re:Q & A SCM?

[Anonymous Coward]

The replacement has to be roughly as performant as BK,

OK, now let me suggest a slightly tougher task than defining what qualities are needed and how existing software stacks up: defining what qualities are needed and how existing stacks up, and doing it with real English words that you can find in a dictionary.

Re:Q & A SCM?

[cakoose]

Actually you do (if you're in the US)

The USA has a lot of people laid off due to outsourcing and the dot-com bust.

These people literally having nothing better (or at all) to do.

person_who_loses_job_after_bubble_burst != donald_knuth

Re:Q & A SCM?

[Libor Vanek]

- handling 10.000s of files in one "release" very very quickly
- "change-set" oriented (change in more files as single transaction)
- support for off-line repositories and easy merging up
- no big stress on handling binary files (the aren't many in Linux kernel)
- easy work with many many (10s) of different versions of single "file tree" with easy solving merging conflicts (imagine having linux-uml, linux-acpi, linux-sata, linux-real-time etc. trees)

Re:Q & A SCM?

[shutdown -p now]

In theory, we could just recycle the existing one.

Re:Q & A SCM?

[Magic Thread]

It's funny you should say that. Have you taken a look at ArX? The author may be no Knuth, but ArX looks quite good. The name is also inspired by TeX.

Re:Q & A SCM?

[Anonymous Coward]

I would say that points directly to Arch, doesn't it?

Re:Q & A SCM?

[Bloater]

There is something I missed off... It should be scaleable wrt history size, so you don't have to have separate histories for each year (which is the recommended practise with arch). It should also not impose such a silly naming policy.

perpendicular bitkeeper

[Doppler00]

Maybe he's looking for a Perpendicular bitkeeper from Hitachi!

Re:perpendicular bitkeeper

[Anonymous Coward]

Yes! That way we can store 10x the data in the same ammount of space. This will solve the 32bit int number problem they are having.

Lets get perpendicular!

Netcraft confirms it

[Anonymous Coward]

Slashdot cliches are dying

Re:Netcraft confirms it

[Eric(b0mb)Dennis]

I, for one, welcome our new karma-whoring cliche mixing overlords.

Re:Netcraft confirms it

[saskboy]

In post 9/11 Slashdot, Cliches about BitKeeper dye you.

Re:Netcraft confirms it

[Anonymous Coward]

In Soviet Russia your new karma-whoring cliche mixing overlords welcome you for one

Ah!

Distributed Version Control with SVK

[Anonymous Coward]

SVK is a very sweet extension to SVN and actually rocks my world! :)

Slashback? Some info on Bit Keeper

[saskboy]

I had no idea what Bit Keeper was, so I did some research. OK, I googled it, but what other kind of research do you do anymore these days?

--
BitKeeper is a CM system from the BitMover corp. http://www.bitkeeper.com/
--
OK, so that first find on google wasn't too descriptive, off to try again...

"Coverage on the bitkeeper announcement from earlier this week is also available." At least the editors admit when they are posting a Dupe not as a Slashback.

Re:How about...

[bonch]

Subversion uses a centralized model. In TFA, Linus talks about his disdain for such a system, and given the distributed way kernel development is done, Subversion would be quite a headache. That's not to say Subversion isn't great for some projects, but certainly not for the kernel.

It really is too bad, though, that with the complaints about BitKeeper before, nobody seems to have written a 100% free equivalent in that time, and there doesn't appear to be a worthy alternative for Linus now that he's dropped BK. Maybe one will pop up, or someone will hack in the needed features into an existing project like Subversion. Here's hoping.

Re:Other reasons...

[Senjutsu]

Bitmover dropping it's free version of Bitkeeper isn't a good enough reason?

dupe, but anyway.....

[diegocgteleline.es]

..I hate how everybody has critized Linus (specially the RMS/GNU front), when he never forced people to use BK.

I mean, when Linus started to use BK he promised things would be at least as good as they were before - and it's true (they've been better in fact), people still gets -rc's in GNU diff format at kernel.org. The official way of distributing patches has always been "clean", I don't know why people whined so much about BK, it's OK for me if Linus wants to use a propietary tool himself, as long as I'm not forced to use it. I've certainly not used or needed to use it for years, and I'm one of those people who tries -mm and -rcs all the time...

(and those who claim that people should behave differently and "give example" just because they're "leaders" can go to hell)

Re:dupe, but anyway.....

[bonch]

Certain people criticized because they view Linux as the golden child of the free software movement, and to have it "tainted" by commercial development software just wasn't politically correct enough for them. Like you, I never understood the big deal--McVoy made a kick-ass commercial product and gave it to the kernel developers for free. No harm done. But what happened? The community bitched and bitched at BitMover, tried reverse-engineering it several times, and made it not worth BitMover's while to provide a free version anymore.

I read a comment from someone that brought up a great point. If BitKeeper had been 100% commercial and Linus decided to buy a license to use it, there wouldn't have been as many complaints. But because it was commercial yet also free to the kernel developers, people complained, tried to reverse-engineer, and so on.

It should be noted that McVoy offered to comp the licenses for Andrew and Linus if they left OSDL. I guess Linus decided it wasn't worth it to be caught in the middle of this anymore. The struggle between the so-called RMS-wing of OSS and the practical Linus-wing of OSS is eventually going to have to resolve these differences. Linux's development pace doubled with the use of BitKeeper, and here's hoping it's not affected by a tool switch.

Re:dupe, but anyway.....

[BlueCodeWarrior]

RMS was just here (University of PIttsburgh...first time he'd been here since the CMU printer incident...) on Thursday, and what he had to say about it was this: It sends out mixed messages if we mix proprietary and Free software. If we're using Free sofware because software ought to be free, then using proprietary software doesn't make sense. It's like saying that you're going on a hunger strike to protest something...but eating a few crackers every day or so. This is what he said, I don't necisarily agree or disagree.

Re:dupe, but anyway.....

[Anonymous Coward]

Yes he DID force other people to use Bitkeeper, at least if they wanted to continue development on the Linux kernel. Eventually a CVS-to-Bitkeeper gateway was cobbled together, but for a time you HAD to use it if you were a kernel developer. Forcing hundreds of people to use a proprietary tool that many had problems with IS a reason to criticize Linus.

Re:dupe, but anyway.....

[diegocgteleline.es]

If we're using Free sofware because software ought to be free, then using proprietary software doesn't make sense. It's like saying that you're going on a hunger strike to protest something...but eating a few crackers every day or so. This is what he said, I don't necisarily agree or disagree.

Then RMS should apply that to himself because he needed to use non-free software when he started to develop GNU. And until linux was there, I guess he couldn't use a free kernel either.

BK is not different from this particular bit of GNU history. Linus started to use BK because he needed it. Not using BK (or other propietary tool) and using a free (but inferior) SCM would have mean that the kernel development rate would have slowed down, and it'd have been worse for the free software (we) overall, even GNU people. I mean, why a propietary tool has always to be harmful? BK has been certainly good for the linux kernel, and the fact that linus uses it has not taken away my freedom...

Re:dupe, but anyway.....

[ckaminski]

Development pace doubled because of BitKeeper, or because of moving to a distributed SCM? Tough question, one which Linus will surely find out soon enough.

Re:dupe, but anyway.....

[Anonymous Coward]

Linus actually doesn't give a damn if it's "free" or not. The below quote is from Wikiquote:

"Anybody who tells me I can't use a program because it's not open source, go suck on rms. I'm not interested. 99% of that I run tends to be open source, but that's _my_ choice, dammit." (October 26, 2004) ~Linus Torvalds

Re:dupe, but anyway.....

[Samrobb]

It's like saying that you're going on a hunger strike to protest something...but eating a few crackers every day or so.

Wow! Are you telling me RMS knows Vince Eirene?

(OK, OK, it's an obscure reference. Anyone who attended CMU in the late 80's - early 90's should get it, though.)

Re:dupe, but anyway.....

[jarich]

If we're using Free sofware because software ought to be free, then using proprietary software doesn't make sense.

Exactly!

I'm using free software for two reasons. One, it works. Two, it's free.

If software ought to be free, then so should furniture and cars.

Cheaper? Sure! Free? No.

Re:dupe, but anyway.....

[bubkus_jones]

Yes, you should use the best tool for the job, but if you're developing free/open source software, and a large component of your development rig is a proprietary piece of software, what kind of message does that send to other developers, or to the public? It's not so much that it didn't harm development or the end product, it's that it could give the impression that even the "big shots" of the OSS world still use proprietary products because they're "better" than any available OSS project. It says that the OSS "equivalents" were not good enough, and that can reflect negatively on other OSS projects.

There should have been an organized effort at creating an OSS project that would work as well (or, preferably better) than BK. If there wasn't an OSS tool that was the best choice, one should have been developed.

Re:dupe, but anyway.....

[bobbyjack]

Furniture and cars ARE, essentially, free - to 'modify' and 'distribute'. You're using free software because it's gratis, not because it's non-proprietary; otherwise, you agree with the argument...

Re:dupe, but anyway.....

[BlueCodeWarrior]

You're confusing Free as in speech and free as in beer. I'll revise the sentence I wrote that you quoted to "If we're using Free software because software ought to be Free, then ..." One of the few typos in that post...

Re:dupe, but anyway.....

[jcr]

Thanks for another example why I shouldn't take Richard seriously.

-jcr

Re:dupe, but anyway.....

[Cramer]

Both -- as they are the same thing. BK brings conciderable productivity to the process. No open source tools provide anything close to what BK is doing. Linus knows this already, but he's been left no choice. ('tho, through no fault of his own.)

Any SCM would help vs. the decade old process of patch and manually resovle the conflicts -- and there are always conflicts. This is simply because it forces a common ancestry in every development arm. diff and patch don't carry any history or version graphs. BK takes this a little too far, IMO, by hanging everything off change sets -- a single change suddenly is dependant on every change in every file that comes before it. (This is good for commercial development as every tree is supposed to be the same. However, for the linux kernel, there are too many side branches with changes unfit for the main branch that prevent easy patch integration -- main can never pull patches from a diverging side branch.)

Re:dupe, but anyway.....

[Cramer]

• There should have been an organized effort at creating an OSS project that would work as well (or, preferably better) than BK. If there wasn't an OSS tool that was the best choice, one should have been developed.

That's just it. The OSS world never creates one wheel, and rarely finishes one before embarking on the next.

Linus chose BK because it was the only thing available at the time to do what was needed. It's not about one being better than another. It comes down to meeting a set of criteria. Several years (and a lot of bitching) later and we're still at the same spot... there still aren't any OSS projects to meet all of the requirements. BK/Free has been around for 5 years; Linus has been using it for 3 years, so where's all the alternatives from those that have been bitching constantly against the use of BK? Well? I'm f***ing waiting. If you don't have something that'll do the job, then you should shutup and let me use what ever'll do the job. That's the thing the opponents refuse to accept. ("I don't need to invent a wheel; Larry has kindly give me one of his.")

Re:dupe, but anyway.....

[Cramer]

Wrong, o' annonymous asshole. One can continue to get patches and tarballs from the usual place(s) and send patches to maintainers and lists the same as has been done for over a decade.

There is absolutely zero requirement for anyone to use bitkeeper. It's easier for Linus to bk pull from someone's tree but there was never a requirement made of anyone to do it that way. Plus, the main tree is only writable by Linus (and maybe a few others), so one cannot bk push their changes into the main tree anyway. (They still have to be sent as a patch or pulled by Linus.)

So, I'll say it again:

• There is
• absolutely zero requirement for anyone to use bitkeeper.

Re:dupe, but anyway.....

[aichpvee]

Yes, you should use the best tool for the job, but if you're developing free/open source software, and a large component of your development rig is a proprietary piece of software, what kind of message does that send to other developers, or to the public?

I would think that the message would be to put up or shut up. If RMS cares so much, wtf is his tool for doing what needs to be done? I seem to see him whining about how Linux should be misspelled "GNU/Linux" more than working on anything particularly important.

Re:dupe, but anyway.....

[Anonymous Coward]

Asshole?

I don't agree with the parent, but I don't resort to name calling either.

Grow up!

Re:dupe, but anyway.....

[Anonymous Coward]

Just because he called him an asshole (and I'd agree with his assessment) doesn't make him a child. He did it to be antogonistic, a deliberate strategy to piss the poster off. Just because you think that's immature doesn't make it so.

Re:dupe, but anyway.....

[aichpvee]

I use free software because I like it better. Doesn't mean I'm going to use a toy to do a Man's job.

Why do you have to agree with that argument in order to like open source software when you aren't in it just for the price?

RMS's argument is completely ridiculous, no matter how right in certain areas. But sometimes open source just isn't there, or maybe it can't be. Saying that one should use inferior products if the disparity in quality is significant just because they are "Free" is what "doesn't make sense."

It's similar sorts of zealots who always say "it's open source, fix it yourself." So it's time to put up or shut up. Which hopefully won't be too hard, since if McVoy can do it, I'm sure there are tons of people who can put together something similar to BK with an open source license, though it does make one wonder, if that is true, why they haven't done it yet.

Re:dupe, but anyway.....

[Anonymous Coward]

He did it to be antogonistic, a deliberate strategy to piss the poster off. Just because you think that's immature doesn't make it so.

Interesting... So exactly what mature reason could the poster have had for concocting a deliberate strategy to piss off a guy on Slashdot?

[I'm not the poster of your parent post. I haven't posted in this thread]

Re:dupe, but anyway.....

[aCapitalist]

The struggle between the so-called RMS-wing of OSS and the practical Linus-wing of OSS is eventually going to have to resolve these differences.

There are fundamental differences that most likely won't be resolved any time soon (as long as RMS is around).

Here's an excellent writeup about the different philosophies of the RMS-wing and the OSS-wing of FOSS

http://linux.omnipotent.net/article.php?article_id =12503

Re:dupe, but anyway.....

[aCapitalist]

If we're using Free sofware because software ought to be free, then using proprietary software doesn't make sense

Who is we - Stallman and a few thousand of his fellow extremists? He doesn't represent Open Source.

And the fact of the matter is that open source is growing up (Linus is 35 now), and people are realizing that proprietary and open source are the best for choice...something that Stallman hates.

Stallman's silly analogies are always weak except to weak-minded people that follow his dogma.

Re:dupe, but anyway.....

[mrchaotica]

Then RMS should apply that to himself because he needed to use non-free software when he started to develop GNU. And until linux was there, I guess he couldn't use a free kernel either.

If RMS had refused to use non-free software, he would have to program the entire system by literally twiddling in bits one by one. In contrast, if Linus had refused to use BitKeeper he'd be inconvenienced, at the worst. Linus used BitKeeper because it was the best tool for the job, but RMS used non-Free software because it was the only tool for the job. I believe that if RMS could have bootstrapped GNU any other way, he would have.

Re:dupe, but anyway.....

[mrchaotica]

If software ought to be free, then so should furniture and cars.

Do you mean free as in speech, or free as in beer? If you mean free as in speech (capital-F Free), then furniture and cars are Free -- you can look at a chair or a car, figure out how it works, and make your own compatible one and it's perfectly legal. If you mean free as in beer, then yes, furniture and cars ought to be free, once they can be duplicated for zero cost, like software can. But since we don't have Star Trek-style replicators yet, it's reasonable to charge money for them.

Re:dupe, but anyway.....

[Anonymous Coward]

One can continue to get patches and tarballs from the usual place(s) and send patches to maintainers and lists the same as has been done for over a decade.

What about maintainers ? They do have to use bk

Re:dupe, but anyway.....

[bhalo05]

Who is we - Stallman and a few thousand of his fellow extremists? He doesn't represent Open Source. Of course! He has said it himself. That's why he always talks about Free Software.

Re:dupe, but anyway.....

[Cramer]

• What about maintainers ? They do have to use bk.

No they don't. And in fact, many don't. But, yes, at some point patches have to reach someone who does (obviously.) However, that bk user can be Linus himself.

Just to be a data point, I use bitkeeper. I like it and have no problem with the "free license." (I use it for more than just the linux kernel, btw.) However, every change I've made that has been sent "up the line" was done so as a gnu patch via email. (No one has ever pulled from my repo(s), because they can't -- there are other changes in there they don't want that cannot be selectively ignored.)

Re:dupe, but anyway.....

[Anonymous Coward]

If RMS cares so much, wtf is his tool for doing what needs to be done?

GNU Arch

Re:dupe, but anyway.....

[bubkus_jones]

I think we're both saying basically the same thing. People complain about there being no OSS tool that'll do the job, but no one has made (from what I can tell) any serious effort to develop one.

I'd just like to know why.

Re:dupe, but anyway.....

[jarich]

But it's not free to ~create~ software, only to copy.

I think that software companies are getting rich by exploiting this fact and people like RMS are much more fringe that they could be by ignoring it.

We probably should be paying less for software, but just like cars and sofas, someone (or a team of someones) has to create the item in the first place.

Re:dupe, but anyway.....

[mrchaotica]

The volunteer and patronage models seem to me to do a good enough job of creating software...

Re:dupe, but anyway.....

[kjots]

'Cause most people prefer to complain then do any actual work. Sad, really.

Re:dupe, but anyway.....

[jarich]

Man, are we wandering off on a side trail or what? :)

Reminds me of a Mark Shuttleworth quote from a recent article http://interviews.slashdot.org/article.pl?sid=05/0 4/04/1859255&tid=90&tid=163&tid=160&tid=11

But YOU try motivate someone to hack on a sewerage management system in their spare time.

His point was that there are lots of projects that people ~want~ to work on, but there are many more that you just have to pay people to write.

I'm not flaming, just asking... do you really believe that ~all~ software should be free or are you trying to make a philosophical point?

Re:dupe, but anyway.....

[mrchaotica]

His point was that there are lots of projects that people ~want~ to work on, but there are many more that you just have to pay people to write.

I completely agree! But just because you're paying somebody to write the software doesn't mean you can't give it away for free; that's the "patronage model" I alluded to. Take the sewage management system example: the public utility isn't a software company, and it's not as if keeping their system proprietary gives them some sort of competitive advantage, so there's no reason why they couldn't pay programmers to make the software, and then give it away. It's basically normal in-house software development, except with a Sourceforge page.

In fact, patronage is conveniently complimentary to volunteerism -- any software with a wide audience would find sufficient volunteers to build it, and software with a narrow audience would have patrons sufficiently interested to fund development.

I really do believe that ~all~ software could be free, and we'd be perfectly fine. The only place where I think proprietary stuff is necessary is games, and I only believe in proprietary artwork, not code. I doubt that commercial software will ever actually go away, though.

Re:dupe, but anyway.....

[jarich]

there's no reason why they couldn't pay programmers to make the software, and then give it away.

Only if there's hardware involved. I work at SAS... our software ~is~ our product. It's not like the sewage plant.

We pay ~lot~ of people money to do a lot of work. Giving away the source would bankrupt us fast.

The only place where I think proprietary stuff is necessary is games

Hmmmm... couldn't you substitute any pet area? I know a few people who might say statistics instead of games. Someone else (at a database company) might say indexing routines.

In my opinion, there's good creative ways to work in every field. There are various ways to reward people. Money is good for me, but I also contribute to some open source projects. I like to give back as well. I guess I'm a hybrid developer. :)

I guess what I'm trying to say is that I think having a mix is the best way. Some people make a very good living writing software and a lot of people write software for other rewards.

I doubt that commercial software will ever actually go away, though

I suspect you are correct. However, as long as their is a Microsoft making a ton of money, there will always be open source alternatives nipping at their heels. :)

Re:dupe, but anyway.....

[mrchaotica]

Only if there's hardware involved. I work at SAS... our software ~is~ our product. It's not like the sewage plant.

We pay ~lot~ of people money to do a lot of work. Giving away the source would bankrupt us fast.

Well, that's my point -- companies like yours would go out of business. But that's okay (for the industry in general, not for you!) because that kind of software would still get made -- just in-house by the kinds of companies that are currently your clients. It may sound less efficient, but the companies could work together and the software would be Free (which makes it better, in my opinion).

The only place where I think proprietary stuff is necessary is games...

Hmmmm... couldn't you substitute any pet area? I know a few people who might say statistics instead of games. Someone else (at a database company) might say indexing routines.

No, because I'm not talking about the code at all. I don't believe the fancy 3d algorithms Carmack invented have any more right to be proprietary than statistical or indexing ones.

No, instead I'm talking about the artwork and the story -- the non-code parts of the game. The reason for this is that to make a coherent, complete game with a good story you need to have one single vision that everyone involved needs to work towards. That seems difficult to achieve with "software by committee." That's the only reason I make an exception for games. And Creative Commons-licensed content would be infinitely preferable, of course!

Also, making the game code Free is important -- games are the only type of program that doesn't become obsolete, specifically because it does tell a story. How many thousands of games out there are still loved even though they can pretty much only be played through emulation? Think how much better it would be if all that abandonware was Free Software, so that it could be ported to modern operating systems and get upgraded with modern technology. Wouldn't it be great to patch Starcraft to run at resolutions higher than 640x480, for use with modern laptops? Don't you wish all games were updated the way Half-Life:Source was?

Some people make a very good living writing software

This can be achieved with the models I've described! In fact, that's what most of the software industry is already -- if you write software that your company doesn't sell, you're already working under the patronage model. The only difference is that the software isn't as useful as it could be, because people outside your company don't have access to it. If it's not something that your company depends on for a competitive advantage, there's no reason why they couldn't just slap up a Sourceforge page and call it a day.

Re:dupe, but anyway.....

[jarich]

The breakdown I see in this model is the companies who need the same software packages tend to be competitiors. Since they compete with each other, they don't want to give the opposition equal footing. So they would keep their code in house, everyone would duplicating the same work, and someone (like SAS) would show up to sell everyone software and we're right back where we started.

Re:dupe, but anyway.....

[pthisis]

What about maintainers ? They do have to use bk

Just to back Cramer up here...

Alan Cox never used Bitkeeper and yet he continued his role as one of the core maintainers/developers.

(I hate the BK free license and am not surprised it eventually caused a rift, but let's be factual--Linus was DEFINITELY against any system that locked anyone else into using non-open-source tools).

Re:How about...

[RalphLeon]

I was thinking at comment #4, wow I should mention subversion, but clearly something as wonderful as subversion is the first thing on anyones mind that has used the SCM.

