Tridge Releases BitKeeper-Compatible Tool

Peter Willis writes "Looking at Freshmeat today (a part of OSTG) it seems Andrew Tridgell has released the BitKeeper-compatible source code management client mentioned on slashdot recently, called SourcePuller. As part of the downloads available for the project you can also get dump files which detail how to pull data from BK trees without the use of libsp. From the README: 'SourcePuller is not intended to be a full replacement for BitKeeper. Instead, you should use SourcePuller as an interoperability tool for situations where you cannot use bk itself. SourcePuller is missing a large amount of core functionality from BitKeeper, and thus is not suitable as a full replacement.'" Article available about the release on The Register.

Comments

As Tridge says in the README

[Anonymous Coward]

Finally, I would like to point out the obvious fact that Linus was perfectly within his rights to choose bk for the kernel. I personally would not have chosen it, but it was his choice to make, not anyone elses. Linus is now in the unenviable position of changing source code management systems, which is a painful task, particularly when moving away from a system that worked as well as bk did. If you want to help, then help with code not commentary. There have been enough flames over this issue already.

so let's keep it civil, eh?

Re:As Tridge says in the README

[kfg]

so let's keep it civil, eh?

Oh! Civility. I'm afraid we don't have much call for that around these parts. May I interest you in some irony?

KFG

Re:As Tridge says in the README

[qortra]

Nice quote, but the thing is, that's all we CAN do here. Slashdot is not a place to do software development; it isn't set up for that. It's a place for discussion. So basically, discuss, praise the author, try & review the software, but if you want to "help with code", you're in the wrong place.

Re:As Tridge says in the README

[Elwood P Dowd]

You heard it hear first, Tridge is anti-comments. If you want GCC to continue to support comments, speak up now or forever hold your peace.

Re:As Tridge says in the README

[FidelCatsro]

Irony always goes astray here , Perhaps i could intrest you in some apathy.

Re:As Tridge says in the README

[0x461FAB0BD7D2]

From the source code I've seen and debugged, it seems most programmers are anti-comments.

Re:As Tridge says in the README

[Elwood P Dowd]

"heard it hear first"

Jesus God, that's one of the dumbest fuckups I've seen in a /. comment, and it's in my comment.

Re:As Tridge says in the README

[Anonymous Coward]

No call for it? But it's the single most popular cheese in the world!

Re:As Tridge says in the README

[Anonymous Coward]

You're mistaken.

It isn't that programmers are anti-comments. They are rather, comment agnostic.

Re:As Tridge says in the README

[Anonymous Coward]

You can however refrain from posting and contact someone about helping with the code. It's really not a thing I have time for or even interests me, so I doubt I'll be much help. I'm mostly a measly web/database programmer anyway. :)

-- gid

Re:As Tridge says in the README

[killmenow]

Perhaps i could intrest you in some apathy.

Meh.

Re:As Tridge says in the README

[snorklewacker]

> If you want to help, then help with code not commentary

Fact is, this is slashdot. It's the nee plus ultra of peanut galleries. Note well that Linus doesn't read or post to slashdot -- he's been busy working on git, which is not a great SCM itself, but it's turning out to be pretty hot as a library. Arch is adopting git as a backend, and darcs is talking about using libgit (once there is a libgit) for a performance boost.

Re:As Tridge says in the README

[SnowZero]

Sometimes I am anti-comment. When I get some example code that has 10+ lines of comments per line of actual code, I usually delete most of the comments so I can see what the heck the code is actually doing. There is such a thing as too much verbosity, or at the very least, some documentation shouldn't be in the middle of the function.

Re:As Tridge says in the README

[indifferent children]

it seems most programmers are anti-comments

You're kidding right? What do you call all of that whitespace that we sprinkle around our code? Those are comments.

If a block of code is especially self-contained or tricky, then it is surrounded by two carriage-returns before the block and two carriage-returns after the block.

If some statements are part of a loop, then we gratuitously indent them. That's not for the compiler's benefit; those are comments.

Don't even get me started on our extreme generosity in supplying names (not just types!) in our function and method prototypes. What, you want us to draw you a map?



BTW, I would submit more Insightful comments in my code if only my peers with good karma had Mod Points at code review time.