The biggest advantages:

• It can move folders what a novel concept? (looking in CVSs direction).
• Good support for local repositories or remote ones manages through Apache or SSH
• Most (good) web viewers for SCMs support it
• Great intuitive clients for Mac OSX, Windows, and *nix
• Easy conversion from CVS to Subversion

There are lots of other good things to say but did I mention that it moves directories? I mean you don't have to go onto the server and manually move all of the files therefore screwing up your teams local checkouts! (If you can't tell I am a little bitter about CVS)

Re:How about...

[wakejagr]

From the start of the thread:

PS. Don't bother telling me about subversion. If you must, start reading up on "monotone". That seems to be the most viable alternative, but don't pester the developers so much that they don't get any work done. They are already aware of my problems ;)

RTFA!

[Anonymous Coward]

Linus is looking into Monotone as you can see here.

Might want to start figuring out how to migrate to that, instead of pestering him about subversion.

Re:DUPE

[Anonymous Coward]

No, in fact, we are talking about Linux, as in the kernel development.

Linus happens to head Linux, and made the call, but the entire team (And thus Linux) will be dropping BK, not just Linus.

Re:Slashback? Some info on Bit Keeper

[mcc]

Perhaps it would be helpful if you would read the Wikipedia article on Bitkeeper, which explains what Bitkeeper is, why it is relevant to the linux kernel, and why its relevance to the linux kernel might be controversial enough to make it a slashdot news item.

Maybe the slashdot eds should start [?]-linking Wikipedia articles in story blurbs the way they used to with Everything2...

CVS?

[MHobbit]

Why not use CVS? CVS is a very mature method of source repositories, and has proved itself many times. Though Subversion may hold potential, why not forego the wait and use something that is used widely, especiall on SourceForge.net?

Re:CVS?

[diegocgteleline.es]

Why not use CVS?

Now that's a great idea. I'm sure Linus didn't even look at that option when he decided to use a SCM!

Re:CVS?

[Taladar]

Subversion is basically CVS with some problems fixed (e.g. moving of directories is now possible). The Subversion people say they wouldn't use Subversion to manage the kernel...

Re:CVS?

[Senjutsu]

CVS's centralised repository design is more or less exactly the opposite of what Linus wants when it comes to a version control system. No distributed repositories, no foreign branching, no dependency tracking between changesets, no atomic commits, expensive branching, etc, etc.

CVS may be the solution to some things, but it's not a solution to the problem Linus is trying to solve.

Re:CVS?

[mpcooke3]

i dunno about linus but some issues i have with CVS:
no support for atomic commits of sets of changes
no support for rolling back change sets.
it only really supports files as the basic item you are checking in.
very slow to do complete diffs, no silent background syncing whilst you work.
Handles directories in an odd way

On the plus side it is better than Visual source safe. but then, what isn't.

Re:CVS?

[ckaminski]

CVS isn't the solution to anything, anymore. It's horribly broken file based semantics suck. Actually, I take that back. CVS would be good for managing your /etc directory, since most files in /etc don't have many dependencies on other files in /etc, so your need for an atomic multi-file commit is lessened.

But for most projects, CVS will be supplanted by subversion over the next few years. Now that subversion has a filesystem based repository, and not only a BDB based one, I expect the process to accelerate. The only holdouts will be those few CVS users who insist on tweaking the RCS files themselves, which only is useful for those instances when CVS breaks, which is what Subversion replaces...

So CVS in two years will be for the text-file pedants.

Re:CVS?

[SA Stevens]

CVS is the basis for a LOT of Open Source development projects. I know NetBSD makes good use of it. I suspect the other BSDs use it as well.

I've used it professionally in highly critical environments. People figure out how to use it and it works.

Arguements pro- and anti-CVS almost always boil down to politics. Please identify yourself as one of the partisans (you have already, actually)

Re:CVS?

[statusbar]

The thing that burns me about cvs is the lack of proper handling of branches & dates at the same time for cvs updates and diffs:
Here is more info
That alone is good enough for me to switch to subversion, but I've found the switch to subversion (pre ffs) really problematic and so s-l-o-w (almost unusable) via the apache module.
--jeff++

Re:CVS?

[bani]

But svk would probably be just fine.

Re:CVS?

[Anonymous Coward]

What do you think (Score:5, Funny) means?

Re:CVS?

[ckaminski]

What burns me about CVS is changesets. The idea that my touching 6 files == one change. CVS doesn't have this without resorting to tags. That's why I hate it. I use it for projects where most files are independant of others, such as managing my management tool tree or my /etc files. For my Java and C++ work, I've moved to subversion.

There's no politics here for me. I like both, I use both, but CVS is getting surpassed. Without some serious development work by the maintainers, I'll probably be off of it by years end. With cvs2svn floating around, and some work around the edge cases, the migration from CVS to Subversion will be unparalleled in simplicity. Hell, it's practically command (line) compatible.

The real politics is around CVS/SVN or Clearcase/Perforce. That's where 90% of the resistance to CVS comes from in corporate environments...

How will Linus keep his bits now?

[mrRay720]

I think he needs to research this now. Those bits don't keep themselves, you know.

Re:How will Linus keep his bits now?

[Narishma]

He should teach his bits how to get perpendicular.

Other reasons...Oath breakers.

[Anonymous Coward]

"Bitmover dropping it's free version of Bitkeeper isn't a good enough reason?"

Because some people can't "play by the rules".

Re:Other reasons...Oath breakers.

[Rakshasa Taisab]

Why arn't you playing by *my* rules?!? Shut up and give me all your money.

Re:Other reasons...Oath breakers.

[Anonymous Coward]

Because some people can't "play by the rules".

It's pretty tough to play by the rules, when you're not told what they are.

Re:Other reasons...Oath breakers.

[Webmonger]

I'm not sure who you're making fun of there. Bitmover for making people pay now that folks are trying to reverse-engineer bk, or the Open Source community for expecting companies to GPL all their software and hand it out free?

Re:Other reasons...Oath breakers.

[Anonymous Coward]

There are reasonable rules and unreasonable ones.

For a software license, prohibition of reverse-engineering is unreasonable and unenforceable in some parts of the world.

Prohibition of competition is even more ridiculous in a software license, especially without properly defining what constitutes competition.

He seems to be writing his own

[iabervon]

Primarily due to frustration with the speed of existing solutions, and the fact that they do badly with huge numbers of tiny changes to large projects (which is what he deals with), he's started writing something that he claims isn't an SCM. What it does is store versions of trees of files and allow them to be annotated. This is different from an SCM in that it lacks support for merging changes to get a version that nobody created before out of multiple versions that people created independantly. This makes it a history archival program rather than useful for collaboration directly.

On the other hand, it's self-hosting and sparse (the previous thing he wrote when confronted by the inavailability of something he wanted) is also available stored in it. Probably, be the middle of next week (if not the end of the weekend), he'll have if set up as sufficient for his purposes, and other people will be filling in support for the way they work.

Re:He seems to be writing his own

[Anonymous Coward]

That's what Prof. Knuth would've done. Take a few years off and solve the SCM problem first, that OS stuff can wait.

Re:He seems to be writing his own

[Anonymous Coward]

Isn't that exactly what SVN does?

SVN manages a history of trees, and has really no merge capability beyond what you can do by hand with diff and patch.

But I would argue that version control is NOT about the "nodes" on the graph of changes (like subversion), it's about the "edges" on the graph (the changes between nodes).

Think about it, which is easier to work with: 1) apply the differences between history state 234 and history state 628 to the files in history state 934, or 2) apply changeset #85 to the files in history state 934. Especially when you have hundreds of changesets to work with, coming from all kinds of different directions.

This is why Darcs or Arch will be the foundation for the next-generation version control, not CVS or SVN (which are basically the same).

GNU Arch?

[jameson]

Does anyone happen to know whether GNU Arch has been considered? I've been using it for a while and find it quite good (it's not perfect, but it's the best versionning system I've used so far).

Re:GNU Arch?

[Anonymous Coward]

Yes it has. Not much was looked into it because it was amazingly slow for a tree the size of linux.

Re:GNU Arch?

[Bloater]

> Does anyone happen to know whether GNU Arch has been considered? I've been using it for a while and find it quite good (it's not perfect, but it's the best versionning system I've used so far).

Then you haven't tried darcs

Re:GNU Arch?

[jameson]

Wow, darcs looks really nice-- the "local copy=repository" makes perfect sense. I didn't know of it, but will give it a try. Thank you!

Re:GNU Arch?

[Anonymous Coward]

Note that Monotone is another very similar to Darcs but with more features. I don't know about the performance of either.

Re:GNU Arch?

[streak]

I believe darcs ended up being slower than arch on a tree the size of the kernel. See LKML SCM thread and reply messages for some concerns about darcs.

Re:GNU Arch?

[Anonymous Coward]

I don't know, but Arch is way too complex. The learning curve is too steep. It has like, what, 185 commands??? I started using it, went through a tutorial, and gave up, about 3 times. You have to do all these steps for the simplest activities. And the naming conventions for files are just awful.

Arch may be a wonderful system technically (and I find I agree with a lot of Tom Lord's rants) but the UI is inexcusably bad.

Then somebody on /. wrote "try Darcs. It's like Arch but simple."

And finally I did. WHOA! Darcs kicks some serious ass. It has an incredible elegance and tight set of commands. It also has a theoretical foundation which I like a lot (I'm one of those people that cries about the lack of truly relational databases for instance, because the theory is so wonderful). I hope Darcs continues to gain users and features.

The problem with Darcs is the same with Arch: too slow (there is about zero optimization, which is good because premature optimization is evil, but bad because Linus needs something NOW).

If only BitMover waited 1-2 years to turn evil, eh?

Re:GNU Arch?

[elbobo]

I'm not sure what you're on about. My daily arch use all cycles around only a few central commands

tla missing
tla changes
tla update
tla commit
tla star-update

It's really no complex at all. It just takes a little while to comprehend arch's version control model, if you're coming from something like CVS or SVN, which function somewhat differently.

Re:GNU Arch?

[nathanh]

Does anyone happen to know whether GNU Arch has been considered?

,,Oh+God@I_hope_he,,doesnt+use++arch.

=I,couldn't++bear=the=pain.

Re:GNU Arch?

[Elwood P Dowd]

Yes, he considered GNU Arch. Linus knows as much about OSS version control as anyone on the planet.

I believe the main problem with GNU Arch is that certain operations are way too slow, especially for a project the size of the kernel. Linus said in one of his emails that Monotone looked like the best contender, and the Monotone website says that they use an essentially different paradigm from Arch. Iduno what they might mean by that, but it might be another thing that won't fit for the Kernel with gnu arch.

So yeah, he considered it. There is definitely no OSS bitkeeper killer yet.

Re:GNU Arch?

[Anonymous Coward]

Normally a post like this would be moderated 'funny', but for that to happen, the moderator would have to have arch experience and both arch users are out of mod points today.

Re:GNU Arch?

[Bloater]

Yes, I knew it was slow. Thanks for the link

Re:GNU Arch?

[Nevyn]

Well, I'd say it's pretty much required to use "tla changes --diffs" instead of without the option. And I'd have to add:

tla add
tla tree-lint
tla mv
tla rm

The later two, not so much ... but still. Saying that, I've been pretty happy with Arch.

Re:GNU Arch?

[elbobo]

Right :) I usually do a "tla changes" then right after do a "tla changes --diffs", for clarity. And anything that's listing patches, I'll add "-s" or "-s -c" on the end.

But really it's not that complex. I've also been quite chuffed with it.

My only complaint is that arch doesn't do anything to ensure its own files stay in a permissions state that'll be acceptable to all people committing to the tree. So I occasionally run into "permissions hell" when working on the same branch with another developer.

Re:GNU Arch?

[Nevyn]

My only complaint is that arch doesn't do anything to ensure its own files stay in a permissions state that'll be acceptable to all people committing to the tree. So I occasionally run into "permissions hell" when working on the same branch with another developer.

Intersting ... I'd have assumed for multiple core developers you'd just have a release repo. and one developer repo. each pulling changesets into each others repo. and doing a single pull into the core repo. every now and then (probably with a testing repo. in between). Did you try that?

That's kind of how I work, as a single core developer ... the release repo. is just a branch with remote tags

Re:Slashback? Some info on Bit Keeper

[Anonymous Coward]

Following up as AC so as not to Karma whore:
Welcome to BitMover! We build and market enterprise level development tools for software, hardware, and web developers. Our flagship product is BitKeeper, a reliable, powerful, and distributed configuration management system. BitKeeper is supported on most platforms, such as Microsoft Windows, Linux, Unix, and MacOS/X.

Re:In post-9/11 Soviet Korea

[Anonymous Coward]

Linus Torvalds, naked and petrified, covered in hot grits

Dupe

[Anonymous Coward]

In the post 9/11 world Zonk imitates Taco.

Re:In post-9/11 Soviet Korea

[natrius]

In post-9/11 America, the Department of Homeland Security has noticed an increased level of chatter on the terrorist communications lines. Among the discussions has been unusually prevalent talk of cliche enrichment. We suspect that these terrorists have allied with "Korea", "Soviet Russia", and "Natalie Portman covered in hot grits" to obtain the sophisticated equipment necessary to enrich cliches. We have no choice but to amend the Constitution to disallow speech that contains cliches.

To enforce this amendment, any discussion involving cliche proliferation will be punished with negative moderation. For good measure, the poster will also be sent to Guantanamo.

open source

[puffy311]

If they had used an opensource SCM from the start, they would not have to worry about changing things around now.

Re:open source

[Jeremi]

If they had used an opensource SCM from the start, they would not have to worry about changing things around now.

If a comparable open-source SCM had existed at the time, I'm sure they would have considered it. (and no, CVS didn't count as comparable :^))

Re:open source

[Nasarius]

Yes, but there weren't any open-source SCMs with equivalent functionality when Linus made the decision to use BK. Unfortunately, there still aren't any. Hopefully this need will lead to significantly better open-source version control systems in the near future.

Re:open source

[iabervon]

If they had used an open-source SCM from the start and not changed things around, they'd be using RCS, which is totally useless for distributed development on this scale. If they had used an open-source SCM from when they started using an SCM, they'd be using CVS, which isn't much more applicable to the actual problem. BitKeeper was something that worked through the period when the available systems weren't really appropriate, and served as something to use while the problem space become sufficiently well understood that it could be solved correctly.

Re:open source

[harlows_monkeys]

If they had used an opensource SCM from the start, they would not have to worry about changing things around now

If they had used an opensource SCM from the start, we'd still be somewhere in the 2.4 kernel, and Linus would be insane. Linus picked BitKeeper because it was the only SCM system available that could handle the job.

Re:open source

[Anonymous Coward]

if a comparable open-source SCM had existed at the time, I'm sure they would have considered it.

"they" is not correctly used there. Linux Torvalds is a single person.

BitKeeper Website

[dduardo]

So when are they going to remove the quote by Linus?

"BitKeeper has made me more than twice as productive, and its fundamentally distributed nature allows me to work the way I prefer to work - with many different groups working independently, yet allowing for easy merging between them." -- Linus Torvalds

Re:BitKeeper Website

[Otterley]

Why would they do that? It's not possible to "un-say" something.

Re:BitKeeper Website

[Anonymous Coward]

probably never, it is political why he is stopping using it, nothing to do with it making him not productive. why should they remove it?

Re:BitKeeper Website

[diegocgteleline.es]

Never?

The quote continues being true. IMO the reason why Linus is dropping BK is the license and flamewars - BK is great, but maybe the free alternatives have got better in those 3 years and now they're "good enought" to use with the kernel. It'd be certainly better if bitmover would make BK free, but that's not going to happen.

Re:BitKeeper Website

[tftp]

There is no need to remove the quote, it is correct. The BK software works fine. The problem is that Linus is not happy about BK licensing any more.

Re:BitKeeper Website

[damgx]

Why should they? The statement is stil valid.

The reason for not using BitKeeper is because of philosophy mostly within the community (my guess), not because something is wrong with the tool.

What BitMover has done is change the price of their product so to speak. Alot of Open Source / Free Software people didn't like the old price, and don't like to pay the new price anymore. That is why the tool is droped.

Re:BitKeeper Website

[Anonymous Coward]

They should recalculate to include the productivity lost while searching for a replacement and migrating.

Re:BitKeeper Website

[Cylix]

Originally... it was a philosophical arguement, but when push came to shove it was the right tool for the job. This quelled things in the past and toned down the flame wars.

Skip to the present, someone in their spare time, who was contracted by OSDL, was working on a competing open source client or implentation. This in itself is not wrong at all, but the terms of their license say this is a big no no.

Some time ago, when such an arguement came up in the past the BK guy said they would just keep changing the protocol and breaking it, but that apparently wasn't really enough of a deterant.

Suffice it to say, their product is their livelyhood and in the end it was cost effective to drop linux support. In my own thoughts, dropping the linux side of life really means there is little reason for the competing faction to create an open source client. (I'm really hard pressed to remember if it was a client or not)

In the interm, they have offered free licenses for a good chunk of developers, but no one at OSDL. (if linus leaves, he qualifies for a free license).

Unfortunately, this is from memory and a little of it is mixed in with my interpertation. The original stated motives for dropping linux support in the long term were profit driven, but if you think about it you may reach the same conclusion I did.

Honestly, I don't really feel BK did anything really wrong here. It's a closed source proprietary product with a few open sourced pieces. Someone might offer some history on the matter, but they may owe their roots to linux. Still, the old saying comes to mind, "You knew I was a snake when you took me in." (Yeah, I know that has a negative connotation to it, but the idea is that when it's someone elses court it's their rules)

All in all, makes for a good book... can't wait to see the ending.

Since this is from memory... any updates or additions would be helpful. Either that or you can read all the threads elsewhere... it's not like these were secret phone calls.

Re:BitKeeper Website

[streak]

Its not that Linus wasn't happy with the BK licensing, its that there were issues involved with developers (specificially a contractor under OSDL) trying to to reverse-engineer some of the features in BK.
Even though an agreement was reached that this developer would stop doing this, apparently he continued.
Larry McVoy (and BitKeeper) responded by saying that they were removing the Free BK license, and employees of OSDL (which Linus is) were not eligible for a free license under any conditions.

At this point Linus really didn't have a choice whether to switch SCM systems or not.

Kerneltrap story.
Read the 2nd paragraph under "Free Versus Free".

Re:BitKeeper Website

[Frank T. Lofaro Jr.]

Yeah, nothing free makes any economic sense whatsoever.

Like Google, providing a free web search service running on a free OS (Linux) using free/open protocols (HTTP over TCP over IP) and serving pages in a free/open standard (HTML).

And no one used Sendmail and BIND, those free programs aren't used at all by any one with any economic signifcant right?

(Note, I was being sarcastic - just thought I'd make that clear before someone with an itchy trigger finger mods this post as "Flamebait".

time to roll your own...

[radarsat1]

looks like the kernel developers will have to take some time out to develop their own bitkeeper.. ah well, a bit of variety will probably be good for them.. :)

Bitkeeper Drops Linus

[Anonymous Coward]

Shouldn't the title be "BitKeeper Drops Linus"?

Mindshare and image bloodbath for BitKeeper

[bADlOGIN]

Until Linus started using it for Linux, I'll bet most developers had never heard of BitKeeper. CVS, Perforce, VSS, and ClearCase for the past 4 years mostly seems to be what people would be using. Now all anybody knows is that these idiots dropped Linux support and burned a a great source of publicity (e.g. "our product is so good, one of the biggest OpenSource poster-child projects uses it despite it being closed and commercial!").
Now they'll be known for screwing over Linus Torvalds. I wonder how well they'll fare in future technology evaluations? I can hear the discussions now: "Gee boss, BitKeeper is nice and all but if they screwed over the guy who writes Linux , how do you think they'll treat us after they have our money?"

Re:Mindshare and image bloodbath for BitKeeper

[interiot]

It's possible that BitKeeper wanted to get out of the open-source business for some time, but just waited until OSS folks made a "bad" move, so that at least some potential customers would would see OSS people as the bad guys instead of Bitkeeper.

Re:Mindshare and image bloodbath for BitKeeper

[LittleLebowskiUrbanA]

You do know that BK is tired of people trying to reverse engineer their software, right? That's where this all started.

Re:Mindshare and image bloodbath for BitKeeper

[Anonymous Coward]

"Until Linus started using it for Linux, I'll bet most developers had never heard of BitKeeper"

In the linux world maybe... . In windoze shops it seems quite popular.

" Now all anybody knows is that these idiots dropped Linux support and burned a a great source of publicity (e.g. "our product is so good, one of the biggest OpenSource poster-child projects uses it despite it being closed and commercial!")."

It seems advertised that the product helped doubling the workflow in linux development. And reading linus comments, it did!

Zealotry doesn't count in business. I don't see that stopping releasing free versions of bitkeeper will harm their business. It isn't a technical reason, so the quality of the software doesn't change.

"Now they'll be known for screwing over Linus Torvalds. I wonder how well they'll fare in future technology evaluations?"

Reading linus comments it seems that even after screwing (sic) him, he still have positive remarks about bitkeeper and/or the technology that helped quit a bit in the development of linux.

The message that it sends seems rather negative for linux and the community in whole, then bitmover... .

Re:Mindshare and image bloodbath for BitKeeper

[frank_adrian314159]

BK is tired of people trying to reverse engineer their software...

Uh, last I heard, that's neither illegal, immoral, or unexpected when you release a product to the world. I'm sorry they're unhappy about it, but it seems a bit naive to think that it won't happen and a bit stupid that they'd PO one of their most visible references over it.

Re:Mindshare and image bloodbath for BitKeeper

[Anonymous Coward]

"Uh, last I heard, that's neither illegal, immoral, or unexpected when you release a product to the world."

That only applies to WORKING WITH someone elses product. e.g. client. NOT COMING UP WITH AN EQUIVALENT ALTERNATIVE.

That's why you haven't heard of it, because it doesn't exist.

Re:Mindshare and image bloodbath for BitKeeper

[tftp]

s/tired/delusionally paranoid/

Re:Mindshare and image bloodbath for BitKeeper

[civilizedINTENSITY]

tired of people trying to reverse engineer their software

This is exactly the core issue, and exactly why BitKeeper isn't worth considering.

Re:Mindshare and image bloodbath for BitKeeper

[Anonymous Coward]

Yea-- Why not just rip off all companies that write software and expect you to payment for it. Great "core issue" you moron.

Re:Mindshare and image bloodbath for BitKeeper

[mcc]

I don't care what they're tired of, if a software vendor is very plausible to at some point become suspicious of me (or decide they don't like something that an organization I'm affiliated with is developing) and suddenly start refusing to sell me their products, that isn't a piece of software I would feel comfortable becoming dependent on.

BK doesn't want me using their products in certain ways? Well, I don't want to be paying money for a product from a company who maintains such a massive degree of interest in dictating what I do with it. More to the point I would expect people for whom the day to day functioning of their versioning system is business critical might be less than interested in buying products from a company that may at any time choose to cut off support because they're "tired" of something you're doing.

Re:Mindshare and image bloodbath for BitKeeper

[Anonymous Coward]

That's why you haven't heard of it, because it doesn't exist.

Are you saying whole GNU toolkit doesn't exist? Almost all of GNU fileutils _is_ reverse-engineering of traditional Unix tools---now they are so much better that my organization, which uses Solaris, installs GNU toolkits on the side for use.

Re:Mindshare and image bloodbath for BitKeeper

[abulafia]

Um, "WORKING WITH", like reverse engineering? Products that work with BK don't exist, because the fucking vendor doesn't want to play nice. Now that the gloves are off, I'm sure everything will soon be fine, and you can STOP TYPING like a caffinated lemur. BK probably has a nice life as a proprietary vendor, we'll all learn from their greed, and life goes on.

/me, about to go reverse engineer a proprietary product, yo. Eat my 'l33t Cobol spewing p@ntz. It's totally about right justified fixed width, ya'll. Patent in da house.

Re:Mindshare and image bloodbath for BitKeeper

[Anonymous Coward]

Fair enough--but an OSDL employee was trying to reverse-engineer it, paid by OSDL

According to Bruce Perens (I think), in posting on /., the situation is not so simple.

Larry kept changing the terms of the free licenses

The OSDL employee was doing the reverse engineering in his own time, he was not being paid by OSDL to do it.

How would you like your emplyer to tell you what you can and cannot do in your spare time?

Re:Mindshare and image bloodbath for BitKeeper

[a_random_geek]

"...Now all anybody knows is that these idiots dropped Linux support..."

Says who? They haven't dropped linux support; nor have they dropped support for the other dozen or two other platforms they support. They do, however, claim to be ramping up their windows support some.

Re:Mindshare and image bloodbath for BitKeeper

[cahiha]

You do know that BK is tired of people trying to reverse engineer their software, right? That's where this all started.

Who the hell cares what they are "tired off"? Reverse engineering is legal in many countries and it is an essential part of an efficient free market economy. The only exception to reverse engineering is patent protection, and that is temporary. And if you don't have that, it is completely acceptable if a cheaper copycat product drives you out of business tomorrow.

Re:Mindshare and image bloodbath for BitKeeper

[Anonymous Coward]

I guess this must be a fundamental difference in philosophy between HP (the old HP) and Sun. Yes, Larry left Sun, but comments by their top management seem to support that how IP is managed was not a philosophical difference between Larry and Sun.

HP was very tepid on establishing patent portfolios and "IP" building in general, and did not worry about the inevitable reverse engineering that would occur in the field. Reverse engineering was the way many ingenius ideas got of the ground. The overall feeling was that HP not only had an obligation to produce technology, but also to market it and produce enough real added value that they were clearly better than people that simply tried to imitate them.
This served HP well for many years until that bitch and her cronies came along.

A company uses modus operandi is to try to kill direct imitators has no clothes. Bitmover is openly admitting that they have no value add if an equivalent functional product were to come to the market as free software. Larry is an idiot. If Bitkeeper was felt to be a major opportunity, commercial companies would be buying licenses and reverse engineering it.. which is legal. This is probably happening right now, and Larry is not going to stop. If he (the CEO) is wasting his time trying to, then I don't hold out much of a future for Bitmover. This company apparently feels they have no problems in the product development and tech segment (though I actually think they do), regardless, if they feel that way, they (and Larry 100%) should be focusing on expanding their market. The fact that Larry has the time to gripe on lkml about licensing crap pretty much proves all is not well.
I might be way off, but Larry sounds like a first class asshole. Too much of that and he is going to lose the only thing his company has going for it.. its top notch tech staff.
Larry reminds me of a less intelligent Shockley.

Re:Mindshare and image bloodbath for BitKeeper

[civilizedINTENSITY]

Reverse engineering is doesn't have anything at all to do with "rip off all companies that write software". Reverse engineering is legal, acceptable, and morally OK. There is nothing wrong with reverse engineering. That *is* my point.

Re:Mindshare and image bloodbath for BitKeeper

[Anonymous Coward]

Isn't BitKeeper based on SCCS? How hard is it to reverse engineer something that's been around for 20 years?

Re:Mindshare and image bloodbath for BitKeeper

[Baki]

And you know that Linus reverse engineered UNIX? He recreated a system with the same API. What is wrong with that? There would be no progress at all if we weren't "stealing" each others ideas all the time. Neither in science, nor in art, nor in (software) engineering.

I find the current brainwashing efforts of "intellectual property" proponents to make us believe there is anything wrong with reverse engineering highly immoral, contradictory with human civilization and ignoring its history.

Re:Mindshare and image bloodbath for BitKeeper

[gl4ss]

* You do know that BK is tired of people trying to reverse engineer their software, right? That's where this all started.*

in other words.. bitkeeper is tired of people using their software? if you continue to use bitkeeper as basic building block of your business you take the risk that one day bitkeeper will move to produce a software that competes with your product and just revoke your license on basis that you compete with them! bitkeeper is not really developer friendly.. and well, source management system would benefit from being developer friendly.

me? I had not even heard of bitkeeper until linus began using it, and now in my eyes they have failed because of beign petty about their product - as if there wouldn't be competitors if they just said to people that they're not allowed to create them.

Re:Mindshare and image bloodbath for BitKeeper

[k4rm4_p0l7c3]

I thought Linux implemented Linux from certain specifications. 'Reverse engineered' is not the term for that. The data was documented publicly, why you'd do it the hard way I don't know.

Re:Mindshare and image bloodbath for BitKeeper

[omb]

Now this is way over the top, and what I resent, more and more, on /.

McVoy is ethical, to a fault, and has both (a) a long and distinguished history as a developer and, (b) a temper, and reactivity which I am sure he usually regrets.

There are lessons to be learned, and mistakes not to re-made, ... but OpenSorce learns these lessons and adapts, and inspite of many flames, goes forward.

(1) there is no OS solution, all the SCCS, CVS, svn, arch ... and then PVCS, Clearcase were developed long ago and do not scale well to a large and weakly co-ordinated sets of developers. Try maintaining a Clearcase repository on 5 continents and 7 time zones, (2) Bitkeeper is a good, now closed source, solution to the problem

Larry has ranted, published white papers, huffed and puffed, and I suspect, finally, pissed Linus off --- Linus didn't whinge, he went and built a bridging tool

And, as often the case, only good will come of this.

Re:Mindshare and image bloodbath for BitKeeper

[LittleLebowskiUrbanA]

"tired off?" Dude, you created a typo that I didn't make and criticized me for your poor eyesight..

Re:Mindshare and image bloodbath for BitKeeper

[Threni]

"Mindshare"? I think you'll find that's a metric which was dropped the same time the tech stock it was applied to did a few years back.

Re:Mindshare and image bloodbath for BitKeeper

[Anonymous Coward]

In Linus's younger and more petulant years, he would complain how it is difficult to achieve Posix compliance without paying for the specs, or something to that effect.

Re:Mindshare and image bloodbath for BitKeeper

[Anonymous Coward]

Dude, I didn't criticize you for your eyesight, I criticized you for what you were saying. It doesn't matter what Bitkeeper is "tired of". Reverse engineering is legal. If McVoy doesn't like it, it's a problem with McVoy, not the people doing reverse engineering.

Actually...

[Senjutsu]

The problem is that Linus is not elligible for bitkeeper licensing anymore.

In addition to dropping the free version, Bitmover is refusing to sell even commercial licenses to the OSDL or it's employees, which includes Linus Torvalds and Andrew Morton.

Re:Actually...

[eraserewind]

So why doesn't our Open Source hero just fork the codebase, oh wait...

Re:Actually...

[killjoe]

If this whole absurd scenario is not an object lesson on why not to choose proprietary software nothing is.

Imagine that. A company pulling their license and then refusing to even sell licenses to you or your employees.

Re: Actually...

[po8]

Oooh---there's an attractive offer. Quit your job, but receive a free BitKeeper license. I wonder if Larry would make me the same generous offer?

Re:Actually...

[Anonymous Coward]

Actually, Larry is one of more unpredictable business managers. He has his share of fears (competition), and what you see in this debacle perfectly proves that you should trust no one when you bet your future on someone's promise.

Re:Actually...

[Anonymous Coward]

"Bitmover is refusing to sell even commercial licenses to the OSDL or it's employees"

er, what? You can do that?

Sorry if I've been living in a shoebox, but could I have details or a link please? Thanks.

Re:Actually...

[tftp]

Yes, you can do that. You can do anything you want with your own product. Exceptions are very rare.

Also, you indeed were living in a shoebox :-) because BK was always licensed on a condition that the licensee does not work for a competitor, and does not work on a competing product. The definition of "competing" was at BK owner's discretion, and a lot of Subversion folks were denied the license two years ago. This was a bad deal from the start, and quite a few people said so on LKML. They were right. Search Google or LKML archives, they should have lots of discussion about that.

Re:Actually...

[mrm677]

In addition to dropping the free version, Bitmover is refusing to sell even commercial licenses to the OSDL or it's employees, which includes Linus Torvalds and Andrew Morton.

There is probably some law against this somewhere. Companies buy competing products ALL THE TIME to evaluate and steal ideas. Anybody else have other examples of a company refusing to sell a product to a certain customer?

Re:Actually...

[ergo98]

If this whole absurd scenario is not an object lesson on why not to choose proprietary software nothing is.

Imagine that Linux is writing commercial software, and for competitive purposes decides to keep the source for himself. Suddenly all of the makers of GPL software libraries come running, arms in the air screaming about thawing and re-activating the Stallbot. If that isn't a lesson on why not to choose GPLd software nothing is.

Re:Actually...

[tftp]

Anybody else have other examples of a company refusing to sell a product to a certain customer?

Sure. You can't purchase explosives without a license. You can't buy many drugs without a prescription. You can't buy many weapons in USA, and in other countries you can't buy any weapons period. These are external restrictions imposed on the company, but the company refuses to sell you just the same.

But if you want an example of company's own decision, a most obvious is when the company refuses to sell itself to a potential buyer.

Finally, every store owner reserves the right to refuse the sale to anyone for any reason (read the fine print somewhere in the corner.) Even big stores, like Wal-Mart, have this rule, and some even put it to use. Try to buy something expensive and return it after 29 days; do it again and again and again, and see for how long the store will allow you to take advantage of them.

But I must admit that BK went farther than anyone else. I haven't heard before of a company that refuses to sell the product on basis of other activities of the buyer, or on basis of his employment. Still, the free market, as it is, permits even such unreasonable conditions. But you are also free to not use BK - and that's what I did on behalf of the business I work for.

Re:Actually...

[killjoe]

This is about USING bitkeeper not modifying it.

Nobody will complain about anything if you are USING gpled software.

GPL only kick in when you MODIFY the softare AND redistribute it. Yes people will complain loudly when you attempt to leech off of their labor without agreeing to their terms.

I just don't get you guys. You want to be able to take other peoples work and do what ever you want with it. What makes you think people owe you their time and labor anyway?

Re:Actually...

[cahiha]

The situation is not at all symmetric.

With the GPL, you are never at the mercy of anybody: you can continue to use GPL'ed code no matter what its original author decides to do.

With a license like the Bitkeeper license, you are at the mercy of the copyright holder: when he decides that he has no use for you anymore or that he doesn't like you anymore, you are in trouble. It's his legal right, it's your stupidity to have agreed to such a license in the first place.

Re:Actually...

[Frank T. Lofaro Jr.]

If software companies want to play that way (excluding sales to particular people or entities) and/or any other actions inconsistent with "mass marketing" of software (e.g. Pepsi doesn't know or care exactly who buys their product - just how much money they make overall - and they sure don't blackball potential customers) then they shouldn't be able to have mass-market licenses (click thru EULAs, etc).

Make them get a signature on the license from each individual customer.

That would be consistent.

Re:Actually...

[Frank T. Lofaro Jr.]

A true free market wouldn't have government enforced monopolies, such as copyright.

Re:Actually...

[imroy]

Imagine that Microsoft is writing commercial software, and for competitive purposes decides to keep the source for themselves. Suddenly all of the makers of commercial software libraries come running, arms in the air screaming about thawing and re-activating the Balmerbot. If that isn't a lesson on why not to choose commercial software nothing is.

Re:Actually...

[Tim C]

If a company does that often enough, it will lose all its customers as they migrate away to a less risky company.

I have no idea why the BK people have done this, but it strikes me as being an incredibly foolish thing to do.

This whole absurb scenario is an object lesson to not choose proprietary software produced by idiots.

Re:Actually...

[m50d]

That's not enough though. You never know if/when your good company will be bought out by idiots.

Re:Actually...

[Tim C]

True, but life is full of uncertainties. Maybe the next time I cross the road I'll be killed by an out of control car on the other side that plows across and into me.

All you can do is try to mitigate the risks. You have to weigh up the risk of a supplier of a preferred product going belly up (or just plain bat shit insane) against risks associated with using a less suitable but potentially safer alternative.

Re:Actually...

[Anonymous Coward]

Although it is true that this was what everyone said was going to happen when linus started using BK, I don't think you need to draw any lessons from it. Linus got 3 years of use out of BK. He hasn't lost anything by using it. If he hadn't started using BK, less would have gotten done. Leaving BK only hurts future performance, it doesn't affect past performance.

So, in the end, the lesson still is: use the best tool for the job, regardless of politics.

Re:Actually...

[Matje]

according to the original story, someone who was doing paid work for OSDL (Linus' employer) was reverse engineering the bitkeeper client in his spare time. For bitkeeper, this meant that OSDL was indirectly funding the creation of a free bitkeeper competitor, which goes explicitely against their license terms.

It may not have been the nicest thing to do, but you can't really blame them for protecting their investment can you?

Re:Actually...

[Anonymous Coward]

> according to the original story, someone who was doing paid work for OSDL (Linus' employer) was reverse engineering the bitkeeper client in his spare time. For bitkeeper, this meant that OSDL was indirectly funding the creation of a free bitkeeper competitor, which goes explicitely against their license terms.

Yes. And I heard someone studying at the university of Berkeley also tried to reverse engineer bk. Larry must remove all the free license to student, as it helps university and they train people that want to reverse engineer bk.

And, I think someone had a grant from darpa, and, in its free time, tought about reverse engineering some part of bk. Clearly, larry must stop selling bk to any government entity.

And, beleive me or not, but some not-that-brillant computer user used microsoft windows to surf to slashdot, and thought about reverse engineering bk. Larry must prevent microsoft to get an bk license, as they are indirectly helping to create a competitor.

Bk may be a fine product. Larry used the kernel for marketing advantage. But, by acting like he did, instead of "preventing" competition, he actually send *hundreds* of the most talented developers in the world in SCM systems.

Hacking is about "scratching an itch". Thanks to his paranoia, Larry made himself a big itch to scratch.

Bitmover is fucked. He could have played nice and let people reverse eng bitkeeper. Linus would have still used the "real" one, it would have evolved into a standard, and bitmover would have made a truckload of money, possibly *owning* the scm space.

Re:Actually...

[Felinoid]

If that isn't a lesson on why not to choose GPLd software nothing is.
It's actually a very good example as to why you should chouse GPLed software.

Linus didn't write all the Linux code so if he were to keep the code for himself he should face the wrath of all those people for whom he just ripped off.

I am willing to contribute code to non GPLed projects as long as you pay me for my work becouse I sure as hack don't expect non-GPLed work to remain free for very.

Re:Actually...

[Mant]

So the choice was, go with proprietary software that does what you want or... do what exactly? There wasn't any open source software that filled the need.

If Linus' assement about the impact it has had on his productivity is right, if he had chose open source software that didn't fit his needs, the Linux kernel wouldn't be as far along now as it is.

Even having to abandon it now, that extra productivity isn't lost, and the kernel is more developed that it would have been.

Pragmatically it doesn't seem like a bad choice, plus it may have helped spur on some more development on open source alternatives.

Re:Actually...

[bheading]

It's the double standards about this argument that bother me. If it was not for proprietary software, the GPL movement and the GNU free software directory would not exist. The GNU tools were all initially developed on Solaris and other proprietary operating systems before a kernel even existed.

Even the Linux kernel is far from perfect. On 90%+ of the installations out there it is bootstrapped by a proprietary bootloader/BIOS. I don't hear any of the GPL zealots complaining about that. It supports the loading of non-GPL proprietary modules because - gasp - most users would rather have their graphics hardware supported rather than get religious about the GPL - what a disgrace that people expect to be able to *use* the hardware they paid for with their favourite OS. How shocking it is that people would abandon their free software convictions just so that they could use the wireless access device on their laptop! Isn't that outrageous ?

Well no it isn't actually. Free software has made outstanding achievements and in many ways has revolutionized the way many people and corporations think about their software infrastructure. However there are gaps that must be plugged by proprietary tools in the highly specialized areas that open source - for some reason or another - does not seem to be able to reach.

Re:Actually...

[Anonymous Coward]

This is off topic, but I'll write it anyway. I've always wonder what prevents Nvidia from releasing the specs for the last generation cards. I don't want the source code for their proprietary driver, just the specs so I can write my own, like the DRI code support for ATI cards. NVidia is a fucking joke wrt to open source support, I cannot get DRI 3D support for a TNT card made in 1999 because the specs were never released. Fortunately I'm not much into gaming and my Radeon card works fine with DRI.

Re:Actually...

[bill_mcgonigle]

In addition to dropping the free version, Bitmover is refusing to sell even commercial licenses to the OSDL or it's employees, which includes Linus Torvalds and Andrew Morton.

That's so bizarre. The only reason I've ever heard of BitKeeper is because it's used for the kernel.

Next up, Gatorade will refuse to sell to the NFL.

Re:Actually...

[bheading]

I suspect the reason why they won't release their driver code is because it may reveal too much about the specialized workings of their hardware. I suspect they have hundreds of man-years of work tied up there.

Re:Actually...

[Discoflamingo13]

Finally, every store owner reserves the right to refuse the sale to anyone for any reason (read the fine print somewhere in the corner.)

In the US, that is not true - you may refuse service to anybody provided that your decision is not based on the customer's age, race, creed, national origin, sex, marital status, or physical disability. Some states have included sexual orientation in this list as well.

Re:Actually...

[ergo98]

GPL only kick in when you MODIFY the softare AND redistribute it.

And in this case the commercial license came into play because, as is classic in the innovative world of OSS, someone was working their hardest to steal the design of Bitkeeper. It was hardly an arbitrary decision.

Yes people will complain loudly when you attempt to leech off of their labor without agreeing to their terms.

Leech...I love that term. It's stated as if everyone of the people out there using Linux are busy contributing to the kernel. The reality, of course, is that overwhelmingly, probably 100,000:1, it's "free software", and the fact that it is GPLd is no different from it being freeware- it's free as in beer.

However I, and many others, will negatively mention the GPL because it is a selfish and self-serving license (yes, JUST LIKE COMMERCIAL LICENSES), yet GPL fanatics dream that it is the second coming. Thanks, but no thanks. My software stack is commercial and BSD.

Re:Actually...

[killjoe]

"And in this case the commercial license came into play because, as is classic in the innovative world of OSS, someone was working their hardest to steal the design of Bitkeeper. It was hardly an arbitrary decision."

Again this points out the one of the biggest problems with using proprietary software. If a company feel that you are trying to compete with them or trying to co-operate with people who compete with them they can pull your license and prevent you from using their software.

In this case some individual was developing software in his own time that might one day compete with bitkeeper. When OSDL refused to fire the guy bitkeeper prevented linus and thousands of other linux developers from using bitkeeper.

Whenever a commercial company wants to they can prevent you from using their software. Even if you paid for it.

"However I, and many others, will negatively mention the GPL because it is a selfish and self-serving license (yes, JUST LIKE COMMERCIAL LICENSES), yet GPL fanatics dream that it is the second coming."

It's not the second coming, it just prevents leeches like you from getting rich off their software. Lucky for you there is lots of BSD code to steal from.

GPL is great because it prevents welfare programs like BSD. BSD is just a giant welfare program for commercial software writers. It's just people volunteering their services so that companies can make money off them.

That's great for some people but not for me and not for tens of thousands of other people. We prefer that our code remain our code. It's a self healing commons, one that can never be taken away.

Re:Actually...

[cheesybagel]

RMS was right and Linus was wrong this time. Next time someone calls RMS a paranoid or that Linus says the only thing that matters is if it works, not the license, you know what to do.

PS: I have also called RMS a paranoid before about licensing issues, but later I learned that he is usually right. The annoying git. :-)

Re:Actually...

[theLOUDroom]

It's the double standards about this argument that bother me. If it was not for proprietary software, the GPL movement and the GNU free software directory would not exist.

You really should read some history of the whole open source movement. All the software Stallman was originally using WAS at least somewhat free (at least implicitly, if not explicitly). The FSF was created as a result of Stallman's later encounters with proprietary software, not because the mood just suddenly struck him after having been previously relying on proprietary software.

Your claim here fails for any number of reasons, the most obvious being that it doesn't even make logical sense. If there was no proprietary software, then by your own definition of the situation, everything else would HAVE to either free software, or we just wouldn't have any software at all.

Re:Actually...

[ergo98]

GPL is great because it prevents welfare programs like BSD. BSD is just a giant welfare program for commercial software writers. It's just people volunteering their services so that companies can make money off them.

Oh god....gut busting. This is too fucking rich. Tell me something - are you really this incredibly stupid? (I suspect so) Are you so blinded by your idiotic zeal that you think this is rational and well-thought out?

Perhaps you missed the prior point (of course you did), but the overwhelming majority of GPL software users are not developers, and they will never contribute anything back. You understand that? Many of them are using this software to replace commercial software. Understand? Many of them are running businesses, and GPL nutbars like yourself have them laughing to the bank, allowing a critical part of their infrastructure to be free.