Re:As Tridge says in the README

[Anonymous Coward]

At least it's not

"Ewe herd it here first"

Re:As Tridge says in the README

[Anonymous Coward]

Against Tridges' wishes I am reverse engineering GCCs comments subsystem:

cat code.c | awk '{ gsub("\/\*.*\*\/", ""); print $0}'

Re:As Tridge says in the README

[OrangeTide]

And when they do have comments, the comments are wrong because someone redesigned that bit of code 10 versions back and never bothered to update the comments.

Re:As Tridge says in the README

[Psiren]

Arch is adopting git as a backend...

Really? I have to say I find that quite amazing. How long has arch been around? And git is, what, 3 or so weeks old? Either Linus really is as shit hot a coder as many would have us believe, or the arch guys have been doing something terribly wrong. Is git so fundamentally different from every other SCM out there that no one but Linus saw the need for it? Open source really does move breathtakingly fast sometimes...

Re:As Tridge says in the README

[jhantin]

"Ewe herd it here first"

Oh. That's just baaaaad.

Time to get back on topic, that's enough subversion of this thread.

Re:As Tridge says in the README

[Anonymous Coward]

If some statements are part of a loop, then we gratuitously indent them.

It's only gratuitous if you use something extreme like 8 spaces per indentation level. And on the contrary, some of these space misers only use 2! But most of us graciously use 4 spaces per level of indentation.

Re:As Tridge says in the README

[0x461FAB0BD7D2]

At least you don't get your code-reviewers moderating your code as -1,Redundant and -1,Offtopic. That's almost mean-spirited enough to make me stop submitting "Hello, World" patches.

But I shall persevere. And perhaps, I'll add some CRLFs this time.

Re:As Tridge says in the README

[Anonymous Coward]

Not 'round here, sir.

Re:As Tridge says in the README

[Fnkmaster]

GnU hurd it hear first?

Re:As Tridge says in the README

[Foosinho]

Two "+5 Funny" out of the same post? I bow down to your slash-skills.

Re:As Tridge says in the README

[MikeBabcock]

And some of us indent the code randomly, and counter-intuitively to how the flow works.

Afterall, if you can't count braces, you don't deserve to code.

(that's a joke ... )

Re:As Tridge says in the README

[Anonymous Coward]

Wow. Yet another humor-impared knee-jerk religion-hater mod'ing down any occurance of 'God' that isn't followed by 'damn'.

If you read v..e..r..y.. ..s..l..o..w..l..y, you might see that the post was a humorous response to the guy slamming his own post.

Re:As Tridge says in the README

[RVT]

Why do you think it's called programming Language? Only rarely should it be necessary to 'footnote' your code. It should speak for itself.

Documentation is another thing. It provides context for whatever it is you are expressing. It is always good to have some of that with a function.

Re:As Tridge says in the README

[cfsmp3]

ctrl-t will help.

Re:As Tridge says in the README

[Sircus]

Is git so fundamentally different from every other SCM out there...

No. A lot of git's concepts are taken from monotone. It's basically monotone without the sanity-checking and without the SCM built on top of it.

Re:As Tridge says in the README

[sik0fewl]

BTW, I would submit more Insightful comments in my code if only my peers with good karma had Mod Points at code review time.

Yeah, I would submit more comments in my code, too, but I would be aiming for +5 Funny so it probably wouldn't help anybody understand the code anyway.

Re:As Tridge says in the README

[eh2o]

If some statements are part of a loop, then we gratuitously indent them.

Really? I always thought that was the ghost of RMS reaching out of emacs to strangle me...

Re:As Tridge says in the README

[einhverfr]

Nice quote, but the thing is, that's all we CAN do here. Slashdot is not a place to do software development; it isn't set up for that. It's a place for discussion. So basically, discuss, praise the author, try & review the software, but if you want to "help with code", you're in the wrong place.

Hardly.
1) telnet thunk.org 5000
2) help
3) ????
4) quit

Ok, normally 4 should be Profit, but this actually works....

Re:As Tridge says in the README

[indifferent children]

Yeah, I logged onto /. today thinking, "I'm going to type in a comment that cures cancer". A +5 Funny will have to do :-(

Re:As Tridge says in the README

[einhverfr]

Wow.... I have to say... I wish all programmers commented the way you do.