Yet all your raving nuts can see is some developer who might link a piece of GPLd software into a commercial product. Hilarious. It's like a bunch of dumb, insect-like lobsters all crawling around at the bottom of the tank, desperately trying to pull other lobsters down, all while the cooks (users and business) are getting ready for their lobster meat feast.

Re:Actually...

[Anonymous Coward]

IMO, RMS is generally correct except on one thing: he needs to get off the whole idea of calling it Linux distros "GNU/Linux"... There is way too much non-GNU software in the mix that it doesn't make sense.

Re:Actually...

[killjoe]

"Oh god....gut busting. This is too fucking rich. Tell me something - are you really this incredibly stupid? (I suspect so) Are you so blinded by your idiotic zeal that you think this is rational and well-thought out?"

Everything I said is 100% factually true. When you code using the BSD license you are volunteering as a developer for apple, MS or whever uses that code make proprietary products with. Not that there is nothing wrong with that, some people don't mind volunteering to large corporations. Other people do and that's why we choose GPL.

'Perhaps you missed the prior point (of course you did), but the overwhelming majority of GPL software users are not developers, and they will never contribute anything back. "

Sure, OK. But hey guess what, even users can (and do) help by submitting bug reports, deisgning logos, helping out on newsgroups, irc, mailing lists etc.

"Many of them are running businesses, and GPL nutbars like yourself have them laughing to the bank, allowing a critical part of their infrastructure to be free."

More power to them. As long as they don't steal the code they can do whatever they want. GPL ensures fair competition. We like that.

Re:Actually...

[boots@work]

someone was working their hardest to steal the design of Bitkeeper

They were working their hardest to get their own data back out of a proprietary data jail. Is that so wrong? What started this was an export tool, *not* a bk clone.

The reason Larry got pissed off is that he sees the history of the kernel (and other projects in bk) as his proprietary information, not the property of the original authors. If he was worried that in the course of getting their data out people would discover his secrets then he should have provided a decent export tool in the first place. (Practically every other version control system does, including Clearcase and p4 IIRC.)

This is like Microsoft cancelling your whole company's Office licences because somebody tried to decode a .DOC file -- leaving you with all your documents unusable, and many people saying "I told you so."

What about Bazaar-NG?

[Anonymous Coward]

How about Bazaar-NG? It looks like it's a design based on GNU Arch that can do either centralized, or decentralized stuff.

Re:What about Bazaar-NG?

[Anonymous Coward]

How about Bazaar-NG? It looks like it's a design based on GNU Arch that can do either centralized, or decentralized stuff.

Bazaar-NG is indeed the ideal solution, but it isn't ready for prime-time yet. Try plain Bazaar, Monotone, or svk.

Here's a link to LWN where he talks about it

[SDrag0n]

He talks about some other products he's tried, why he wrote his own, and a little about how it works.

http://lwn.net/Articles/131313/

Check the "made the first version available" link towards the top

Re:Here's a link to LWN where he talks about it

[abulafia]

"Because it sucks?"

Best response ever. And I say this as a serious DB advocate.

not surprising though

[roman_mir]

I remember reading this article and leaving comments on it. RMS was quite clear:

JA: What about the programmers...

Richard Stallman: What about them? The programmers writing non-free software? They are doing something antisocial. They should get some other job.

My answer to that was not quite polite.

BTW. don't be surprised to get -2Troll in a short while here.

Re:not surprising though

[HalfFlat]

So which do you take issue with?

1. Writing non-free software is antisocial

or

2. If a job is based upon antisocial behaviour, one should get some other job

Personally I can't see how you could argue against 2, unless you are in fact antisocial and just don't care. RMS has been making a case for point 1 for years now, and the recent news with regards to BK is more supporting evidence for his argument.

Reading your comment to the interview, I can't understand your vehemence. It's irrational. If you have a choice of doing something bad or doing something good, why insist that it's okay to do the something bad? I don't want to put words into RMS' mouth, but I get the impression that when he states something like: "It is better not to program at all than to program non-free software", it's not a statement about the utility of free versus non-free software -- as you noted in your comment, a lot of non-free software is very useful -- but instead it is a statement concerning how programming as a practice in today's society should be carried out. Programming non-free software is not bad because it makes useful software -- that would be silly. It's bad because it supports a system in which non-free software proliferates and causes huge amounts of waste. It is essentially the same point as the one you quote.

The horrible irony of it all is that most non-free software is non-free by default, not because the owners of that software are profiting significantly by having it be non-free. Most software is written for internal use for internal projects. Embedded software is usually specialised to its hardware and of limited use to competing manufacturers. In as much as it is not limited, a culture of free software allows the programmers to write original software, not re-write what others have done hundreds of times before.

If hypothetically free software were to become the only legal form of software as of tomorrow, the vast majority of programmers would still be in a job. And those jobs would be better.

Re:not surprising though

[tomhudson]

Writing non-free software is antisocial

Not necessarily. This is something that sounds good in a sound-bite, or a slogan, but I can't see this being true in all cases.

A lot of software wouldn't get developed if someone weren't paid to do it - after all, most people, given a choice, would rather "scratch an atch" - do what they enjoy working on - instead of working on stuff they find boring. So the boring stuff ends up having to be either paid "work for hire", or "contracted out", or "proprietary".

Unless you're suddenly going to rewire a LOT of people's brains so that they suddenly find boring projects interesting, there will ALWAYS be room for closed-source/proprietary software.

It's not anti-social. It's not even asocial. It's just the way things are. To castigate others for exercising their freedom of choice by using a mix of open- and closed-source programs as being antisocial is itself antisocial behaviour.

Heck, I use open-source for almost everything, but there is ONE program from MicroBitch that I actually find useful (Image Composer - I know, it's lame, so what ...) for the quick-and-dirty task of making logos, buttons, etc. I *could* use other programs ... that I don't is my choice. It doesn't make me antisocial, any more than using open-source makes me a smelly hippy.

Re:not surprising though

[Anonymous Coward]

troll? troll? wtf?

Re:not surprising though

[HalfFlat]

A lot of software wouldn't get developed if someone weren't paid to do it - after all, most people, given a choice, would rather "scratch an atch" - do what they enjoy working on - instead of working on stuff they find boring. So the boring stuff ends up having to be either paid "work for hire", or "contracted out", or "proprietary".

You are casually conflating "work for hire" and "non-free"!

Almost all the programming work I do is for a private company. It is work for hire in that sense. Some of it incorporates GPL software and thus is by necessity GPL licensed itself. The other software could just as well have been licensed as free software as well. It doesn't matter. Being free or non-free is orthogonal to being payed or not for programming.

It is only an issue when it comes to selling software itself, unenhanced by other software, as a product. Because it is what makes the Microsofts of the world, and because it is very visible, it feels like this is what programming is all about. But it really carries a disproportionate impact.

On your other point: we all have a choice, it is true. But you could say it is a choice to be antisocial :) Just like other behaviour which supports institutions which are bad in the long-term, such as wasteful use of energy, water, etc., there is some degree of culpability. As to how antisocial it is to support non-free software, or for that matter, to drive SUVs in an urban environment, is a separate but important question. Personally, I am no environmental saint, but I am aware that I am choosing not to be one. This is a form of antisocial behaviour, just a very pallid and common one.

Re:not surprising though

[tomhudson]

You are casually conflating "work for hire" and "non-free"!

Not really - I used the term as one of a series, when I said "So the boring stuff ends up having to be either paid "work for hire", or "contracted out", or "proprietary". Only the first case equates to your scenario. In the other 2 cases, they do not have ownership of the code in question, so at least 2/3 of the cases I cite are "non-free" from the start.

And, of course, they may also decide to keep the fruits of the "work for hire" non-free.

Now on to the "real world" ...

The definition of the term antisocial

1. Shunning the society of others; not sociable.
2. Hostile to or disruptive of the established social order; marked by or engaging in behavior that violates accepted mores: gangs engaging in vandalism and other antisocial behavior.
3. Antagonistic toward or disrespectful of others; rude.

Certainly, using closed-source software such as bitkeeper doesn't meet the definition of antisocial behaviour. RMS might be brilliant in some things, but his statement that closed-source software is antisocial is itself antisocial as per definition (3) above. Maybe he needs to look up "irony".

It's talk like that that keeps perpetuating the stereotype of open source being the domain of a bunch of antisocial long-haired fat smelly moms-basement-dwelling communist crunchy-granola-munching unrealistic anti-business obsessive-compulsive hippy creeps with borderline personality disorders, no sense of humour, and totally lacking in both perspective and basic social skills.

As long as we keep it up, we're just giving more ammo for the Bitch from Redmond to slap us back with.

Re:not surprising though

[Anonymous Coward]

You sound very rude and antagonistic.

Re:not surprising though

[cahiha]

Why don't you quote your actual response?

I have two words for RMS in this case: Fuck You, RMS. I will write all the non-free software I wish and no-one is going to stop me.

I myself don't actually fully agree with RMS's statement: I think whether writing non-free software is anti-social depends on the circumstances. And not all anti-social activities are necessarily to be avoided (they may be the lesser of two evils).

But "fuck you, I can get away with doing whatever I want" is not an appropriate response, in particular to someone like RMS, who has consistently stood for his views and made a lot more personal sacrifices than you are likely ever to make for any cause in your life.

There are logical responses one can make to RMS, and one can disagree with him. But you have merely proven yourself to be quarrelsome and insulting--bad qualities in a commercial programmer. I hope your future commercial employers will take notice.

Re:not surprising though

[Sigl]

RMS: Writing non-free software is antisocial

tomhudson: [deciding to use Image Composer (ie non-free software)] doesn't make me antisocial, any more than using open-source makes me a smelly hippy.

Did you mean to compare USING software to WRITING software? You're right about choosing to USE non-Free software doesn't make you antisocial. I agree.

Try limiting your thoughts to writing software. If you think of the code written as an expression of the programmer, I can interact with Free software to a degree that can't be matched with non-Free software. In this case antisocial would mean the actions of one programmer are prevented from interacting with another programmer.

Re:not surprising though

[Anonymous Coward]

It wasn't polite and you didn't really defend your position with anything more than, I'm going to do what I want, and you can't stop me. One could call that a rather antisocial philosophy, which would support RMS's point.

Re:not surprising though

[roman_mir]

Well, precisely, who the hell is RMS to tell people who chose to do whatever they do - write nonfree software for exampe - that they must not do it and should switch to another occupation or a hobby?

He is a guy with his believes, I am a guy with my believes. I believe RMS is pushing his believes against mine and the only response to that is tell im to sod off.

Re:not surprising though

[roman_mir]

1. Writing non-free software is antisocial

or

2. If a job is based upon antisocial behaviour, one should get some other job - I disagree with both statements.

If my writing non-free software helps some user to do whatever it is he/she wants to do I don't see how that is antisocial.

Secondly even if someone is working and in process he/she excercises antisocial behaviour (and I disagree that programming nonfree software means that,) I don't see how it is of anyone's business except for that person. I chose to be antisocial in many ways but writing closed source software is not one of these ways. When I do chose to be antisocial it is my choice and I will take offence at anyone pointing out that I should change whatever 'antisocial' thing that I do. While some antisocial behaviour is illegal - like I suppose killing people, some is not - like not going to the parties and not caring to talk to people and whatever antisocial means.

Many firms write closed source software for their own use and many private people do that too. Is it antisocial to create something and not let anyone in the world use it? Maybe, I don't know. But writing software and letting people use it even if for money is not what I call antisocial.

Re:not surprising though

[roman_mir]

But "fuck you, I can get away with doing whatever I want" is not an appropriate response, in particular to someone like RMS, who has consistently stood for his views and made a lot more personal sacrifices than you are likely ever to make for any cause in your life. - oh really, don't talk about personal sacrifice to me. Who the hell are you to judge what personal sacrifices for what causes I have made?

My future commercial employers? I am a contractor and I always get things done and on time. That's what commercial employers care about above all.

I don't need to reply logically to a statement that is based on a system of believes and feelings of moral superiority rather than on logic.

Re:not surprising though

[aCapitalist]

Here's another one where Stallman compares proprietary developers to perjuring cops and murderers.

http://www.gnu.org/philosophy/my_doom.html

Re:not surprising though

[hazah]

A lot of software wouldn't get developed if someone weren't paid to do it

Are you SURE you're interperting the word FREE correctly? You can develop free software in a veriety of situations. First, in your basement, for nothing. Second, in your basement for someone for nothing. Third, in your basement for someone for some $$. You are *free* to do it.

The only thing that changes between free/non-free software is that at the point where the software is done, if it is free, who ever paid you to develop it, is as *free* to distribute it further as you are. Even with modifications.

Re:not surprising though

[Tim C]

If hypothetically free software were to become the only legal form of software as of tomorrow, the vast majority of programmers would still be in a job. And those jobs would be better.

I'm a programmer; how would my job be better if free software became the only legal form of software?

Re:not surprising though

[Anonymous Coward]

The problem is your basic assumption that it is possible to replace all non-free software with free software. Clearly businesses who use software benefit from free software over non-free software if it does the job equally well. That so much of software out there is still non-free is a testament to the inability of the free software community to deliver products that are useful in the real world. GNU/Linux is nice, but it's only a unix clone, and not a very good one, and when you get to specific functionality, like business logic software, there is simply nothing free out there than can do the job.

Now, you could argue that if the system was redesigned to ensure all software was free, those business logic projects would become free software. That may be true, but it's up to you to demonstrate a method that that can happen, and so far I haven't heard anything from the Free software community that makes me think they have a credible way of replacing proprietary development wholesale in a way that still puts bread on the table of the people who make the software. The current Free software development still rests on a foundation of proprietary software. Almost all businesses that use/develop free software use it in support of proprietary software where the real money is made that funds the free software development. There are few examples of projects I can think of that only release under a Free license, that actually pay their developers. And those projects that do generally use donation mechanisms that won't scale to industry-size levels (for the same reason communism failed: people are genetically engineered to be selfish and greedy). If all software becomes free, all software developers must be paid directly for their work on the free software. I see no way of doing that. Do you?

It's like people who complain about hunger and poverty. Nobody disagrees we should eradicate hunger and poverty. It's just that there is no credible method to do it. The more resources we throw at the problem, the bigger it becomes, due to the basic way society is engineered. Until the protesters outside of economic conferences demonstrate a credible way to re-engineer society to eradicate hunger and poverty, they say little of use. Same with the Free software crowd harping on about the evils of proprietary software. They're right in their criticism, but like "show me the code", to them I say "show me the process".

Re:not surprising though

[tomhudson]

The topic of this thread was USING non-free software - in this case Bitkeeper.

Try limiting your thoughts to writing software. If you think of the code written as an expression of the programmer, I can interact with Free software to a degree that can't be matched with non-Free software. In this case antisocial would mean the actions of one programmer are prevented from interacting with another programmer.

It's stil not antisocial. For example, I can choose not to interact with people who don't pick up their dog crap (a pet peeve of mine). This doesn't make me antisocial. It means that I have a choice, and I'm exercising it.

Same with code. People have a right to keep their code private, for whatever reason (for example, would you REALLY like anyone to look at your first coding attempts? I wouldn't. Some was god, some was embarrassing.)

Another reason: maybe I don't want just anyone fooling around with the code because I don't want them adding buggy stuff to it, redistributing it, and tainting my reputation. Again, not antisocial. Just cautious.

Everyone has the right to choose under what regime they will release their code, or not to release it at all. To label someone as antisocial when they're not a monopolist is n itself antisocial behaviour, because it disregards their rights.

People should release their code because they want to, not because there's any pressure to conform to a particular social model.

Re:not surprising though

[roman_mir]

Writing closed source software is not antisocial, lookup the meaning of the word. Writing software that can be sold and used by whoever needs the software is definitely not antisocial. Even writing software for yourself that you will never release to anyone else ever is not antisocial.

Re:not surprising though

[Sigl]

..., lookup the meaning of the word.

At Dictionary.com antisocial: 1. Shunning the society of others; not sociable.

Also, Shunning: To avoid deliberately; keep away from.

Now if noone is interested in seeing the code then I wouldn't label it deliberately avoiding. However, releasing in non-Free manner would preclude the interest so might still be considered deliberate avoidance. In a society that that uses code to express some ideas to the point the code is considered Free Speech, I absolutely think this definition fits with releasing a program but keeping the code to yourself. I'm not saying it's destructive behavior but it does fit the definition.

Note: For this definition you don't have to be bucking a social norm to be antisocial.

Even writing software for yourself that you will never release to anyone else ever is not antisocial.

Certainly, just keeping something to yourself is not antisocial. But if all you did is code and kept it all to yourself would that be antisocial? That might get you "marked for behavior deviating sharply from the social norms" (from webster.com). That might be a different definition of antisocial but the word still applies. I don't care about this definition because it will just devolve into a debate about what the "social norm" is. For the definition I refer to it's not required.

Re:not surprising though

[Sigl]

The topic of this thread was USING non-free software - in this case Bitkeeper.

roman_mir started this thread with Stallman talking about people writing software. I can't find anywhere in the thread where someone said using software was antisocial until you brought it up.

Re:not surprising though

[Anonymous Coward]

oh really, don't talk about personal sacrifice to me. Who the hell are you to judge what personal sacrifices for what causes I have made?

You seem to have trouble with distinguishing a suggestion and supposition from "judging".

I don't need to reply logically to a statement that is based on a system of believes and feelings of moral superiority rather than on logic.

You can reply in whatever way you like, and people can think of you whatever they like in return. Being illogical, insulting, and using vulgarity in response to a statement by someone who has been principled, consistent, and altruistic for all his life tells us a lot about yourself.

I am a contractor and I always get things done and on time. That's what commercial employers care about above all.

You just go on believing that.

Re:not surprising though

[tomhudson]

Guess you didn't bother following the linky-linky that roman_mir linked to, (or you missed this):

JA: What if your job requires you to use non-free software?

Richard Stallman: I would quit that job. Would you participate in something anti-social just because somebody pays you to?

Here Stallman is stating that he would consider it antisocial behaviour to use non-free software.

Re:not surprising though

[roman_mir]

Shunning the society of others; not sociable. - only this is not a sentence about software but of general human behaviour.

Now if noone is interested in seeing the code then I wouldn't label it deliberately avoiding. However, releasing in non-Free manner would preclude the interest so might still be considered deliberate avoidance. - If someone pays for the software and plays by the license it is entirely possible to release the source code to them for viewing. This does not mean that the source code has to be GPLed, but only released for a review.

In a society that that uses code to express some ideas to the point the code is considered Free Speech, I absolutely think this definition fits with releasing a program but keeping the code to yourself. - what does free speech have to do with this? I am free not to release the source or to release it. I imagine that this freedome is more important even than the freedome of speech. You are just confused, there is no logic tying together being antisocial and not releasing source under a Free license.

Certainly, just keeping something to yourself is not antisocial. But if all you did is code and kept it all to yourself would that be antisocial? That might get you "marked for behavior deviating sharply from the social norms" (from webster.com) - I doubt that people care to mark someone as antisocial because this person never released his software into the wild. I wrote plenty of software for my own use - various tools, games even that I never gave to anyone. It's a learning process for oneself, it has nothing to do with being antisocial.

confusion all around here

Re:not surprising though

[Sigl]

Guess you didn't bother following the linky-linky that roman_mir linked to, (or you missed this):

Why would I refer to the section about using software when roman_mir was specifically upset about the part about writing software?

Richard Stallman: I would quit that job. Would you participate in something anti-social just because somebody pays you to?

Here Stallman is stating that he would consider it antisocial behaviour to use non-free software.

I would say it is not antisocial but would enable it. You are right about his statement. I do infer that using is actually participating in something antisocial. I'm not sure I see how because, his participation may enable it, but, it doesn't become antisocial by association.

Re:not surprising though

[roman_mir]

So, now tell me if I wasn't prophetic. My original post is already Troll, it just has to lose a few more points to get to -2.

You on the other hand did not give any comments to my answer to you. Think I am trolling? No, I am genuinely disgusted with RMSs position on the issue. But here you can't be genuinely disgusted with that position and not affect the psychosomatic problems of many of /. readers. I think they start having diarrhea when someone is genuinely against something that they accept as an ideology. Because to them it is not an idea, it is an ideology. RMS has an idea - a all consuming paranoia really. But to him it is a genuine idea as well. To the rest of the most this is too complex and/or too crazy to fully diagest the idea, so many just accept it as a higher 'truth'. Sort of like a religion, and RMS is the priest. And anything that attacks this 'saint figure' must be evil and cannot coexist with the neat little ideology hammered in their little heads.

Think I care about the score? I've been here for years, 2000 posts, the karma is so high, I get Mod points at least 1-2 a month. No, it is not the score that concerns me. It is the fact that there is a new religion in town and it is based upon something I came to love over the past 12 years - programming computers.

How about... Arch or Monotone

[ikewillis]

Arch and Monotone are both GPL-licensed distributed development tools, and retain a BitKeeper-like distributed development model which Linux prefers.

Somehow Arch was immediately mentioned on the original thread about Linus's intent to switch away from BitKeeper, but somehow only Subversion has been mentioned on this one. Arch was created specifically with the goal of replacing BitKeeper as the SCM for the Linux kernel source, as it says on their web page:

It is somewhat well known, these days, that some of the core developers of the Linux kernel are using a revision control system which is not free software. There is a need to create a free software alternative to that system and to do so is one of the goals of the arch project

Re:How about... Arch or Monotone

[Anonymous Coward]

What's needed is for the people who work on the kernel and some of the other major OSS projects (glibc, gcc to name a couple) to take a few days and talk about what they need in terms of source control. The fact that there is rarely any requirements/specifications documented is probably the number one cause of opensource projects dying off. Nobody remembers what it was supposed to be doing after a few years, and by then its an emacs clone, but with fewer features and more suckage.

Re:How about... Arch or Monotone

[the eric conspiracy]

Monotone has been discussed and appears to be too slow. SVK, Darcs, and Arch are still in the running.

Re:How about... Arch or Monotone

[Anonymous Coward]

Too slow? Where is this discussion.

I ask because I'm considering switching away from the CVS/SVN model myself (I currently use a mixture of CVS and SVN).

Monotone or Arch are the only decent and cheap systems I found. Monotone looked "neato" but I have no idea if it's any good. If it has performance problems with something as small as the Linux kernel then I'll need to look elsewhere.

Arch looked interesting but it's a GNU project and for some reason non-trival GNU projects almost always suck.

Re:How about... Arch or Monotone

[Anonymous Coward]

Monotone is too slow and SVK isn't??

SVK is layers upon layers of bloat. A gazillion Perl modules + Subversion. All in a big 12-layer cake. I.e. it uses Subversion as its storage engine (and Subversion ain't exactly "slim and trim").

I installed it on my Gentoo development server (currently armed with CVS, SVN, and Darcs, and tailor.py to mirror between them all). It was WAYYYYYYYYY too slow, even for a project with one file in it. And it got mighty confused about connecting to non-SVN repositories like CVS. And apparently none of these guys have ever heard of.. BRANCHING?? It botched all my CVS branches (missing files for instance).

I'm keeping my eye on Darcs for the time being. 1) extremely easy to use and 2) at least an attempt at a theoretical foundation, rather than ad-hoc "command-oriented" design like some of the competition. Also the author is friendly and intelligent. Tom Lord (arch) is merely intelligent. :-)

Re:How about... Arch or Monotone

[Anonymous Coward]

darcs' dependence on haskell pretty much kills it for most people. cross platform haskell is still really squirrely (the win32 cygwin port is highly unstable for example).

the fact you're still using cvs speaks volumes though.

Re:How about... Arch or Monotone

[MassacrE]

Actually, slowness in subversion is usually easily shown to be due to the stateless point at which it starts operations. It has to scan all the directories (and some # of files) to decide what the local filesystem state is before it can begin the network operation. In most cases the actual network operation is as fast as it can be, it is the local working copy operation which hurts you. There are certainly things at the filesystem level which could be done to make all operations (on all fileystems, across all source control systems) faster.

Re:How about... Arch or Monotone

[Anonymous Coward]

arch functionality is good but the enforced archive structure blows.

it also suffers from poor cross platform support (the win32 port is pretty broken).

svn is far more flexible, and much better supported.

Re:How about... Arch or Monotone

[Anonymous Coward]

This was pointed out in kde-devel mailing list,
they had such problems when trying to import
KDE CVS into subversion...

The transition to SVN is still "in progress"...

It was supposed to be done by April 1st,
then April 4th, then...

Re:How about... Arch or Monotone

[the eric conspiracy]

Monotone is too slow and SVK isn't??

Here is a quote from Linus:

"I'm playing with monotone right now. Superficially it looks like it has tons of gee-whiz neato stuff... however, it's *agonizingly* slow. I mean glacial. A heavily sedated sloth with no legs is probably
faster."

Darcs is interesting, but it already has a bad rep in regards to speed and scalability because of the Haskell infrastructure. I wouldn't put the kernel on it. There are also concerns about getting otehr developers to support it because of the Haskell code base. Right now it's a one man show.

Subversion already is well proven on large projects - the question is whether SVK is mature enough.

My guess is what is going to happen is that the kernel guys will do something custom, perhaps on top of or in conjection with SVK and SVN. Monotone, Darcs, etc aren't very likely, and Subversion by itself doesn't fit the distributed work flow of kernel development.

Re:How about... Arch or Monotone

[stor]

the fact you're still using cvs speaks volumes though.

What do you mean by that? That the parent is ignorant for still using CVS or the fact that there's no reasonable alternative?

We're still using CVS at work. CVS isn't great but it's problems are well-known here. For most projects CVS does OK and when it doesn't I go in and manually fix stuff with confidence.

I haven't found any free replacement for CVS that offers compelling (read: no brainer) reasons to switch. Subversion seems to be the closest thing but going for that will require a painful transition that requires justification.

As an admin rather than a developer I'd be futzing with the developer's data so I better have a damn good reason. The developers would have to be absolutely convinced that this new and shiny revision control system is the way to go. It would have to be really easy for them to use. It would have to be put in a testing environment first, with the developers helping me test it thoroughly.

Give me a compelling reason for the hard work and for the abuse I'll receive if one tiny thing goes wrong. "CVS sucks" doesn't cut it: I'm well aware of how much it sucks and I'm OK with it for the time being, as are the developers that rely on me.

Cheers
Stor

Re:How about... Arch or Monotone

[paskie]

You are incorrect w.r.t. Linus quote, that's what Chris Wedgwood said. Linus is still considering Monotone, believing that the speed issues can be fixed.

Judging from the level of enthusiasm, the second top contender is probably Bazaar-NG, if it really delivers in time. SVN and derivatives aren't really so popular amongst kernel developers, for various reasons, so I wouldn't say it is that much likely that they/we will use that.

Note that we are now working on Git (you might like my branch now, which provides some humanly usable user interface ;-) ), Linus' interim storage system suitable for tracking of trees history. It is not a full-blown versioning control system, but it is lightning fast and probably can do enough stuff to be usable for the development at least until some full-blown VCS matures.

Re:How about... Arch or Monotone

[evvk]

The important speed and scalability issues have nothing to do with Haskell, but the use of an advanced O(n^2) merging algorithm for everything.

Monotone does not provide anything close to such a thing, and thus no cherry picking. Arch provide both stupid and star merge, which again makes it more complex to use.

Re:How about... Arch or Monotone

[the eric conspiracy]

You are incorrect w.r.t. Linus quote, that's what Chris Wedgwood said.

Yes, I stand corrected - Chris was quoting Linus which confused me as to the author. Nonetheless Linus did agree with the assesment as to the current speed. Of course we don't know if it can be fixed - but until it is it is out of contention.

Bazaar-NG is an interesting project - but as you noted isn't ready now and I am sure you realize it can't really be evaluated until it is. And SCMs really need to age for a while to get the bugs out.

As far as svn, everybody understands that per se it doesn't fit the development patterns used by the kernel team. However I think you guys may be missing out on the boat as far as SVK goes - it offers a lot of what you find missing in svn and probably is closer to being a workable system than any of the other alternatives.

Git is a kind of interesting study, but as is noted is a tracking tool for interm use, not a real SCM.

Re:How about... Arch or Monotone

[the eric conspiracy]

We're still using CVS at work.

It is surprising how many small shops don't use version control at all. For them the move to CVS is a giant step forward, much bigger than the move to something better than CVS.

Ultimately though most commercial teams move out of CVS - it just has too many issues. The one that got us was that it has no defense, diagnostics or warnings against repository corruption. We found that a system like that as storage for the crown jewels of our company and the source of our livelihood to be unacceptable.

Re:How about... Arch or Monotone

[IamTheRealMike]

Importing CVS branches is really hard because CVS itself is just a huge hack. I've been using SVK lately for local Wine development and it's worked pretty well. The startup time is a pain in the ass, but there are ways around it and I think that's been much improved in the recent versions. I'd definitely say it's mature.

Best of all, the SVK maintainers are incredible. If I find a problem, I literally report it to them on IRC and a few minutes later it's fixed. I've yet to encounter a bug in SVK that required more than a 20 line patch from them to fix. It also has test coverage of over 90% - unbelievable.

Basically, I don't see why Linus is not giving SVK more attention. It's pretty fast, already stores the kernel tree (this is one of its main test cases), has the same sort of features as Arch or BitKeeper except with a friendly interface, and the code is clean and robust. The only downside is the tricky install - if CPAN bails out, things get tough.

Re:How about... Arch or Monotone

[MassacrE]

I believe KDE and GCC are now both waiting on a 1.2 release, that fixes many of the 'slowness' issues these projects have hit.

CVS import is unfortunately underoptimized, because usually people only do it once.

It is also very difficult to do, because you have to infer correlation between files in order to create a proper revision history - file-level commits means you need to do a substantial amount of scanning in order to figure out which files were part of a commit, and non-atomic commit means that the files may actually have different timestamps.

Distributed operation?

[tepples]

Is Subversion decentralized? Even the Subversion maintainers admit that the Linux kernel development atmosphere doesn't mesh well with the centralized CVS/SVN model. There exists SVK, a decentralized hack on Subversion, but is it mature?

Re:Distributed operation?

[Anonymous Coward]

considering svk is in wide use and has good cross platform support now, i would say yes it most definitely is mature.

Re:How about...

[Anonymous Coward]

Haha, good troll.

Re:How about...

[Spunk]

Wow. Amazing how Open/Free software can work together to the point of one project recommending that we should use their competitors' project in some instances. :)

Can you imagine a commercial software vendor releasing something like that?

OpenCVS.org Anyone?

[Anonymous Coward]

Cripes. For a freedom advocate, Linus is not very good at it...

not surprising though-My way, or the highway.

[Anonymous Coward]

It'll eventually bite RMS in the ass. Just watch.

Anyway I'll not be using (L)GPL and recommending that others not do so as well. Good code, and good things aren't the exclusive province of GPL'ers.

bitkeepers website

[burntash]

maybe bitkeeper shouldn't be including Linus' praise about their software, since now he has a new opinion

Re:bitkeepers website

[a_random_geek]

..but he doesn't have a new opinion. He's dropping bitkeeper for licensing concerns, not because it is an inferior solution.

Pay attention.

Re:bitkeepers website

[Anonymous Coward]

Hell, if you cannot get a licence anymore it's not even a solution, let alone an inferior one. And paying attention is indeed a good idea.

Re:bitkeepers website

[bheading]

That's the thing, he doesn't have a new opinion.

Isn't that what Subversion does (really well)?

[toby]

What it does is store versions of trees of files and allow them to be annotated

Re:How about...

[brett_sinclair]

svk may be a reasonable compromise. It is a distributed scm, but built on top of the very stable Subversion. It is very fast, even for large trees, despite being written in perl.

It is also quite easy to use svk together with a central Subversion repository, to get the best of both worlds.

Re:dupe-comments

[tomhudson]

Your complaints about people complaining about dupes is a dupe.

You must be new here ...

opencm anyone

[Solilok]

I am considering various cm's for a project.

Arch and Monotone have been mentioned. I also came across OpenCM, which seems to have merit too.

Is it that openCM is little-known, or because of some impediment in the software? Or lack of track record for a newly developped cm?

Re:How Does It Feel Linus?

[Anonymous Coward]

proprietry was, and still is the best tool for the job.

If he had not used Bitkeeper then the alternatives he is looking into now would not even exist.

Re:How about...

[MemoryDragon]

Actually subversion has a commerical backing. But those who wrote the thing simply wanted to keep the fanboys out of the LKML. And they are right in their analysis, they have a kicksass version control system but their approach is not suitable for the needs of the Linux kernel developers. What Linus and others really need is a highly distributed system and subversion (even SVK) is not exactly that, SVN is more like a drop in replacement for CVS without the problems CVS has, and it has its fair place in the world, and most who use it love it. The answer is just the style the people who do subversion generally have, very honest and professional.

Consistancy with OS

[OSXexpert]

Wasn't it Stallman that came out a while ago in interview and said Linus and his Version Control usage of BK was inconsistant with the open source movement and Stallman's views on such? Interesting, I wonder if Linus had a change of mind due to Stallman's public distain for Linus' version control choices.

Geeks put out a hit on BitKeeper.

[Anonymous Coward]

"Now they'll be known for screwing over Linus Torvalds. I wonder how well they'll fare in future technology evaluations? I can hear the discussions now: "Gee boss, BitKeeper is nice and all but if they screwed over the guy who writes Linux , how do you think they'll treat us after they have our money?""

Translation: We want REVENGE!

Seriously, give it a rest. No one's going to fight your battles for you.

Re:Other reasons...

[omb]

1. Linus robustness in these issues is legendry, he appeases NO ONE

2. Bitkeeper is a vary, very good product.

3. Because of lm, it is unusable in the Kernel context.

4. The good reason is lm's attitude, and he owns the product. Sadly, he has just completely destroyed all _TRUST_ in his product

If you read about 'git', think and stop making silly posts, you will understand what is going down.

The question no one has asked

[Anonymous Coward]

How much has Linus use of Bitkeeper hampered the development of a free software alternative? I mean, if Linus had used a free SCM since the beginning, chances are that much more effort would have invested in that very same system to make it better. Unfortunately all the past publicity has helped only Bitkeeper instead of the free software movement.

Thank you

Re:The question no one has asked

[tftp]

Use of BK probably stimulated the development of a free software alternative. Back then SVN was in early betas, arch was convoluted (some say it still is) monotone did not exist, and there was nothing else. But today, after BK has proven what feature set is necessary, some alternatives do exist.

Re:The question no one has asked

[SA Stevens]

One of the prominent facts of Open Source programming history, going waaaay back, is that something that works good and commercial has to exist first to copy and replace. This can be said about large parts of the GNU toolchain. It's part of the 'culture' for the name of the GNU replacement to be a pun or joke on the name of the original tool.

Big conglomerate groups of random people on the Internet aren't that good at developing something totally original. It's much more common to let some commercial entity 'develop the product spec' first.

Re:The question no one has asked

[Anonymous Coward]

Emacs was the start of the GNU project waaaay back then what non-free program did it copy?

The reason that GNU copied Unix has a very solid practical reason: modularity. This allowed to replace the system bit-by-bit instead of working in an all proprietary and hoping that some day your all origial OS free os would be ready to

Re:The question no one has asked

[BobNET]

Emacs was the start of the GNU project waaaay back then what non-free program did it copy?

EDLIN, I believe.

Re:The question no one has asked

[Anonymous Coward]

httpd is the first thing that comes to mind as open source software that didn't copy commercial software.

Unless you're going to try to point to other hypertext systems that were around before, in which case it's easy to just say that with very few exceptions, most software is just iterative improvements over old stuff. Revolutionary new software is pretty fucking rare.

Re:The question no one has asked

[SunFan]

It probably has not hampered anything, because good change management is _very_hard_ to do well. CVS has sucked tremendously for a decade or so already, and only now do better options exist (Subversion, e.g.).

Re:The question no one has asked

[RedWizzard]

How much has Linus use of Bitkeeper hampered the development of a free software alternative? I mean, if Linus had used a free SCM since the beginning, chances are that much more effort would have invested in that very same system to make it better.

Linus could not have chosen a free alternative because no such thing existed. The options (CVS, Arch, Subversion, etc) where not up to the task. So his decision hasn't hampered their development at all, in fact it's stimulated their development by highlighting the missing functionality.

Re:The question no one has asked

[Anonymous Coward]

Funny, but let's inject some fact here: Emacs stands for Editor Macros (for the ancient editor TECO).

Re:The question no one has asked

[leshert]

Emacs was the start of the GNU project waaaay back then what non-free program did it copy?

Emacs existed before GNU. It originated as a macro library for the TECO editor.

Re:The question no one has asked

[winwar]

Actually, a lot of people have asked the question.

You must just have been hiding in a hole somewhere. Say, you aren't a terrorist are you? :)

Re:The question no one has asked

[SA Stevens]

There were and are non-free versions of Emacs.

There is Gosling Emacs, which is non-free and the source code is not available. Stallman did NOT 'invent' emacs.

Re:The question no one has asked

[dvdeug]

One of the prominent facts of Open Source programming history, going waaaay back, is that something that works good and commercial has to exist first to copy and replace.

Like what, Emacs or Perl? There's a number of examples of original open source work. I'm not sure there's much difference between the copy rate of Open Source and the copy rate of proprietary software; I mean, I'm typing into IE (Netscape, Mosiac) on Windows (Macintosh, Amiga, Xerox) and could run Microsoft Word (WordPerfect, WordStar, many ancestors).

Re:The question no one has asked

[cheesybagel]

Gosling Emacs (1981) is posterior to Stallman's Emacs (1976). See this history timeline. At best you may say it started as something based on TECO.

Re:How Does It Feel Linus?

[jacksonj04]

I absolutely adore the way that people don't have the balls to post with a username any more.

The whole point of Open Source is that people have CHOICES ABOUT WHAT THEY USE TO DO A TASK. If they don't like it they can tweak the code so it does what they want.

Last time I checked, using BitKeeper constituted a CHOICE on the part of Linus. One of the core principles behind OSS.

CVS is abysmal, and Subversion doesn't fit the Linux development model. Proprietary software may be evil, but if it's the best tool for the task then why should people deal with inferior OSS crap, which may need serious hacking to make it do what they want, when there's the option - the CHOICE - of using proprietary software?

Re:How about...

[the eric conspiracy]

From reading the traffic on this topic it looks like SVK is a strong contender. Monotone is turning out to have really insufficient performance for managing the kernel.

Some of the other open source candidates also have performance problems.

DARCS

[Anonymous Coward]

DARCS

Clearcase?

[Anonymous Coward]

How about IBM opensource Clearcase? Its the best SCM I have ever used, and its got all the facilities that Linux needs. It would demonstrate if IBM is truely comitted so opensource, rather than using it as a marketing tool.

Re:How about...

[Tadghe]

"... but clearly something as wonderful as subversion.."

As I've pointed out before....

Until the Subversion folks learn about a little thing called "Security"*1 Subversion is pretty much unless for the real world.

Until Subversion has the ability to use a repository a bit more robust than sleepycat's db *2 (hey, sorry I'm corrupt, think I'll just go wipe 3000MB of commits now...) Subversion is pretty much unless for the real world.

notes:

1. Layering something as weak as HTTP AUTH/Basic on top of the Apache2 interface is not creating a "Secure" Subversion.

2. I know they have some sort of Filesystem repository now, but honestly, after watching Subversion's db repository "eat" several gigs of data in trials, I'll pass...

mod parent up

[Anonymous Coward]

ClearCase is a state of the art SCM for fast LANs. That's not what we have here. And yes, you do need a dedicated ClearCase administrator AND an IBM Rational support contract.

Re:How about...

[Artega VH]

My experience is that the berkley db orgiginally the only choice in SVN sucked. I had a full database lock with a small project witihn 2 week. However now that I'm using the fsfs repository its been smooth sailing.

I really like subversion, better than CVS in every way I know... except the checkout and lock functionality.

In terms of security I fail to see your point. How is subversion not secure? Perhaps you set it up incorrectly? It sounds to me like you're simply trolling... whoops i bit..

Re:Slashback? Some info on Bit Keeper

[sonoluminescence]

wow good link, I'd give you a mod point if you didn't have so many already.

When I read this on slashdot I thought, "I can't be the only one who has no idea what SCM means, my first thought was google, then wikipedia for SCM, never even thought of looking for bitkeeper on wikipedia.

This is really quite illuminating on a a topic I know nothing about. It's amazing the average linux user can be so unaware of controversy going on in the kernel development world.

OSS/FS Software Configuration Management (SCM)

[dwheeler]

See Comments on Open Source Software / Free Software (OSS/FS) Software Configuration Management (SCM) systems for more information on OSS/FS SCMs. There are several relatively mature centralized SCMs, but the distributed ones are less mature. See paper for details.

Re:OSS/FS Software Configuration Management (SCM)

[Anonymous Coward]

I think that author gives too little attention to Darcs. He basically treats it as a curiosity.

The strength of Darcs is in the *theoretical* foundations. Don't worry about the speed or the fact that's written in Haskell, those implementation details.

Because Darcs attempts a theoretical foundation, it means that as its features become more complex and/or require optimization, there will always be a safety net to tell the developers what the results should be and what invariants should apply. Other products designed in an "ad hoc" manner will simply hit a brick wall at some point.

I don't think this should be underestimated.

Re:OSS/FS Software Configuration Management (SCM)

[dwheeler]

Actually, that's more or less what it says. Here's the text:

Darcs, in particular, is very interesting for its technology. From what I've seen, darcs is currently more of a prototype of some very innovative ideas for SCM, and maybe a tool for smaller projects, rather than a useful tool for large projects, though it can be used.

A few people are looking at SCMs as an interesting technology area to do research in. We need those people, because they create the great new ideas that we want to use! Please research!

But most people are looking for a tool to use, today, to solve their problems in managing very large, complex systems. There are risks to using Darcs that way.

Microsoft vs. BitKeeper

[jbolden]

I'm sure Microsoft doesn't like the .mono project very much or the 1/2 dozen word document readers but you don't see them trying to enforce license agreements like the one BK was going for.

ah, that was half of my intent!

[Phil Urich]

I've got good enough Karma for my uses anyways; I felt like just laying on the irony as heavy as I could slosh it out.

Re:ah, that was half of my intent!

[tomhudson]

I hear ya. Especially since everyone's all in a twist about this, and that most of them haven't read up on it, and that it wasn't Linus who dropped Bitkeeper, but that Bitkeeper was more than a little pissed that

1. OSDL had two staffers who were trying to reverse-engineer their product, contrary to the terms on which they were donated the licenses
2. Bitkeeper tried to reach some resolution with OSDL over this over a period of about 5 weeks, but was unable to
3. Bitkeeper has stated that they would have no problem comping (giving complementary licenses) to Linus and Andrew again, but not while they are employed with OSDL

Seems to me that a deal is a deal. The deal for the free licensing included a "no reverse engineering" clause. You can be sure people would scream bloody murder most foul if someone violated the GPL license, so it's kind of hypocritical to get all upset about Bitkeeper being upset that someone's violated their license agreement after getting freebies that were widely acknowledges as being extremely helpful.

There's room for both proprietary and open and free and shareware and public domain software - it's a big world out there.

This whole thing is a tempest in a teapot compared to, say Jonathan Schwartz's latest brainfart against the GPL.

We need a new poll:

Sun and the GPL/Open Source schizophrenia thingee

[ ] Sun doesn't "get it"
[ ] Jonathan Schwartz is off his meds again
[ ] If this is Tuesday, open source is bad
[ ] MicroBitch paid them off ... again ...
[ ] "There's no such thing as bad publicity"
[ ] We have to extract *some* value out of our SCO Unix license
[ ] All the above

For example, I have no problem with Sun keeping Java closed. That's their decision, and their right. Just like its' my right to either use it or not. What I *do* have a problem with is the lies and fud (well, maybe except for Troll Tuesdays :-).

So, back on-topic, OSDL was, IMO, in the wrong on this one. That the employees in question were doing the reverse-engineering on their own time is irrelevant. It still broke the terms of the deal. It's also a pretty shitty thing to do, akin to biting the hand that feeds you. I'd have been pissed, too.

So long and thanks for all the fish, Larry!

[borgheron]

Guess it's finally happened. When we found out about your restrictive license that was enough. I guess Linus has finally seen the light, thank goodness.

Told ya larry! See ya, wouldn't wanna be ya!