What I personally hate is the way many programmers often try to tell me how their code works using comments. These are dangerous. Whitespace provides hints as to how to read how it works, but the code has to be your authoritative point for that: SPOT principle at work (comments should not get outdated).

If the code does not speak for itself, it is bad code and should be rewritten.

Now........ When you do comment... Don't tell me HOW your code works! However, it is perfectly reasonable to tell me why you make a specific decision. For example: "Sorry, this code is ugly. However, I can't come up with a better answer at the moment." or "This code here is to prevent an SQL-Injection attack." Or even "As per RFC. Sorry folks, FTP to Mars won't work."

Re:As Tridge says in the README

[Iffy Bonzoolie]

If some statements are part of a loop, then we gratuitously indent them.

It's only gratuitous if you use something extreme like 8 spaces per indentation level.

I think his point was that any indentation is gratuitous to the compiler - it is used for human readability, and thus could be considered a "comment."

But, I think a comment is a certain class of readability measures applied to code. Formatting is another. Naming is another. I personally subscribe to the self-documenting code philosophy - I'd generally rather pull some code out to a seperate method and give it a terse, yet descriptive name than break the flow of the code with some inline comments. Refactoring and all that.

Of course, sometimes I get stuck because I can't think of an acceptable name for my new class or function.

-If

Re:As Tridge says in the README

[fsterman]

Studies have been done that show commenting helps coders write better code, especially when they comment beforehand. It's just like any profession, thinking out what you do before you do it helps you do it better. So while it might be bothersome for you it helps them. Try it.

Re:As Tridge says in the README

[dmaxwell]

If the code does not speak for itself, it is bad code and should be rewritten.

Is that truly universal? What if a really ugly regex is the most efficient way to get something done? What if the code in question implements a not immediately apparant algorithm? As far as algorithms go some rely on some rather hairy math that not everybody has minored in mathematics to understand. In situations like that, the last thing we need are arrogant programmers who think their code is so elegant as to not need comments.

Re:As Tridge says in the README

[walt-sjc]

From "Real Programmers Don't Eat Quiche":

Real Programmers don't comment their code. If it was hard to write, it should be hard to understand and even harder to modify.

Re:As Tridge says in the README

[gunnk]

I'm going to MOSTLY agree with the grandparent, but in the strictest since you are correct to wonder if it is truly universal. However, AS AS GENERAL RULE code that is truly obscure is generally poorly written. Cramming a gazillion manipulations into a single line is not generally better performance-wise than breaking it apart into readable chunks.

Of course, assembly is ALWAYS ugly, but can also be the most efficient code by a long shot...

What was my point?

Re:As Tridge says in the README

[Anonymous Coward]

Precisly! I prefer to write a generous description of what the function/class/code block is supposed to do and then code to that description. I only intersperse comments where it will be helpful to clarify confusing code bits. Otherwise, a good coder should be able to write code that is clear and concise that does not require too many comments to understand.

<sarcasm>
Code should be really clear and concise, like this:

/* This is an integer */
int a;
/* Add them together */
a=zdjj + hh;

</sarcasm>

Re:As Tridge says in the README

[Anonymous Coward]

what you want is an editor that allows folding (as in folding a piece of paper) to hide a block of lines.

Re:As Tridge says in the README

[einhverfr]

Is that truly universal? What if a really ugly regex is the most efficient way to get something done? What if the code in question implements a not immediately apparant algorithm?

What if your really ugly regex is broken? Should I be debugging your comments or your regex?

In this case, I would like a comment which says something about why you chose that way. Something like:

"This is ugly but seems better than the alternative."

This is an immediate flag that this piece of code is difficult to read and deserves additional attention.

As far as algorithms go some rely on some rather hairy math that not everybody has minored in mathematics to understand. In situations like that, the last thing we need are arrogant programmers who think their code is so elegant as to not need comments.

Ok, but who is going to be debugging this math app? Someone with no math background? If so how will that person know if it is working or not? I would say you are in trouble in this case even with the comments you ask for.

So I will rephrase this:

If the code does not speak for itself when read by a qualified programmer, it is bad code and should be rewritten. Also ideally the qualifications required should be as low as reasonably possible.