GJC

GIT is coming along

[Anonymous Coward]

I've seen stuff from the subversion people asking people not to bug Linus about using subversion (it isn't distrubuted, can't easily be made distrubuted, and they ask people to quit bugging Linus about using it (they agree that it isn't what Linus is looking for). Linus built something called GIT that uses part of the Linux kernel to manage merging large trees and portions of trees with date/time/directory information. He has already shot down people who ask about using SQL (and related databases) as he states that "they are too generic and too slow", and has stated that he can stat an entire tree several hundred times more quickly than the fastest database, becuase his entire engine is optimized for kernel work. Linus has already taken a week off to do bitkeeper replacement/tool creation. It may take another week before the whole thing is finished. It seems Bitkeeper clarified how Linus can maintain large chunks of tree (big merges and so on), and Linus wants to take the parts he liked in Bitkeeper, replace the parts he hated, and go with that. Tweaks to follow.

Re:GIT is coming along

[kabz]

Oh well, I guess that means my mention of the totally kick-ass MKS Omniworks source control system would get shot down as it is based on Oracle.

I am really pleased that Linus has some stuff he's working on though. I hope it all comes together for him, and us. I'd hate to see the Linux kernel end up getting forked over this kind of fighting.

Re:How Does It Feel Linus?

[xxavierg]

i think the issue is Linus is so dead set against closed source drivers in the kernel, but it is ok to develop used closed source tools? i think that is wherein lies his hyporcrosy. if there were no open source tools to do what he needed, he should have written his own. isn't that what the fanboys are always screaming?

it's like a vegetarian eating a hamburger because he it's to inconvienant to go to the health food store for a tofu sandwich.

i have no problem with closed sources if it is a better choice, so what i say matches what i do. (btw, i am a vegetarian, so if i go out with people, and there is nothing for me to eat, i just do not eat. sometimes i get really hungry, but i do not go against my ethics because it's easier...)

Open Source to the rescue!

[Duncan3]

Software: If you charge for it, they will flee. Paying for software is so 1980, and why should we, someone is more then willing to make a free version of anything important in their spare time.

Humans will only pay for food, shelter, and entertainment (i.e. sex, drugs, and rock and roll). NOTHING ELSE!

Perhaps the SCM Solution is not the problem

[Anonymous Coward]

In all these discussions the question seems to be:
What tool can we find to fit Linus' methodology.

The question that nobody ever asks is:
Do Linus' methodologies actually benefit the project?

I'm going to say something that will make people angry, but I don't mean it as flame bait:
The Linux kernel is in absolute disarray, has been since 2.2.x, and is only getting worse as the project grows.

The kernel development process is a highly distributed one, with individuals all over the world implementing their own corner of the universe and sending it in diff form to the Great Linux Mutex in the Sky: Linus.

All these people, all over the world, doing things their own way. Limited cohesion, inability to pick a good design and stick with it for more than 5 minutes. Here are two examples from the stable kernels ...

Firewalls:
2.0.x: ipfwadm
2.2.x: ipchains
2.4.x: iptables

Event Handling:
2.0.x: None
2.2.x: None
2.4.x: dnotify
2.6.x: inotify

The VM system has been swapped out - several times! - in "stable" kernel branches. Nowdays Slashdotters are saying that it is the distribution's responsibility to maintain a stable kernel tree.

So, here's my point. Maybe the BSDs really have something going for them by maintaining a centralized development model where major changes are planned, documented, and subjected to stringent review across the board before being included in the kernel.

Here are the above examples for FreeBSD:
Firewalls:
2.x: ipfw & ipf
3.x: ipfw & ipf
4.x: ipfw & ipf
5.x: ipfw2 & pf

Event Handling:
2.x: none
3.x: none
4.x: kqueue
5.x: kqueue

FreeBSD consistently implements well thought out systems and incrementally improves them over the years. Time and time again FreeBSD issues solid stable releases. Even the releases that are publicly declared not ready for production are more stable and have fewer dark hairy corners than Linux.

Maybe it's time for Linus to re-evaluate his model and bring some true software __engineering__ to the Linux development process. Instead of being the Great Mutex in the Sky, he can be the great organizer and enabler of an organization of people working together to implement well thought out, well designed systems.

Just a thought from an anonymous coward.

Re:Perhaps the SCM Solution is not the problem

[DarkMan]

I'm not convinced by your arguments, although I do note that most of the sub-points are valid. Specifically, I reject your conclusion.

If Linus did Linux the *BSD way, a few things would have to change.

Firstly, and foremostly, he would have to develop, and guide, the userland as well. That is part of the reason for the apparent stability of BSD interfaces - for some kernel->userland interfaces the implemetnation is hidden (for example, things that are used only by libc). Whether this is a good or bad thing on the partof the BSD's is irrelevent, it is a fundementaly different approach.

Currently, Linux provides a set of kernel level interfaces, and the GNU libc utiliese some of these to provide the standard C library. Some of the kernel interfaces are standard, (POSIX file stuff, sockets etc), some are not (Firewall, some threading interfaces etc).

Most fundementally, the reason I reject your conclusion can be best expressed as a question: "If the linux process is so poor, why is it more popular than *BSD?"

One can discuss that question ad nauseum, but I suspect that part of the reason is that it is _not_ tightly controlled. If you have a method for making some part better, the Linux apprach says: show me the code, convince us it's better. The *BSD approach says: Show me the code, convince us it's better, get somone on the core team to sponser it, shedule it for the next (or prehaps one after next) release, and wait.

Thus, Linux has a lot more mobility, and can utilise new approaches sooner than *BSD - for good or ill [0].

My personal thoughts is: Linux is not broken. The development model works as well now as it used to, thus there is no need to change. Let the two approaches co-exists, let people use whichever they feel like. If that means that Linux becomes a testbed for ideas, and *BSD cherry pick, then so be it, that's fine. If that means that Linux evolves and *BSD stagnates, so be it, that's fine. If that means that Linux is unstable and *BSD is stable, so be it, that's fine. In practice, the last two will be indistinguishable for a long time after one of them occurs.

Let them be different - choose your favourite model, and be happy with it. But if the other one results in something you prefer, consider that both are trade-offs, and neither is perfect.

[0] This can bring problems. The various VM saga's are prehaps a canonnical example here.

Re:Perhaps the SCM Solution is not the problem

[Anonymous Coward]

Most fundementally, the reason I reject your conclusion can be best expressed as a question: "If the linux process is so poor, why is it more popular than *BSD?"

The reason I reject your rejection of the conclusion can also be expressed as a question: "If the Windows process is so poor, why is it more popular than Linux?"

You seem to be arguing that successful implies good. Yes, there is some correlation, but it's just so far from a universal law that it's meaningless.

Re:Perhaps the SCM Solution is not the problem

[Anonymous Coward]

Linux has consistently more contributions because the barrier to entry is very low. This does not necessarily a good thing, as hilighted above.

Re:Perhaps the SCM Solution is not the problem

[Tim C]

Most fundementally, the reason I reject your conclusion can be best expressed as a question: "If the linux process is so poor, why is it more popular than *BSD?"

No offence, but logically-speaking an argument by appeal to popularity is pretty weak. There are a great many reasons why a thing may become popular, of which superior quality is only one.

Re:Perhaps the SCM Solution is not the problem

[m50d]

I'm just fed up of the mess he's made of CD recording, which comes from the bigger mess there is of the whole device layer. The scsi module should be moved up a level, and ide should work under scsi like usb or anything else does. 2.6 is too unstable because of the stupid decision to have all the development in the stable tree. Also, I want my reiser4, if it needs to go up into the VFS layer let it. Anyway, want to start a fork? Fact is there's too much hero-worship of Linus to ever change him in the main kernel tree, or persuade him he needs to change. The whole point of open source is that when he does something really boneheaded (when, not if, everyone makes mistakes) and won't go back on it, we can fork. I think the time to do that is now.

Re:Perhaps the SCM Solution is not the problem

[Anonymous Coward]

Firstly, and foremostly, he would have to develop, and guide, the userland as well. That is part of the reason for the apparent stability of BSD interfaces - for some kernel->userland interfaces the implemetnation is hidden (for example, things that are used only by libc). Whether this is a good or bad thing on the partof the BSD's is irrelevent, it is a fundementaly different approach.

Your way of thinking is the exact reason linux kernel development is broken, and linux development at large is broken. Everyone works on their own little task, very little give and take between the projects happen, and the result is a cacophony where the user is left picking up the pieces. Users don't care about how technologically marvelous their kernel is, they want stuff to just work. See how long it took to get the kernel to a point where you could insert a usb device and have it mount on the desktop, like windows and mac os have done for years and years. That's a direct consequence of the lack of integration in the design. And this happens thorughout the entire linux community. The most interesting stuff happens only when projects are forced to work together, like how the cooperation between X.Org and the KDE and GNOME projects is producing the new X extensions that will allow radically new UI behavior.

You don't necessarily need someone telling everyone what to do. What you do need is to organize a method of constant conversation between all the major projects that make up the OS. In that case, userland would quickly make kernel developers realize what designs they need to change, and what designs they shouldn't change. Likewise, userland could more easily cope beforehand with kernel design changes. Stuff that could be shared between projects would be recognized. Overall development speed would pick up, and overall usability owuld rise by an order of magnitude.

A bazaar is nice, but it only works if not everyone tries to talk at the same time. Otherwise you just get a cacophone where you lose your way.

Re:Perhaps the SCM Solution is not the problem

[slux]

On the other hand the BSDs have been forked to four separate versions because someone has wanted to redo a certain part. The rewrites on Linux have always been improvements and in the case of ipfw etc. the kernel still retains compatibility to the older schemes.

You seem to be saying that everything should be done only once and done right. I don't buy that. What may initially seem like a good idea can always be shadowed by a better one someone has later. You also can't write a piece of software as large as a kernel by planning each part to be the best and as feature-complete as it can or you'll end up like the Hurd. This is particularly true in the open source development process. Had Linus originally planned for Linux 2.6 when he started writing it I doubt we'd have Linux now.

Re:Perhaps the SCM Solution is not the problem

[Anonymous Coward]

"If the linux process is so poor, why is it more popular than *BSD?"

equally,

If windowsXP is so poor....

Re:Perhaps the SCM Solution is not the problem

[Anonymous Coward]

People use windows because...Microsoft will eat baby kittens if they don't.

Re:Perhaps the SCM Solution is not the problem

[chrysalis]

Yes, critical parts of the Linux kernel have been redesigned and rewritten several times. Is it just for fun or to break things? No. It was always to rewrite things in a cleaner and more modular way.

Yes, iptables has obsoleted ipchains and ipfwadm. But the older interfaces are still there is you need them.

And yes, the VM system is being constantly tweaked. Linux targets a wide range of hardware, from embedded devices to 1024-ways NUMA supercomputers, and it tries to be optimal for all kind of applications. This is not a trivial task. Do you really think that it is humanly possible to design the perfect VM that you can stick with for years? Definitely not.

Rewritting things (in a better way) may introduce breakage but it is the best thing to do for the long term.

FreeBSD tends to be conservative. In order to fix bugs or to implement new things, hacks over hacks over the old code are added. This is a safe bet, but this "stability" hides the fact that parts of the code have just become unmaintainable.

This is why the DragonFlyBSD project is slowly rewritting major parts of the kernel, in a modern, simple and maintainable way.

Also maybe the reason why the Linux and DragonFlyBSD codes are changing so much is because people prefer coding over trolling.

Re:Perhaps the SCM Solution is not the problem

[rathehun]

I don't have the neccesary technical skills to evaluate the rest of your arguement, but this part needs to be looked at more carefully.

"If the linux process is so poor, why is it more popular than *BSD?"

"If the Windows process is so poor, why is it more popular than Linux?"

Discuss.

Re:Perhaps the SCM Solution is not the problem

[Anonymous Coward]

Uh, so your argument is that since BSD advances more slowly than Linux it's a model that should be followed?

I don't get it.

Re:Perhaps the SCM Solution is not the problem

[cpghost]

If the linux process is so poor, why is it more popular than *BSD?

Just as Windows...

the Linux apprach says: show me the code, convince us it's better. The *BSD approach says: Show me the code, convince us it's better, get somone on the core team to sponser it, shedule it for the next (or prehaps one after next) release, and wait.

What is the real difference here? In both cases, code submitters have to get approval by maintainers: with Linux, there's only one mainainer. With BSD, there are many committers. Actually, BSDs approach seems even more failproof than Linux', because you don't have to worry about Linus going on vacation (or some BitKeeper issues). You just ask another committer to submit your code to review.

Thus, Linux has a lot more mobility, and can utilise new approaches sooner than *BSD

Hmmm, the -CURRENT branches are very mobile too, just as with Linux. There's one important difference though: BSD kernel interfaces (external and internal) tend to be a lot more stable than Linux' interfaces. This is both good and bad: with Linux, you can experiment with new interfaces and paradigms. With BSD you can do the same. But with Linux, the experiments quickly make it into the public kernel, even if they have to be retracted later. With BSD, this happens too, but it's very rare.

Let them be different - choose your favourite model, and be happy with it.

Right!

Re:Perhaps the SCM Solution is not the problem

[Monoman]

Most fundementally, the reason I reject your conclusion can be best expressed as a question: "If the linux process is so poor, why is it more popular than *BSD?"

I'm not picking sides but most fundamnetally, the reason I reject your point is:

Quite often, being the most popular is not determined by being best. It is usually about being good enough for most people.

* VHS vs Betamax - a classic example
* Microsoft Windows - if you are here then you know why.
* American Idol winners - a talent show that ends in a popularity contest.

Re:Perhaps the SCM Solution is not the problem

[Anonymous Coward]

"and the result is a cacophony where the user is left picking up the pieces."

FALSEHOOD#1 In some "canonical" point of view, is not users picking up the pieces, but Linus Torvalds, the Benevolent Dictator cherrypicking them (and since Torvalds is just, well, Torvalds all the time, he can follow an iron-hard criterium about what he accepts and what he doesn't for his source tree).

"Users don't care about how technologically marvelous their kernel is"

FALSEHOOD#2 *Some* users doesn't care, *some* users do care, and *all* users have their own choices known to them.

"See how long it took to get the kernel to a point where you could insert a usb device and have it mount on the desktop, like windows and mac os have done for years and years. That's a direct consequence of the lack of integration in the design. And this happens thorughout the entire linux community"

FALSEHOOD#3: It is well known that Windows NT 4.0 didn't have USB support AT ALL on its days, and it still doesn't have USB support today. First Microsoft OS with USB support was Windows 2000, about, well, year 2000; Linux did have quite good USB support by mid 2001, so nothing about "years and years", and that mainly because Linux wasn't considered by those developing USB devices, not because some inherent Linux inhability.

All in all, while your post *seems* to point out valid points and even migth have some consistence, the reality is that it is mostly crap. Linux distributions DO work despite all your presumed cacophony.

Re:How about...

[aardvarkjoe]

Until the Subversion folks learn about a little thing called "Security"*1 Subversion is pretty much unless for the real world.

Funny how a piece of "useless" software can get thousands of users. You'd think that they would notice that they weren't getting anything done.

Until Subversion has the ability to use a repository a bit more robust than sleepycat's db ... I know they have some sort of Filesystem repository now

If you know that that problem has been solved, then why are you still complaining about it?

BitKeeper just screwed the pooch

[tjlsmith]

Having Linus endorse BK was the best thing that ever happened to it, better than everything else put together.

Now they've blown it because of some suit's decision.

Re:BitKeeper just screwed the pooch

[cpghost]

Exactly. Hosting the Linux source code repository was their ONLY publicity asset. How many people in the IT industry ever heard about them if they were not connected to the Linux project?

I'd be tempted to say: Good riddance! but it's not so easy: if people were actively trying to mimic BK's software, sooner than later BK would they have been out of business anyway. Someone in Mgmt is just trying to stop a sinking ship from taking water: they are doomed to fail.

Begin with a solid theoretical foundation..

[Anonymous Coward]

I think we can learn from the fiasco of relational database products on this one (i.e., the standard, SQL, is not relational at all, and the true power of relational databases has still not been implemented in an actual product).

FIRST come up with a rational, rigorous theory of version control. What is a "change"? What does it mean to apply changes and their inverses? What is a conflict? Etc, etc. Come up with theorems and rules based on that foundation.

THEN implement it all in a product. Or better yet, many products, but base them all on the same theory. I.e., let them compete on ease of use, security, speed, etc., rather than just who has the best ad-hoc collection of features.

The author of Darcs does have a "theory of patches" but it's not rigorous (by his own admission). However it's a start.

I think having a solid theory, divorced from any particular implementation, would be a great idea, would yield a consistent set of terminology, and would allow those tough corner cases to be solved with more understanding.

Does anyone else agree?

PS: McVoy himself has talked about the mix of theoretical and practical that BitMover employed to create BitKeeper. It's not just an academic argument. You'll get a better product if you THINK then ACT rather than leaving out the first step. And yes, besides the fact that it's not open source, BitKeeper *is* the best SCM system out there.

PARENT INSIGHTFUL--*NOT* TROLL

[Danuvius]

Modded by codemonkeys or luddites though. %)

In true Open Source fashion...

[Anonymous Coward]

Linux kernel hackers, suddenly finding themselves without the tools they are accustomed to, will simply write a functional replacement in an embarrassingly short amount of time.

The key features that make BitKeeper nice will be quickly implemented in a free and open project, and BitKeeper will fade into obscurity.

Taking away the "free" BitKeeper binary was a huge tactical blunder from the perspective of the people behind the decision, but I'm glad it was done.

Linus didn't scale.

[tres3]

For those of you that don't remember the lost patches and snails pace of Linux development from 2000-2002 I invite you to browse through the LKML history. We had a real problem with the fact that Linus was unable to keep up with the vast number of patches that were coming in through email and when he went on vacation there was no reason for anyone else to even try to submit patches. We all said that Linus doesn't scale! There is no way that 2.6.x would be anywhere near what it is today without some form of Source Control Management (SCM). Developers had been trying to get Linus to use some form of SCM for years but none would do what Linus wanted them to (most were not distributed enough). Larry offered Linus, and other developers, Bitkeeper for free but insisted that they not use it for non Open Source projects and that developers of competing SCM not use it at all (without paying for it). It was nice of him to give us that chance but everyone knew that the strings would, sooner or later, tie us in knots. Bitkeeper served its purpose well (relieving the Linus load) for years; all developers had to do was to send an email to Linus saying "pull the patch from my tree". This solution was a vast improvement on the previous way of doing things and the pace of kernel development increased significantly. It also enabled Linus to hand off large chunks of the kernel to other maintainers so he could focus on the parts he is good at.

The most important thing that Bitkeeper did was to show the developers what was possible with the right tools. Anyone that reads the LKML knew that this day would come sooner or later and many of the SCM developers have used the time to improve their tools. What we really need to do is thank Larry for the use of his program (it was a great help) and move on. I don't think that Linus and the other kernel developers will ever go back to the days before fine grained changelogs, distributed source trees, and the ease with which patches from any one tree can be applied to any other tree. I think, if anything, that the biggest thing that dropping Bitkeeper will do at this point is to accelerate the development of better (and more distributed) SCM's.

Thanks Larry! And more importantly: Thanks Linus, Alan, Andrew, Marcello, Rik, et al. Your work and dedication is much appreciated! ;-)

P.S. Kernel.org has a new SCM written by Linus (in his directory) that is available for your perusal.

Re:Linus didn't scale.

[Anonymous Coward]

Thanks Larry! And more importantly: Thanks Linus, Alan, Andrew, Marcello, Rik, et al.

Two out of those first three never used BK.

-1, troll

[Anonymous Coward]

not even a good one either. try harder next time.

Re:How Does It Feel Linus?

[tres3]

Linus is not opposed to closed source drivers in the kernel. He is against allowing closed source drivers in the kernel from affecting kernel development. He will not keep module interfaces from changing and he will not help people that have incorporated closed source drivers into the kernel when their shit breaks. nVidia has to spend alot of effort keeping up with the kernel changes so that their drivers continue to work and that is their decision. It would be easier if nVidia just submitted source code so that the kernel maintainers could help them make changes when the kernel interface to modules changes. It would also make my life easier so that I didn't have to keep two versions of the kernel and two versions of xorg.conf on my system but nVidia has made their decision. Linus hasn't gone out of his way to make their lives more difficult but he does refuse to go out of his way, even in the slightest, to make their lives easier. In my opinion that does not constitute dead set against closed sources drivers. Obviously, your opinion differs.

Sarcastic post...

[rice_burners_suck]

You know, I wasn't going to say this, but I can't help it...

As we all know, Linus has more resources at his disposal than Bill Gates. Why can't he just have some of his minions design, from scratch, a distributed source configuration management package that can do everything he needs, and have it ready within six months? Then, he and his crew could suffer along with Subversion and the distribution problems it will pose for them, for six months, before Linux can be hosted on Linus' own DSCM software.

It shouldn't be quite that hard to do, with all the resources he has... When Theo had a problem with (I believe it was) the license for the SSH program included prior to OpenBSD 2.6, he thought about it for a while and then busted out his own implementation. If he could do that, then Linus with all his resources can bust out a DSCM.

Yes, this post is totally sarcastic. But seriously, who said you can't take Subversion, rip out its guts, and make it distribution-aware.

Re:Sarcastic post...

[the eric conspiracy]

Why can't he just have some of his minions design, from scratch, a distributed source configuration management package that can do everything he needs,

If you read the threads involved you would not have written this message. The question isn't whether or not they are going to do this, but rather what existing project, if any, they are going to leverage to reach this goal.

Re:Sarcastic post...

[paskie]

And what you write is exactly what Linus is doing now. It is not a full version control system, but it can likely do the basic job for the kernel, especially for distribution of the tree history between developers; it is probably likely that they will use (test, improve) various version control systems on top of this, actually. (Which should be possible.)

Re:Sarcastic post...

[rice_burners_suck]

Bah! Real Programmers (tm) don't use version control and code without breaking the software! Ba hahahaha hahaha haha!

Some ADVOCATING design-by-committee?

[Ars-Fartsica]

It has been known since ancient times that the perfect model of governence is the benevolent dictator. Don't knock it! Seriously though, the other systems have had issues you are not bringing up. How did the apparently perfect governence of FreeBSD allow the 3.x series to come to pass, for example?