Re:As Tridge says in the README

[Bloater]

I think he means stuff like:

if(a!=1){ /* if a isn't equal to one */
a++; /* increment a */
x+=a-1; /* add the old value of a onto x */
foo( x); /* foo is a function, call it passing the value of x */
} /* end of if (a!= 1) block */

I have seen code like that from inexperienced coders because their tutor's slides explained what each line does in a comment and they thought that's what you're supposed to do. They also thought that indentation and whitespace was done in the slides because the tutor fancied it.

Note to programming course tutors. *DO* *NOT* explain how code snippets work using comments. *DO* demonstrate *proper* use of comments and explain indentation and other whitespace, and exactly why you thought it was best to indent as you did.

Re:As Tridge says in the README

[dmaxwell]

Exactly. I wasn't advocating that sort of thing at all. I'm thinking more along the lines of tersely documented document headers and some explanation when the developer feels excessive cleverness is called for. Comments don't have to be and shouldn't be a stream of consciousness.

Another situation where they're called for is if working around brokeness elsewhere is necessary. Without knowing the problem being worked around, it won't be clear why apparantly baroque code was used. I just don't accept the "comments are almost never necessary" school of thought. Code often has to be adapted to external realities that won't be apparent from the code itself.

Re:As Tridge says in the README

[SnowZero]

You completely missed the point. Comments are like water, essential for life. That doesn't mean a raging flood is helpful for life.

Re:As Tridge says in the README

[SnowZero]

You are spot on. Large comments at the top of a function don't "break up" the work nearly as much as inline comments. Inline comments should not state the obvious, but merely state the sections identified in the top level comment, and describe the non-obvious(*). I always like the concept that "comments should state why you are doing something, not what you are doing."

(*) This is way too often overlooked. If it takes more than a bit of thinking to figure out why code is written the way it is, you should comment it. In other words, rather than one comment per line, I'd rather see very few comments on obvious lines, and 7 lines of comments for one line of voodoo.

Re:As Tridge says in the README

[SnowZero]

What I want is an editor that folds stupidly obvious comments while keeping the important ones, but even Emacs doesn't do that yet. Folding treats "set i to 0" the same as "FIXME - this is completely broken".

what?

[qortra]

I assume you mean "have essentially nothing on them". If so, why do you say that? I has the source code for the free bitkeeper tool that he designed. What more would people want?

Re:what?

[Anonymous Coward]

I assume you mean "have essentially nothing on them". If so, why do you say that? I has the source code for the free bitkeeper tool that he designed. What more would people want?

This project has a one-line summary, no home page, no documentation, nothing in the CVS. He's basically using SourceForge to host two files.

Re:what?

[GreyPoopon]

What more would people want?

How about categorization and documentation? When you access a SourceForge site with nothing but the source code on it, it's difficult to take seriously. In this case, though, we all know who the author is. For lesser known authors, lack of any additional materials can kill community interest in the project.

Re:what?

[Anonymous Coward]

well then it's a good thing that he's not a lesser known author, huh?

Re:what?

[Anonymous Coward]

He just put it up, there probably isn't a lot of documentation yet. Do most sourceforge projects spring to life fully formed with all their documentation, home pages, bug-lists, etc. all at once?

Might come in handy now

[Anonymous Coward]

with the move away from bitkeeper. :-D

On a serious note, it's good that this apparently oh so evil piece of software is finally out in the open, so that the people can see that all the fuss was about a tool that allows you to get your data that is managed by a propietary tool. How evil...

Re:Might come in handy now

[LiENUS]

Just a note, this tool assists in no way with the move away from bitkeeper, it was only of benefit when the linux kernel was still in bitkeeper. All this lets you do is pull source from a bitkeeper server, it does not replace the bitkeeper server. The way this benefits the open source community is now developers can begin intregrating this library into other development systems (like IDEs) and allowing bitkeeper integration.

Re:Might come in handy now

[Anonymous Coward]

Just a note, it was a joke.

Re:Might come in handy now

[4of12]

On a serious note, it's good that this apparently oh so evil piece of software is finally out in the open, so that the people can see that all the fuss was about a tool that allows you to get your data that is managed by a propietary tool.

In a way it's good to have issues like this bubble up to give people a chance to really think about the fundamental principles involved in open source development.

For instance, it would be hypocritical to applaud Tridge for his work on reverse-engineering SMB to create Samba and to simultaneously criticize him for doing the exact same thing with BK.

Likewise, to criticize MS for using secrecy to make money with its products, while decrying the effect Tridge's reverse engineering has on the income of the Larry trying to feed his family by selling BK.

The principles have to hold independent of the emotional circumstances and the players of the game.

FOSS is all about the natural migration of more and more software technology into commodities. And that will inevitably be a difficult pill to swallow for anyone who has created new software that makes money for them. At some point, if the software is really useful, other competent programmers will look to produce a work-alike functionality. And it will be for the greater good of humanity as a whole because they will be able to use better tools for less money. Even as it erodes the financial benefits that accrue to one or to a few from having thought and done it first.

A happy ending would be Tridge's sofware encouraging more people to buy the BK core to interoperate with Sourcepuller. But, in the longer run, I expect a free core will eventually be developed and it will displace the proprietary one.

Re:Might come in handy now

[John Harrison]

But, in the longer run, I expect a free core will eventually be developed and it will displace the proprietary one.

And that my friends, is true innovation! Let someone else come up with a good idea and then implement a feature identical replacement.

Actually one of the nice things to come out of this is the work Linus has done on git. At least someone is trying to come up with something new.

Re:Might come in handy now

[bcrowell]

Might come in handy now with the move away from bitkeeper.
Wow, that was a close call. I mean, we almost had a situation where, like, anyone off the street could have just used this piece of software to download the entire source code to the Linux kernel. And -- OMG -- this wouldn't have been just the stuff that the kernel developers had decided was OK to release to the public, but the real complete version that they were working on, with all the metadata included. 'Cause, y'know, once all that highly sensitive information was out there in the wild, there would've been no way to get it back and make it secret again.

Re:Might come in handy now

[SlashDread]

"The principles have to hold independent of the emotional circumstances and the players of the game." Spoke the high morale philosofer. Back in the real world, we behave, and make laws accordingly, that are big fat principle breakers, for perfectly valid practical reasons. We can fly a flag. We cannot fly the Nazi flag. Beng. Principle violation. We can speak. We cannot speak hate propaganda. Beng. Murder bad, war good. Beng beng. The principles dont differ, the practical consequenses differ. From some flag we yell and screech. From some plane crash we war. From some hate propaganda we revolutionize. People do the same on every level, we find 10 bucks, we keep it, we find 100 bucks, we go to the police. We will see who is right; if BK goes bankrupt, it was Linus.

Re:Might come in handy now

[ImaLamer]

Well, being open source it won't be hard to fork and then start working on creating the server and "core functionality".

Re:Might come in handy now

[dmaxwell]

And that my friends, is true innovation! Let someone else come up with a good idea and then implement a feature identical replacement.

That is the history of all technology starting from Ogg The Caveman bashing his kill with a stick and Argh seeing it and trying it with a rock.

Bitkeeper itself copies functionality and ideas from older SCMs; it didn't arise in a vacuum. And we all know the one about how Xerox invented the point and click GUI except that they really didn't either.

Re:Might come in handy now

[Anonymous Coward]

Well I don't know about your shitbag country, but I can fly a Nazi flag if I want in mine.

Re:Might come in handy now

[LiENUS]

Creating the server is probably 10x more difficult than creating the client, all the client does is send requests. The server has to handle merges, ACLs and more.

Re:Might come in handy now

[Daniel Phillips]

Just a note, this tool assists in no way with the move away from bitkeeper, it was only of benefit when the linux kernel was still in bitkeeper. All this lets you do is pull source from a bitkeeper server, it does not replace the bitkeeper server. The way this benefits the open source community is now developers can begin intregrating this library into other development systems (like IDEs) and allowing bitkeeper integration.

Your "just a note" makes little sense. The purpose of SourcePuller is to retrieve our open source code, complete with its entire history, into a neutral, lossless dump file format that any version control tool can pull from. BitKeeper integration doesn't even come into it. The purpose of SourcePuller is to release our code from the clutches of BitKeeper so that we can forget about BitKeeper forever and get serious about improving our own tools. As we should have been right from the beginning (and as some of us were anyway, quietly, through the whole distasteful affair).

The way this benefits the open source community is, we don't have to suffer the loss of three years of versioning history. We now move forward without BitKeeper. SourcePuller is not for integrating with BitKeeper, it is for removing BitKeeper from the open source toolchain.