RIchard Stallman Knew This Would Happen

[Anonymous Coward]

"The spirit of the Bitkeeper license is the spirit of the whip hand.
It is the spirit that says, "You have no right to use Bitkeeper, only
temporary privileges that we can revoke. Be grateful that we allow
you to use Bitkeeper. Be grateful, and don't do anything we dislike,
or we may revoke those privileges." It is the spirit of proprietary
software. Every non-free license is designed to control the users
more or less. Outrage at this spirit is the reason for the free
software movement. (By contrast, the open source movement prefers to
play down this same outrage.) "

- Richard Stallman Oct 13 2002, 3:50 pm

Maybe he is not a mad man, but actually a very wise man.

Here is the discussion link: http://groups-beta.google.com/group/linux.kernel/b rowse_thread/thread/a98de7edab73f365/7d68ee9f364e9 3f6

Re:RIchard Stallman Knew This Would Happen

[linuxdot]

Richard Stallman Oct 13 2002, 3:50 pm Maybe he is not a mad man, but actually a very wise man. FUCKING EH!

Re:RIchard Stallman Knew This Would Happen

[Some Bitch]

Maybe he is not a mad man, but actually a very wise man.

He's both, he's an insane genius. In 20 years time we'll all be saying "We should have listened to RMS". At times he can appear petulant but if I'd been spent a couple of decades warning people about non-free software and been derided while been proved right time and again I'd probably not be a happy bunny either.

One day we'll listen to him BEFORE things start to go all runny but that will probably cause the universe to end.

Re:RIchard Stallman Knew This Would Happen

[burns210]

Him pointing out that a developer of a proprietary system owns and has control of that system doesn't make him wise, that makes him conscious.

Flame me if you want. I don't buy that software development is a religious experience.

Re:RIchard Stallman Knew This Would Happen

[toby]

Maybe he is not a mad man, but actually a very wise man.

...It took you 20 years to realise this???

Re:RIchard Stallman Knew This Would Happen

[syousef]

Maybe he is not a mad man, but actually a very wise man.

Nup, I've met him and he's a mad man, a total fruitloop, and not a very nice man.

Mad and right are not mutually exclusive though.

Re:RIchard Stallman Knew This Would Happen

[Anonymous Coward]

One day we'll listen to him BEFORE things start to go all runny but that will probably cause the universe to end.

No, those are the guys who are saying that man made blackholes could be a "problem" and the proponents of gray-goo and Kurzweil's singularity. Those are the ones we should listen to before the world ends. But RMS is probably right about software issues.

Re:RIchard Stallman Knew This Would Happen

[cpghost]

One day we'll listen to him BEFORE things start to go all runny

Just like... avoiding binary drivers?

trac

[moosesocks]

although it's based upon SVN (which I know has its downsides), I've found trac to be the best package to work with in terms of version control and project managment.

perhaps our model of a versioning system is outdated just because everyone keeps thinking of CVS. maybe we really just need a graphical set of tools to manage source.

there are definite usability advantages to using a visual/sptaial interface for managing large quantities of information (such as the linux source tree).

and yes, I am a mac user. I also run linux. I also see the necessity of the command prompt and use it for scripting in addition to applescript (both of which are soon going to become unnecessary. . Of course, for everyday tasks, when asked to choose between the Aqua GUI and the commandline, one would be a fool to pick the commandline. The linux GUI paradigm is absurd -- X is basically a wrapper for the command line -- most GUIs and programs are based around the limitations of their command-line equivalents. It just pains me to see so much effort being wasted on open source development due to poor project organization.

If you can picture the entire source tree and kernel structure in your head, great, there are about 9 other people in the world like you. If not, then kernel development will slow to a crawl. Granted, any usability expert will argue that in order to be functional, the developer should not HAVE to know the inner workings to contribute, but still.

Re:trac

[Reverend528]

Of course, for everyday tasks, when asked to choose between the Aqua GUI and the commandline, one would be a fool to pick the commandline.

yes, there's clearly no advantage to having every program on the system readily available and the ability to string commands together with pipes. Plus, we all know how GUIs are inherently more scriptable.

Re:trac

[Anonymous Coward]

Trac is an awesome project management web application with Subversion integration. It doesn't get in your way, and the wiki support in it is well thought out.

Re:How Does It Feel Linus?

[malloc]

The whole point of Open Source is that people have CHOICES ABOUT WHAT THEY USE TO DO A TASK. If they don't like it they can tweak the code so it does what they want.

Last time I checked, using BitKeeper constituted a CHOICE on the part of Linus. One of the core principles behind OSS.

I have to disagree. The whole point of Open Source is freedom. Freedom from any group being able to "lock you in", freedom to adapt software to meet your own needs, freedom to share and benefit from each other's work so we can focus on solving other problems. This freedom leads to having more choice, but using closed-source software only reduces freedom and thus the choices you have available.

You could say that Open Source is about preserving choice.

Malloc

p2p?

[robbo]

With all the hand-wringing about coming up with non-infringing uses for p2p software for the Grokster case, I'm surprised no one has suggested that we take an open source p2p client and give it some version smarts...

my 2 cents...

Re:p2p?

[lachlan76]

Because when ANYONE can put code into it it can't really be trusted. You'd need to sign the code, and have trusted keyservers.

Re:p2p?

[Anonymous Coward]

It's less complicated than that, not that it hasn't been done. With distributed version control systems, you don't allow just anyone "put code in"; what you do, is you allow just anyone to run their own repository server, and then have nice tools for browsing each other's stuff and fetching other people's changes into your own repository if you like them.

The keyserver/id part is usually taken care of by piping the commit access through ssh, and you only allow yourself to commit to your own server, other people run their own.

The "Peer" in "Peer to Peer" means a person of equal rank. If some people can put stuff in and some can't, and their is only on repository, it's not "Peer to Peer".

No

[lorcha]

For more information, read Please Stop Bugging Linus Torvalds About Subversion.

Thank you, and have a nice day.

Re:How about...

[bani]

whats wrong with svn co?

ClearCase blows donkeys

[lorcha]

Always has, and always will.

I can't count the number of times I've gotten build failures because I updated my view in the middle of someone else's big commit. WTF is that all about?

If I wanted that, I could use SCCS.

And don't get me started about the voodoo that is config.spec.

Wow, uncomfortable truth ...

[Anonymous Coward]

That coming from a software developer.

not true

[Fourier]

I don't usually reply to ACs, but I believe this to be inaccurate. Arch performance is quite reasonable, particularly when exploiting its revision library and hard-link checkout optimizations.

Perhaps you are thinking of darcs or monotone, both of which do have performance issues on large trees.

Re:not true

[jsebrech]

Just because it's fast enough for you doesn't mean it's fast enough for linus.

I RTFA. Apparently the bottleneck is merging new changesets from email. Seems BK is orders of magnitude faster than everything else out there for doing that.

Re:not true

[Anonymous Coward]

> I don't usually reply to ACs, but I believe this to be inaccurate.

I usually don't reply to people not replying to AC, but...

From the SCM thread:

"However, the SCM's I've looked at make this hard. One of the things (the
main thing, in fact) I've been working at is to make that process really
_efficient_. If it takes half a minute to apply a patch and remember the
changeset boundary etc (and quite frankly, that's _fast_ for most SCM's
around for a project the size of Linux), then a series of 250 emails
(which is not unheard of at all when I sync with Andrew, for example)
takes two hours. If one of the patches in the middle doesn't apply, things
are bad bad bad."

Linus dont want /reasonable/ performance. He want merge operations to be nearly instantaneous.

Cheers,

--fred

Re:not true

[Anonymous Coward]

I think you are the one who has it backwards. Arch proved prohibitively inefficient. Monotone apparently is being considered.

Perhaps arch is fast enough for whatever you manage; but if it's fast in general, why all the forks (bazarng or something like that) that emphasize performance.

Re:not true

[Fourier]

"If it takes half a minute to apply a patch and remember the
changeset boundary etc (and quite frankly, that's _fast_ for most SCM's
around for a project the size of Linux), then a series of 250 emails
(which is not unheard of at all when I sync with Andrew, for example)
takes two hours."

Generally speaking, applying a series of Arch changesets is a fast operation--it's nothing more than 'diff' and 'patch' with rename tracking. The bottleneck is running "tla changes" prior to commit, which determines what files have been altered by the series of patches.

Here is an email in which a knowledgeable Arch hacker demonstrates Linux kernel 'tla changes' benchmarks under 20s. That's a relatively minimal penalty, and it's incurred once per merge, not once per patch.

Re:not true

[Fourier]

In this post I provide evidence that Arch offers reasonable performance on large trees. Here is a post from the lkml thread that demonstrates Monotone's performance problems.

Do you have a link where has Arch been proven prohibitively inefficient? (Any benchmark older than mid-2004 is obsolete.)

Also, bazaar is not a performance-focused fork, although it does have a caching system that is not yet in mainline Arch. Its primary goal is to put a better UI on top of Arch. bazaar-ng is a completely separate project that shares no code with Arch, but is inspired by some pieces of Arch design.

How Does a closed society Feel Linus?

[Anonymous Coward]

By your logic then, living in a society is "closed source", and being a hermit is "open source". After all a society by definition will restrict your freedom to "do whatever you want". However we see that the "closed source" society brings as much if not more, than the "open source" hermit choice.

Maybe there's a clue there, that the "do whatever you want" isn't the best definition of freedom there is.

C'mon, You Knew This Was Going To Happen

[cmholm]

Yes, Linus had a perfectly good reason for selecting BitKeeper. And, one could reasonably anticipate that Bitmover would do what it did. Hell, if you read what Larry McVoy had to say back when Linus jumped onboard back in '02, he made it quite clear that he was going to get pissed if someone tried to reverse engineer the code. Richard Stallman doesn't have to say "I told you so", 'cause he already did, three years ago.

So, in conclusion, big frickin' deal. BK got a few years of valuable, free beta testing, Linus got some work done, and the Open source folks got a reminder as to why the Free source folks got religon.

Someone help me with this...

[JavaRob]

Maybe I'm delving into a discussion that I haven't researched deeply enough yet... but personally, while I do open-source some of my code, I have this dream of writing some difficult, closed-source software, and making money selling *that software*. NOT, for instance, selling support for that software. Support is the un-fun part of programming (does this make me anti-social?).

If I open-sourced this software, my huge time investment would immediately be lost -- other developers would take the code (and could avoid duplicating the years I spent, which is nice for them), make a few improvements, possibly package it up nicer, and distribute that. Yes, more people in the world can now use my software... but what should I do then? How do I fulfill my dream of working my ass off, then being able to *rest* for a bit and dabble in unprofitable work?

I don't like duplication of work any more than anyone else, but to some degree (especially for new products) it seems like a necessary evil to me. Where does the money come from in the all-software-is-libre model, especially for independant developers? I don't want to be RedHat. Selling retail boxes, configuration, and support doesn't appeal to me at all. Support is like solving the same boring problem, with tiny variations, millions of times, instead of tackling a fresh new problem and fixing it once and for all.

Re:Someone help me with this...

[Anonymous Coward]

Support is the un-fun part of programming (does this make me anti-social?).

So you wouldn't be giving any kind of support for your non-free software?

Re:Someone help me with this...

[Anonymous Coward]

I have this dream of writing some difficult, closed-source software, and making money selling *that software*. NOT, for instance, selling support for that software. Support is the un-fun part of programming (does this make me anti-social?).

One thing that might help you is that "support" doesn't necessarily mean "end-user tech support", which is certainly a non-enviable position. I'm grateful that people do it, because I certainly wouldn't want to either. In this context, "support" could just as easily mean

tackling a fresh new problem and fixing it once and for all.

; it just means that the customer (individual or business) might tell you what the "fresh new problem" is.

Re:Someone help me with this...

[hazah]

Have you seen nothing in how things are usually done in open source? First of all, you can sell the software, even if it is opensource. Second, no one is going to take your code, fix it up, and then sell it. That would be quite a large investment on their part. Instead, what you're more likely to find, is that people are sending you bug reports, fixes, suggestions. If you go the open source route, and you actually DO have verywell written software, you're more likely to have a community centered around it than anything evil you imagine.

Re:Someone help me with this...

[roman_mir]

you're more likely to have a community centered around it than anything evil you imagine. - agreed. But this does not mean that writing closed source software is antisocial at all. One can create closed source software and never release it to anyone, not even if they pay for it. Even this is not antisocial. When you create things you don't have to give or show or sell them to anyone, you can just enjoy it by yourself. This is not antisocial, it does not disrupt the norms of society, it is not rude, it is not illegal.

One can create closed source software and sell it and this is also not antisocial. After all, this software will help someone to do something, otherwise they wouldn't buy it, would they? It is very social in fact to create any type of software and release it as Free or free or sell it as not Free. It does not fall under the 'antisocial' tag.

From dictionary.com:
antisocial:
1. Shunning the society of others; not sociable.
2. Hostile to or disruptive of the established social order; marked by or engaging in behavior that violates accepted mores: gangs engaging in vandalism and other antisocial behavior.
3. Antagonistic toward or disrespectful of others; rude.

So RMS comes up with his own definition of antisocial and now he pushes it onto others and they accept it. But I do not.

Re:Someone help me with this...

[JavaRob]

So you wouldn't be giving any kind of support for your non-free software?

Not personally -- that wouldn't be the core of my business, it would be something that the core income could support hiring someone to do.

Re:Someone help me with this...

[hazah]

you're more likely to have a community centered around it than anything evil you imagine. - agreed. But this does not mean that writing closed source software is antisocial at all. One can create closed source software and never release it to anyone, not even if they pay for it. Even this is not antisocial. When you create things you don't have to give or show or sell them to anyone, you can just enjoy it by yourself. This is not antisocial, it does not disrupt the norms of society, it is not rude, it is not illegal.

If you're not releasing the software at all, it's neither closed or open, as there is no license attached to it, only a copyright. As you are the sole creator and user, you have both the source and the binary distribution and in no different of the potential position that of any open/closed source vendor on the planet.

One can create closed source software and sell it and this is also not antisocial. After all, this software will help someone to do something, otherwise they wouldn't buy it, would they? It is very social in fact to create any type of software and release it as Free or free or sell it as not Free. It does not fall under the 'antisocial' tag.

Those people that buy closed source software implicitly trust the vendor. If the vendor is very reputable for having perfected their creations (NASA), then you are correct, there is no antisocial behavior going on. However, given the fact that most closed source vendors do not have such a reputation (MS), there is nothing to base this level of trust on. Because of that, it is vital that anyone who wishes to use a piece of software to be able to see the code. Essentially, its a different method to achieve NASA quality in the long run. By no means does that mean that OSS is there, but as long as it remains open, it has a much better chance of success. The release of software is not antisocial, but given the capability of any one of those vendors, their refusal to release code is becoming more and more so.

From dictionary.com:
antisocial:
1. Shunning the society of others; not sociable.
2. Hostile to or disruptive of the established social order; marked by or engaging in behavior that violates accepted mores: gangs engaging in vandalism and other antisocial behavior.
3. Antagonistic toward or disrespectful of others; rude.

So RMS comes up with his own definition of antisocial and now he pushes it onto others and they accept it. But I do not.

No, RMS did not come up with his own definition. He is clearly describing #3. Closed source vendors are engaging in these activities constantly. Some do not, others just not so much. One of the reasons is what I mentioned already with source. Another ongoing debate is software patents and the legal wasteland that it's creating. Oh, and that nice lady from MS "support" saying, "It's too bad that you already have a key, you're going to have to buy another CD now, full price, because we'll never send you another one".

Re:Someone help me with this...

[JavaRob]

One thing that might help you is that "support" doesn't necessarily mean "end-user tech support", [...] it just means that the customer (individual or business) might tell you what the "fresh new problem" is.

This is how I earn my money now, doing contract development. It's okay -- quite heavy on the people skills (requirements gathering is an art!), plus I'm not working on stuff that's particularly interesting to me. Plus -- I get paid by the hour. When I take time off, I don't earn anything. Hence the appeal of developing a closed-source product.

In the context of building extensions and customizations to the main software I developed... well, that gets back into subtle variations on the same problem, not solving new ones.

Re:Someone help me with this...

[JavaRob]

Have you seen nothing in how things are usually done in open source?

For the record, I have started a few small open source projects, and contributed to some larger ones. I have done for-pay development with the agreement to open-source the final result... I get some satisfaction out of this, but not enough cash to feed myself.

First of all, you can sell the software, even if it is opensource. Second, no one is going to take your code, fix it up, and then sell it. That would be quite a large investment on their part.

You lost me on the second point -- assume for a moment that I have spent 3 years developing and refining a production-ready application, which I am currently selling closed-source. Are you seriously saying that the effort involved in taking my code, changing the software title and icons, and reselling it is anywhere *near* the effort I put into developing it? I don't see it.

Instead, what you're more likely to find, is that people are sending you bug reports, fixes, suggestions. If you go the open source route, and you actually DO have very well written software, you're more likely to have a community centered around it than anything evil you imagine.

In the ideal situation of open-sourcing my software, I would end up with a strong community of developers who would help improve the software beyond what I could achieve on my own. Some company better-equipped for distribution and support will package, sell, and support the end-result software, and if I'm lucky they will hire me. If my software design is good enough, they will not need to hire me, and I will be left with a fuzzy warm feeling for having helped the world while I gnaw on a crust of bread.

In the worst case (but very possible) scenario, my wonderful software will target a small enough niche (or won't have that magic "developer itch" factor) that I *won't* have a good community of developers, plus any company with the resources can still package and sell it, and their only real competition will be other companies doing the same thing. I, with no funds to build a dist and support infrastructure, will be a) bankrupt, because without any income I can't pay off the debts from that 3-year dev cycle and b) not even have any warm fuzzy feeling.

I'm far from being anti-OSS (see above), but I still can't see a realistic picture in *all* software being open source.

Re:Someone help me with this...

[JavaRob]

Those people that buy closed source software implicitly trust the vendor. If the vendor is very reputable for having perfected their creations (NASA), then you are correct, there is no antisocial behavior going on. However, given the fact that most closed source vendors do not have such a reputation (MS), there is nothing to base this level of trust on. Because of that, it is vital that anyone who wishes to use a piece of software to be able to see the code. Essentially, its a different method to achieve NASA quality in the long run. By no means does that mean that OSS is there, but as long as it remains open, it has a much better chance of success. The release of software is not antisocial, but given the capability of any one of those vendors, their refusal to release code is becoming more and more so.

You're not really arguing for open source here -- you're arguing for real developer support, which ideally would include source code access but need NOT grant the rights to redistribute. Plenty of developers decompile closed source software while trying to debug it (and been annoyed that this is against the license!) but don't even think of trying to redistribute it (the bigger open source part).

It's tough to argue that offering comprehensive developer support (including viewable but NOT open source) is still being "anti-social" in any way.

Re:Someone help me with this...

[roman_mir]

Those people that buy closed source software implicitly trust the vendor. If the vendor is very reputable for having perfected their creations (NASA), then you are correct, there is no antisocial behavior going on. However, given the fact that most closed source vendors do not have such a reputation (MS), there is nothing to base this level of trust on. - wow, and releasing software in binary form is different from any other vendor selling proprietory electronics or other types of hardware how? If you don't trust the vendor, don't buy the software - it's that easy. It's like if you don't like GPL - don't use it.

If you don't trust the vendor but you want to buy the software you can probably request a source review or even request that the code must be provided as source and to be compiled with the user's compiler after user does the review. This does not mean that the user gets the right to redistribute the source or steal copyright.

No, RMS did not come up with his own definition. He is clearly describing #3 - what, releasing software under a closed source license is antagonistic to others and rude? That's ridiculous! It's a license, you don't have to accept it! You don't have to buy if you don't like the terms, there is certainly nothing rude about it if you accept the license and buy the software.

Who is being antagonistic here? Who is being rude arguing that closed source software is rude?

Example for other opensource projects

[adepali]

Looking at the BK mess, I believe it's time to reconsider a very common yet very flawed stance taken by several FOSS developers: the idea that venerable legacy software does things the "right" way, and it should never evolve and adapt to the modern programming needs.

We've been tolerating software with crippling capabilities and major usability issues for years, accepting or even praising their limitations as part of the FOSS Way of Things. CVS is a prime example, there are dozens more.

Of course this kind of software has been great, and we are all thankful to the people offering it, but it's time to place it in a museum, with a golden memorial plaque, and move on. The OpenOffice team got it right with their 2.0 version, even if it's still in beta: they evolved some cumbersome legacy applications to something sparkling that's a pleasure to work with (please don't bring up their use of Java issue). We should all benefit from their example.

Re:Example for other opensource projects

[Anonymous Coward]

"We've been tolerating software with crippling capabilities and major usability issues for years, accepting or even praising their limitations as part of the FOSS Way of Things. CVS is a prime example, there are dozens more."

Only because they actually were tolerable. If it were such a pain in the ass, well, why it is still widely used? why didn't you yourself, suffering such big pains aleviate them by putting down some fucking code?

Re:Example for other opensource projects

[adepali]

Who says I didn't?
"Why is it still widely used" indeed... Why are Windows used by 96% of users? Why do so many illogical things happen?

What were the alternatives?

[siljeal]

What were the free alternatives at the time Bitkeeper was chosen? Certainly NOT CVS and NOT subversion. The subversion team posted a statement saying quite clearly that Subversion is not a suitable tool for kernel development at this time.

Re:What were the alternatives?

[True Grit]

What were the alternatives

None.

From reading this thread I get the impression a lot of people simply don't realize that Linux kernel development is fundamentally different from other, typical software projects. It needs an extremely distributed, decentralized SCM because it (the Linux kernel) is perhaps the ultimate example of a distributed, decentralized software project we have so far. Linus couldn't scale, so he had to decentralize, but there simply didn't exist anything that could let him hand off control to sub-components while still giving him the power of oversight and the ability to do fine-grained control over changesets. Except Bitkeeper. Linus made his choice out of pragmatic necessity at the time.

The fact that 3 years later, Linus still can't just use an existing SCM, but has to start building a toolset of his own (git) to do what he needs to be able to do, should be a clue to all the people asking "what about subversion?" that what Linus needs is something most of the rest of us DON'T NEED. For most of us, SCMs based on centralized repositories is a perfectly fine solution for our SCM needs, it works even for large (in LOC terms) projects. Linux though is large not only in terms of LOC, its also huge in terms of the number of contributors and maintainers of its various sub-systems.