Re:Might come in handy now

[ray-auch]

The version history was never "lost".

What do you think Linus is using to test GIT ?

Notice how he didn't announce he got the data with Tridge's help - why ? - because it wasn't necessary, everything could be exported anyhow. But apparently that wasn't enough for Tridge.

Re:Might come in handy now

[Daniel Phillips]

The version history was never "lost"

According to you. How about: the history, the whole history, and nothing but the history. Concentrate on the "whole" part please.

Re:Might come in handy now

[ray-auch]

Actually accroding to Linus. Like I said, where do think he got his GIT data from ?

If this new GIT database is incomplete then perhaps someone should have waited until the lifeboat (ie. the export) was checked before pushing the self-destruct button.

If the GIT database is complete, then what was the problem (that the reverse engineering was solving) ?

So, is Linus' new dataset incomplete or not ? He doesn't seem to think so (anywhere I've seen) and surely he should know ?

Like Grains of Sand in an Hourglass...

[jeff_schiller]

...so goes the soap opera that has become the Linux community

Re:Like Grains of Sand in an Hourglass...

[Anonymous Coward]

Since when is a soap opera 3 members of the communty :P

Re:Like Grains of Sand in an Hourglass...

[blueskies]

Luckily, the free version isn't supported by ads...

Re:Like Grains of Sand in an Hourglass...

[indifferent children]

Suggested names:

All My Children (a forking-good name)

As the Hello World Turns

One File to Give

General Protection

The <B> and the Beautiful

Re:Like Grains of Sand in an Hourglass...

[jeff_schiller]

LOL....if I had mod points, you'd get some.

Re:Like Grains of Sand in an Hourglass...

[Master of Transhuman]

How about "Linus's Landing"?

"Dynasty" - about UNIX, Linux, BSD, FreeBSD, etc.

Bill Gates = J.R. Ewing? (Naah, Billy Boy hasn't got near the style of J.R.! Forgeddaboutit...)

Re:Like Grains of Sand in an Hourglass...

[stor]

"Hackers", sung to "Neighbours"

Hackers
Everybody needs good hackers
With a little bit of source code
They can code a better way!

Hackers
Should share their code with one another
That's when good hackers
Are truly free...

Cheers
Stor

Re:Why not GNU Arch?

[space_dude_27]

Perhaps because Tridge never set out to create a replacement for BK - merely a tool that would interoperate with it and enable you to get source out of a BK repository without actually running the BK client.

Re:Why not GNU Arch?

[sfraggle]

I must say I haven't used it

Come back when you've tried to read the documentation - you'll quickly see why nobody in their right mind is considering GNU Arch.

So?

[qortra]

Why does that matter? Most of the features in sourceforge are basically designed to be responses to the code/applications anyway (bug reports, forums, etc). So it will get plenty of use. But even if it didn't, why not use it to host a couple of open source files that people want?

Re:So?

[Anonymous Coward]

But even if it didn't, why not use it to host a couple of open source files that people want?

Do they? Now Larry's pulled the Linux BK, what *do* you need this for?

(of course that didn't stop me downloading it anyway :-) )

Re:So?

[Bloater]

All your own revision history that cannot be obtained by any other means (and couldn't be even before BitMover pulled the free BK). Tridge is just giving BK users access to their own data.

Re:Why not GNU Arch?

[Anonymous Coward]

I think you messed up the first link. Where was it supposed to go?

Hell is reporting very cold temperatures today

[Anonymous Coward]

"so let's keep it civil, eh?"

What has the world come to?
An Anonymous Coward making an informative post and asking others to keep it civil?
On /.?

Logo?

[Anonymous Coward]

A big huge middle finger!

Re:Logo?

[Anonymous Coward]

For one who read occasionaly the archives of the linux kernel list [URL], it would be obvious as to who McVoy is. An arrogant troll! The worst part is that he thinks that BK is some big invention. And that's exactly what linus is trying to prove with 'git': it's no big deal Larry.

So why did linus use BK in the first place? In my humble opinion linus *used* McVoy and the BK services. McVoy gave them for free, so why not? For as long as it would last. In the meantime somebody would write a free decentralized SCM.


Tridlell? Linus basically didn't wanna piss off larry because larry as a madman he is could pull off the cables and leave the kernel without source. Linus had to avoid only that. And he's not a troll. But Tridell wanted to piss off larry and advertise his reverse engineering skills (gas station owner really look into these things you know)
More here

Re:Why not GNU Arch?

[0x461FAB0BD7D2]

Flywheels has a history of linking articles at his site, PressEsc to un-related stories and comments, in order to boost his PageRank.

He has even linked to Google to redirect to his page so that he escapes detection from mods.

Feel free to check his history.

Re:Why not GNU Arch?

[grumbel]

### I must say I haven't used it, but from reviews and comparisons I've read, it seems to be a good tool.

Well, try to use it then. The feature that it has indeed sound nice in theory, but Arch has huge problems when it comes to usability and performance, which make it unusable for something as large as Linux and unconfortable for most other projects around. A simple look at the 'help' already makes that pretty clear that there is something wrong with the userinterface:

$ svn help | wc -l
41
$ tla help | wc -l
186

Its however not a lost case, Bazaar-ng is trying to fix those problems of Arch:

* http://bazaar-ng.org/

Re:Why not GNU Arch?

[davecb]

Scalability, actually.

--dave

Re:Why not GNU Arch?

[Anonymous Coward]

Nobody...
Yes, you are the very very very very first first first person to come up with something THAT original.

Re:Why not GNU Arch?

[cduffy]

Arch is a good tool -- once you've wrapped your mind around it. Coming from CVS, that's hard.

One of the problems I'm having at work is that, having wrapped my mind around Arch, I'm for all intends and purposes unable to go back to thinking in CVS primatives -- the conceptual model is that much better. However, since Arch isn't practical for use at my place of employment (no usable win32 port, much less one with a GUI the UI folks can use), I've become damn near useless as SCM advisor -- my mental model just isn't aligned for CVS anymore, and the thought of trying to "fix" that (by retraining myself to work within all of CVS's limitations again) is just too damn horrifying.

In a year and a half, maybe, or however long it is, Bazaar-NG will be ready for commercial use, and then we'll have somethnig that'll let me have my pretty conceptual model and actually be usable by the rest of staff. It's a dream, anyhow.

No thanks

[A nonymous Coward]

Not today, thank you. Perhaps you could come around tomorrow, or maybe the day after? I'll check my calendar, but later, ok?

Re:No thanks

[babyrat]

tomorow is bad for me - I plan to stop procrastinating then.

Wow, that's a lot of code for telnet poking around

[Chyeburashka]

zoopark:~/sp-0.1 gena$ cat `find . -name "*.c"` | wc -l
13416
zoopark:~/sp-0.1 gena$ cat `find . -name "*.h"` | wc -l
859
zoopark:~/sp-0.1 gena$ cat `find . -name "*.[ch]"` | wc -l
14275

Re:Who Gives a Fuck?

[n1ywb]

I think it's pretty clear that most people in the FOSS community were NOT happy with Linus' choice to use BitKeeper. Get a clue.

Re:Why not GNU Arch?

[rob_squared]

You can't even use that, McDonalds is going to bring back the Arch Deluxe and that's its new name.

Re:Why not GNU Arch?

[keesh]

I don't want a tool that tries to tell me how to do things. I want a tool that stays the hell out of my way and lets me do some work. Their "way things should be" is not suitable for all software development models, yet their tool is so closely tied to said model that they actively add in features to try to stop you from working any other way.

WTG

[Tharkban]

Congratulations to Tridge.

Sounds like everyone was within their rights. Which means disagreement is simply the by product of everyone being human.

Re:WTG

[Anonymous Coward]

means disagreement is simply the by product of everyone being human.

I agree with you there.

damn!

Too bad that there is no source to pull..

[Anonymous Coward]

Probably it is more or less pointless now that the kernel source is not in BK anymore.

It almost looks like tilting at windmills, it was given that if Tridge succeeded, the kernel would be out of BK.

Why the P.C name?

[thehunger]

A tool that lets you Pull stuff out of BitKeeper. How did he manage to avoid naming it BitPull?

Re:Why the P.C name?

[Feztaa]

I would have called it Keep on Pullin' Bits

Re:Why the P.C name?

[sethadam1]

Then you would have completely missed the joke...

Re:Why the P.C name?

[Anonymous Coward]

How did he manage to avoid naming it BitPull?