The fact is, is that what Linus needs is unusual and not needed by most of the rest of the software development world. This explains why the FOSS community hasn't yet answered this need, because this 'need' is still relatively rare. When it comes to source code management, one size does not fit all. Just look at the different posts in this thread about different SCM preferences. I wouldn't be at all surprised to see the top level Linux kernel developers, 5 years from now, using a custom built SCM system running as a high-level layer over the usual suspects like Subversion, CVS, or whatever. The point is, they're needs are so unusual, a custom solution to their problem might very well end up being what they come up with.

What's interesting is what comes after that. How's this for irony: Larry cuts off Linus's use of Bitkeeper because some other OSDL employee just happens to be working on a competitor to Bitkeeper in his spare time. Linus, because of need and desperation, does what he does best, he throws his own considerable intellectual capacity at the problem and starts hacking away at his own solution. He comes up with something interesting, and unlike Larry, Linus's creation is open-source, thus some of the other kernel developers get interested, and they start hacking on Linus's creation. At first, for a year or two, only Linus and a few top-level (ring 0) developers work on and use it. But in time, the thing gets better and better because its baptism under fire is being used to control the high level source code management for one of the most complex and extremely distributed software projects in existence. It reaches critical mass and some other distributed projects start looking at it as a way to handle their needs, other hackers join in from outside the kernel group. The thing snowballs, and in 4-6 years, the very thing Larry is trying to avoid by cutting OSDL and Linus off at the knees, a direct and serious competitor to Bitkeeper, comes into being precisely *because* he cut Linus off at the knees. Larry ends up creating the very thing he believes he's avoiding. :)

Instead of one obscure OSDL developer working part-time on a competitor, he's now managed to get several of OSDL's best talent, starting with Linus himself, working on a Bitkeeper competitor full time to fill the void left by Larry taking Bitkeeper away from them! As the beer commercial says: BRILLIANT! :)

Yes

[Baki]

It is what subversion is doing.

However, the grandparent article didn't really suggest that Linus use SVN, since as you pointed out he won't due to lack of a distributed repository. The statement on the similarity of versioning concept between SVN and the description of what Linus alledgedly is implementing on his own is correct.

that is correct

[Trepidity]

Although I think you will find that the count is substantially higher than two if you count people who tried to use it and gave up, so perhaps some people (such as myself) will find this funny after all.

GNU stands for software freedom, not "open source"

[jbn-o]

It is a mistake to conflate "open source" and GNU.

• GNU started over a decade before (1984) the open source movement (1998).
• GNU was started to make an entire free software operating system.
• The open source movement rejects software freedom to pursue a message they believe businesses want to hear.

As RMS has said in regards to working on GCC, the GNU Compiler Collection:

"Open source advocates do contribute to our community, when they work on free software packages, but our community is older than that movement, and owes its existence to the idealism that movement rejects. It was built by the free software movement, so it is the free software community. If you help us, please keep in mind that what you're helping is the free software movement."

Darcs?

[Flaming Death]

Not sure why anyone hasnt mentioned darcs - being very similar to bitkeeper and arch. And free.. and so on.. I personally use darcs over anything else, since changing from svn.. I wont be going back :-)

www.darcs.org

Re:Darcs?

[swarsron]

Try pulling the kernel tree and you'll understand why darcs is not really an option. I use darcs too and like it but its simply too slow for anything bigger

Re:Darcs?

[Flaming Death]

I must admit, our current project only goes to 300 MB.. but it seems pretty much fine? records, and pushes/pulls only take a few minutes.. no longer than svn did in anycase. A fresh checkout (pull), for the whole repository is definitely not that slow. Id be surprised if say a 600MB kernel repository was much different in speed with darcs.. as with svn.. cvs and such? - I have previously worked on 1GB+ projects (up to 4GB) that were cvs based.. and they were astonishingly horrible for initial checkouts.. certainly much longer than 4x my current project pulls from darcs. I used to let it run overnight (on the 1GB cvs one).. and this is same machine.. with just differing codebases.

If anyone has some profile statistics for these sorts of test/claims Id be very interested. Id hate to get to 600MB and find out darcs falls over?

Re:Darcs?

[Anonymous Coward]

"records, and pushes/pulls only take a few minutes.."

Yeah, well... Mr Torvalds requires these kind of operations spending no more than about 20 seconds, go figure.

Re:Darcs?

[Anonymous Coward]

"If anyone has some profile statistics for these sorts of test/claims Id be very interested. Id hate to get to 600MB and find out darcs falls over?"

No, it won't "fall over". Yours are different requirements. Say it grows painingly slow and goes as long as ten minutes (say twenty) your morning merge. On one hand, you will be doing it twice/thrice a day on moments you probably don't care (first hour in the morning while you go for some coffee, when having dinner...) on the other probably the tool will optimize on time.

Then consider Torvalds' requirements. He'll merge up to 100 times a day (by his own account). Even 1 minute-a-pop makes 100 minutes a day "lost" just waiting for the tool... and that if everything goes OK and there are no conflicts, further editions and rechecks, etc.

Re:Darcs?

[Flaming Death]

I have done upwards of 20-40 changes in our codebase in a day. A merge doesnt take long to do (seconds not minutes). And it shouldnt. Since darcs applies this as a patch (same as BK does). So even on a massive repository, merging should only take a few seconds. Unless hes merging the entire repository everytime, but even then BK aint gonna be quick anyway.

BTW our repository is only 10 weeks old from 0.. so I'd say it grows reasonably quick - even faster than the kernel at the moment ;-) Id estimate the large amount of time in merging is in conflicts (as it is in most repository's). Merging is _so_ not an issue in terms of speed (even CVS and SVN are fine for merging or commits). My original concern was 'pulls' and 'intial checkouts' - if this is done often, then time is not your friend, and even BK is not wonderful on this (good, but not great - so far darcs has show to be as good, but again or repository is smaller than the kernel, so I dont know for certain how it scales.)

Very interesting.

[Anonymous Coward]

So, when people defend PATRIOT by comparing the US with North Korea and China for human rights, that means you are likening the US to NK?

Really not

[paskie]

No, not really. What Linus is doing looks completely different. It is quite similar to Monotone if anything, in fact. It has quite a good description of itself in the README (skip the top part there :-) ).

One consequence of what he is doing is that it is trivial to do e.g. pulling from remote repositories (basically just two rsync commands), or diffing arbitrary two trees. You can see my scripts as an example.

uh...

[Cryptnotic]

If you've gone through high school and college, you should know that it's more important to be popular than right.

GNU tools

[Anonymous Coward]

WERE created from open software. It used to be that the software was open and you got it with your computer. RMS worked on a lot of that, then MIT told him he couldn't look at the code anymore, even when he had written it. So he took the code that was open and created the GNU tools. And worked from there.

So, it DID start with open tools.

C was an open spec

UNIX was open source from AT&T

Re:GNU tools

[SA Stevens]

UNIX was open source from AT&T

If we had a time machine handy, I'd like to introduce you to a room full of AT&T Attorneys.

Re:How Does It Feel Linus?

[Tim C]

The whole point of Open Source is freedom. Freedom from any group being able to "lock you in"

And yet, so many Open Source/Free Software people try to convince everyone that they should only use Free/Open Source Software, thus attempting to lock everyone in to using only that type of software.

If you truly believed in *my* freedom, you would respect my choice to use whatever I believe to be the most appropriate tool for the job.

not surprising though-Another world.

[Anonymous Coward]

Here's something to think about. Presently the thing that's keeping RMS style ideas out of physical enterprises .e.g. creating physical objects is technology. Just think what his ideas would do to the car industry, or the food industry, or any of the myriad industries.

Here's another to think about. Were communism, socialism, or even RMSism's break down is that they assume the ideal, in the world of the imperfect.

All working systems work, exactly because they retain some principles, but don't ignore human nature. If we can chastize the MPAA/RIAA/ETC for ignoring human nature? Why do we think that RMSism can't be?

Someone help me with this...BSD

[Anonymous Coward]

BSD and a couple other licenses accomplish the same thing without all the idealism (and the destructive effects of idealism).

The GPL (and RMSism) aren't the end all and be all of software as a hobby, and an industry..

Re:How Does It Feel Linus?

[m50d]

Because if he thinks using propriety software is OK, why is he bothering writing a free kernel at all?

not surprising though-Economic Terrorist.

[Anonymous Coward]

I think there's one thing that people are assuming here. The GPL license by itself neither,encourages, or discourages economic behaviour from those who use it. The economic aspect comes from the desires of others to assume the burden of work necessary for OSS to meet their needs[3]. And the amount of that is dictated by how much work the programmer put into the results. For example if a programmer creates a free CMS with all the bells, and whistles, complete with excellent documentation, and a wonderful interface. Then the amount of economic work* needed by others is much reduced. So what the GPL does encourage at worst is poor code quality, and at minimumn so so code. So that the economic benefit will come from the programmer doing "that extra economic mile" to support themselves[2]. Now people are going to point out that the code is good, and yes in a lot of cases it is. However that "goodness" isn't so much a function of the GPL aspect [1], but the economic aspect of someone paying them. Also the whole OSS movement is relatively young, and youth has hidden many a problem, that age reveals.

*Keeping in mind a couple facts about economic work. Someone is being payed, be it internal, or external, and two the economic work doesn't have to be done by the original programmer

[1] Programmers already know this. Code quality is highly variable. being free doesn't automatically make it high quality, and there is in fact a reduction in efficiency in dealing with poor code.

[2} Ego also gets mentioned as a viable motivator, but you can't "eat" motivation. You can at best consider it an entrance fee, to the "economic games".

[3] No I haven't mentioned freeloaders, for they're a plague on pay and free alike. But free has the potential be hurt more by freeloading that pay.

rofl

[Anonymous Coward]

In other news linus has switched to charmin toilet paper.

Re:Slashback? Some info on Bit Keeper

[Anonymous Coward]

I think linking to everything2 was more about promoting everything2 than about being informative. After all, it was made by the same people who made slashdot.

Now what?

[harris s newman]

Mod me down if you wish. I use Linux, but prefer BSD's just because of this kind of shit. A critical piece of software being managed by one person. At his whim. When a team develops software, there are more eyes on the DECISIONS and that results in better decisions. This decision by Linus is just an example of the decisions that could be flawed by having only one person owning a project. Now what? This will likely put back the development of Linux.

Why is it news . . .

[Anonymous Coward]

when Linus done stuff? Who really give a shit if he uses Bitkeeper or not. Grow up.

Fascinating.

[Qbertino]

Reading the Kernel Mailing List and Linus and other debating on the issue is just plain fascinating. A few testruns on monotone and 5 OSS experts are giving commments on what could be wrong, why it is so slow and meanwhile Linus has written "a little toolkit" the last two days that could be usefull as a base for a version management framework.
Just watching them work is a pleasure. These people probably do things in an hour for which I would need half a year. If the monotone people (is it a team?) are smart they'll join right now. There are invaluable testresults and accompaning interpretations from kernel developers rolling in on the tool as we speak. In two months from now monotone could finish what otherwise would take a year or two. ... Imagine Linus 'helping out a little for a few days' on your OSS project. Cool.

MOD PARENT UP AS INFORMATIVE

[akc]

Sorry no mod points, but been reading parents author in lkml.

extend subversion

[jonathanduty]

I wonder why he doesn't just extend subversion or the svk extension and add the functionality he wants. After reading up on it, subversion is almost there, just missing a few key features that he really wants.

Re:extend subversion

[Anonymous Coward]

"I wonder why he doesn't just extend subversion..."

Yeah... and I wonder why people just don't extend peebles into beef and hunger disappears from Earth... maybe coz you can't get apples from oranges?

People laid off in US

[lorcha]

They either got new jobs for 50% more money, or opened their own consulting businesses (like me) for waaaay more money.

The ones who still can't find work are people who should never have been let within 500 feet (that's 15,400cm, just to give you a rough idea) of a computer. They're people I like to call "Morons who got into computing because they heard it was a great way to get filty rich quick but are now on welfare and like to bitch about it on slashdot every day".

They are not going to be writing Linus's next SCM system. Their greatest concern right now is whether or not you would like fries with that.

Re:People laid off in US

[SA Stevens]

Actually, 'adapting well to a business environment' (i.e. being all slick and jiffy about finding another job) does not automatically correlate to 'good code hacker.'

Re:People laid off in US

[lorcha]

Actually, 'adapting well to a business environment' (i.e. being all slick and jiffy about finding another job) does not automatically correlate to 'good code hacker.'

True, but being a 'good code hacker', as you say, does more than just correlate to a high demand for your good code hacking skills. It causes a high demand for your skills.

So if you are good, you are still in high demand. If you never should have been here to begin with, then you better start practicing the "fries with that?" routine.

I can tell you firsthand it is damn near impossible to keep my clients supplied with quality people. If you know any, please give me a call.

Re:People laid off in US

[Anonymous Coward]

What the fuck makes you think that you should have been let within 500 feet of a computer? You manage to bullshit people into paying you lots of money, so that means you're competent? Yeah, right.

Re:People laid off in US

[4of12]

They are not going to be writing Linus's next SCM system.

But, if any of those laid off folks were talented enough to write Linus' next SCM system - and I have heard that Linus is a picky customer and that his product is showcased to the most elite nerderati - then they would have marquee billing.

Visual Studio 2005

[Anonymous Coward]

Will have new SS, same what MS uses.

Re:How Does It Feel Linus?

[jacksonj04]

Because he can. That's the whole point!

The proprietary OS didn't do what he wanted, so he wrote a kernel. However, BitKeeper did what he wanted better than the OSS ones, so he used that.

The me too post

[akc]

exactly, but I still don't have mod points to mark it as such:-(

dupe, but anyway.....Forth.

[Anonymous Coward]

You C/C++ people are a hoot. Truely a "Everything looks like a nail" example. I can get a working system in a shorter time than you all can by using Forth. Then once I've got going, I can lay C/C++ idioms on top.

Re:dupe, but anyway.....Forth.

[mrchaotica]

You've still gotta build a Forth interpreter first, unless you're talking about OpenFirmware, in which case it ain't Free either.

performant

[PMuse]

Well, teach me! It actually _is_ a word. . .

performant (n): a performer (Etymology: based on informant, etc.)

Source: Webster's New Millennium(TM) Dictionary of English

Heh.

[lorcha]

How's that job search going, fry-boy?

I disagree entirely :)

[pjc50]

FIRST, build a revision control system based on what you think people want. Collect hate mail and suggestions for improvement. Meanwhile, people work with the imperfect (but existing!) solution.

Then deploy phase 2: what people ACTUALLY want based on experience of the first product.

Phase 3 is a marketing campaign to change people's expectations of revision control systems to match what your software does. This is particularly easy when things like "change" are so ill-defined to start with.

The result is a system that a majority of people are comfortable with and understand.

Someone help me with this...Conundrium.

[Anonymous Coward]

"You're not really arguing for open source here -- you're arguing for real developer support, which ideally would include source code access but need NOT grant the rights to redistribute. Plenty of developers decompile closed source software while trying to debug it (and been annoyed that this is against the license!) but don't even think of trying to redistribute it (the bigger open source part)."

There are a lot of companies that offer source code with their product. Usually with high-dollar items, were not having the source code would have a greater impact than Nvidia not offering it's source code. Most of the benefits without the BitKeeper disadvantage.

Naturally there's an element of trust with all licenses, and if someone decides to violate it by reverse-engineering it, or taking GPL code and incorperating it into hardware. Then that's what the law's for.

Anyone who uses my product must do as I say...

[doom]

It's the double standards about this argument that bother me. If it was not for proprietary software, the GPL movement and the GNU free software directory would not exist. The GNU tools were all initially developed on Solaris and other proprietary operating systems before a kernel even existed

Too bad Sun didn't think to forbid the use of Solaris to anyone who might be thinking about making a competing product, eh?

(But then, maybe it's too bad that BSD wasn't originally under a GPL license, then Solaris wouldn't have existed, the unix-wars of the 80s wouldn't have happened, Microsoft Windows might have died a well deserved death before Windows NT was out... counter-factuals are fun.)

Re:Anyone who uses my product must do as I say...

[bheading]

"Too bad Sun didn't think to forbid the use of Solaris to anyone who might be thinking about making a competing product, eh?"

Under the BK license this restriction only exists with the free version, which seems like a reasonable enough compromise unless you're one of those people who expects people to give up their life work for free. True, it's great when people do that, but not everyone is the same.

"But then, maybe it's too bad that BSD wasn't originally under a GPL license, then Solaris wouldn't have existed,"

But now you're talking crap, since Solaris is based on SVR4, not BSD (SunOS, retroactively known as Solaris 1, was BSD based).

"Microsoft Windows might have died a well deserved death before Windows NT was out... counter-factuals are fun"

They're not fun when they are misinformed in the way that yours are. The rise of MS Windows has little to do with the UNIX situation and it definitely owes very little to free software.

Re:Anyone who uses my product must do as I say...

[doom]

"Too bad Sun didn't think to forbid the use of Solaris to anyone who might be thinking about making a competing product, eh?" Under the BK license this restriction only exists with the free version, which seems like a reasonable enough compromise unless you're one of those people who expects people to give up their life work for free. True, it's great when people do that, but not everyone is the same.

Reasonable? Well I wouldn't have drank that cool-aid myself, even given the claimed factor of 2 (or was it 3) productivity improvement. (And of course, that factor is comparing BK to no BK, what would it have been for BK comapred to another system?)

And myself, I don't actually think it's reasonable for companies to get too creative about the "agreements" they impose on people (Careful about eating that candy bar before getting a lawyer to read the label! Who knows what's in the fine print.)

(And I actually have my doubts that this BK license stuff would actually apply to someone living in California, which is a "right-to-work" state, and tends to shoot down non-competition clauses.)

"But then, maybe it's too bad that BSD wasn't originally under a GPL license, then Solaris wouldn't have existed," But now you're talking crap, since Solaris is based on SVR4, not BSD (SunOS, retroactively known as Solaris 1, was BSD based).

Well you have me there, I did indeed get my the origins of the older SunOS confused with the current Solaris. (Though I could try and skate on a technicality: would Solaris have ever existed without SunOS?)

"Microsoft Windows might have died a well deserved death before Windows NT was out... counter-factuals are fun" They're not fun when they are misinformed in the way that yours are.

Wrongo. Those are the *most* fun ones.

The rise of MS Windows has little to do with the UNIX situation and it definitely owes very little to free software.

Tell me, did you ever get your hands on a copy of Windows 3.11? Do you remember what an incredible, unbelievable piece-of-crap that was? Microsoft didn't have anything close to a decent product before Windows NT. My claim here is that had it not been for the Unix wars, there would have been a stronger competitor to Windows, and they might not have gotten to the "NT" stage.

Re:Anyone who uses my product must do as I say...

[bheading]

Regarding the productivity claim, Torvalds preferred the no-SCM system to CVS. My understanding - stop me if I'm being misleading - is that he seemed to believe that CVS or Subversion would have slowed him down compared to his own system of manually merging the patches; I can't remember the reasons why Arch was not acceptable at the time. On the basis the productivity claim seems valid to me.

UNIX wasn't a competitor for Windows; the two products do two different jobs. The unfortunate way of things now is that a decent sized business probably needs both Windows *and* Unix boxen.

Re:Anyone who uses my product must do as I say...

[doom]

Regarding the productivity claim, Torvalds preferred the no-SCM system to CVS. My understanding - stop me if I'm being misleading - is that he seemed to believe that CVS or Subversion would have slowed him down compared to his own system of manually merging the patches; I can't remember the reasons why Arch was not acceptable at the time. On the basis the productivity claim seems valid to me.

At this point it's all handwaving of course, but my gut level feel is that they could've figured out ways to struggle along with, say, subversion, and still gotten a productivity boost -- though perhaps not as big has been claimed for using BK.

Personally, I have a good vibe about "gnu arch" and Tom Lord in general, but I think it was still pretty new back then. Also the performance data I've seen does show that arch is still peculiarly slow in some areas. Anyway, in general I can understand why Torvalds would've been reluctant to dive into using arch back then (but if he had, I bet arch would be a lot better by now, and we'd all be better off).

UNIX wasn't a competitor for Windows; the two products do two different jobs.

When Windows NT hit the market, I think it ws the Unix Review that ran a cover story: "Is Unix Dead?"

No, Unix wasn't serious competition to Windows at that stage -- that's what I'm complaining about. Why wasn't it?

(Now, thankfully, we're getting to a position where unix boxes really can substitute for Windows for *most* of what people want to do with them.)

linux is starting to sound like a corporation

[Anonymous Coward]

It's been a couple of years since we changed software-- I guess it's time...

Whoa, drag'n'drop programming!

[Jussi K. Kojootti]

I also see the necessity of the command prompt and use it for scripting in addition to applescript ( both of which are soon going to become unnecessary .

Well that was hilarious. You know, Apple isn't the first to try drag'n'drop programming. They'll fail like the ones before. I'm not saying their stuff won't be useful in some strictly defined situations (like the one in the flash clip), but frankly, if you think you can somehow emulate the power of the CLI with a couple of listboxes and radiobuttons you either don't know much about command line tools or are out of your mind.

Re:How Does It Feel Linus?

[malloc]

And yet, so many Open Source/Free Software people try to convince everyone that they should only use Free/Open Source Software, thus attempting to lock everyone in to using only that type of software.

If you truly believed in *my* freedom, you would respect my choice to use whatever I believe to be the most appropriate tool for the job.

Trying to convince someone to use FOSS != "locking in"! If you make a choice to use FOSS you are not at the mercy of the vendor in the way you are with a closed product. See Vendor lock-in

If I "truly believed" in your freedom I would respect your choice to put on handcuffs? If I "truly believed" in your freedom I would respect your choice to sign away your freedom in NDAs? No, of course not. What I'm saying is that while you certainly have the choice to choose what software you want, by going proprietary you ultimatly (long-term) are reducing your own freedom and choice.

Malloc

opensource stubborness

[xpyr]

So the main reason why linus switched was because some buddies of his that worked with him didn't like bitkeeper because it was proprietary. Nevermind the fact that linus himself said it made him twice as productive when he started to use it. I myself don't really care if a program is proprietary or not. If it works for me, I'll use it. These opensource advocates need to get off their high horse and not judge a piece of software based on whether it's proprietary or opensource, but instead judge it based on what you can do with the program. BitKeeper was one of those proprietary programs that worked great for it's use. But because it was proprietary and not opensource, it was disliked by some.