Sorry, but "funny" is meant only for worn-out inside jokes. All forms of cleverness are forbidden. Apologies for the moderation fuck-up.

Re:Why the P.C name?

[hawkeye_82]

Shoulda called it BitFinger

Re:Why the P.C name?

[williw]

As soon as I read your comment, I immediately thought he should have call it PitBull or KeepBitter (get it... reverse engineering reverse name, unintended consequences. har har har)

Re:Wow, that's a lot of code for telnet poking aro

[imroy]

No, the telnet bit was just for dumping the data out of a BK server with the "clone" command/request. I presume that this code interprets the data and produces usable source files and diffs, or whatever format the "metadata" is in. I've never used BK (and after this little exercise, never will) so I don't know what the other metadata is all about.

Re:Why not GNU Arch?

[spworley]

Its however not a lost case, Bazaar-ng is trying to fix those problems of Arch

And with the support of the community, and a lot of developer work, they'll be able to reduce Arch's 'help' text down to only 10 words, making it the most powerful source control system.

Re:Why not GNU Arch?

[cduffy]

Oh, it's not that bad. Yes, Tom has given Arch features which are tied to how he works -- but none come to mind that actually stop you from working a different way, as opposed to merely being annoying.

That said, I'm anxiously awaiting Bazaar-NG.

Re:Wow, that's a lot of code for telnet poking aro

[HeelToe]

Why not just:

wc -l `find . -name "*.[ch]"`

Re:Why not GNU Arch?

[Anonymous Coward]

Maybe the help just sucks? "cvs help" weighs in at 39 lines, but that only includes the top-level commands, and some only exist to introduce more commands (e.g. "cvs admin").

The author of Arch has a reputation for being difficult to work with. That and arch is still a second-class citizen on Windows, which limits its niche somewhat.

Re:Why not GNU Arch?

[cduffy]

Arch is quite scalable, if used correctly -- and the remaining places where it aren't are either (1) implementation rather than design issues, or (2) issues which have a solution proposed which nobody's bothered to implement yet.

Folks who actually do their setup correctly (greedy, non-sparse revlibs; hardlink trees; reiserfs) have reported some very, very impressive benchmarks - and the remaining scalability issues mostly relate to patch log management, and there've been plenty of solutions proposed and on the table that could be implemented very, very quickly if anyone was feeling enough pain to prioritize them (or hire Tom to prioritize them -- same thing, really).

This is what Larry was complaining about?

[X.25]

If *this* is the project Larry was complaining about so much, I can't wait to hear what he has to say now.

Larry, is THIS the reverse engineering you were talking about? Stealing your ideas? Making OSS version of BitKeeper? Blah, blah.

There were so many cases of people making opensource software talking to proprietary back-end (getting stock quotes with tool via TCP, for example, instead of using Java/Windows clients), and noone really made so much noise.

I have no respect anymore for BitKeeper and Larry if this is all Tridge was "reverse engineering".

Re:This is what Larry was complaining about?

[MattW]

And what the hell was Linus talking about? He can understand not wanting someone "riding on their coattails, solve the problem yourself" - what problem? A tool for pulling snapshots? This is not the vast mysteries of SCM being solved, this is a utility.

Re:Why not GNU Arch?

[Anonymous Coward]

$ svn help | wc -l
41
$ tla help | wc -l
186

... and that's to show what exactly? At least grep the commands. The above includes whitespace and section names. There's nothing wrong with more documentation.

However, GNU Arch does have an abundance of commands as you can see with this command

$ cvs help 2>/dev/stdout | grep "^ " | wc -l
32
$ svn help | grep "^ " | wc -l
28
$ tla help | grep ":" | wc -l
105

Re:Why not GNU Arch?

[krmt]

I tried using arch to manage my debian packages, which have an upstreamversion-packageversion versioning number scheme. Both tla and baz complained that this wasn't an appropriate version number. This is beyond annoying, it makes arch unusable for my fairly simple needs.

Plus, the UI is completely tied to the implementation, so you have to know a ton about the underpinnings of arch in order to use tla. I don't want to know how arch does what it does. I don't care.

The baz people are working on fixing this, but there's a lot of problems to be fixed (see this for the massive list) and I think it'll take them some time to do so. Currently, baz is pretty buggy for me too, segfaulting on things like branching. That said, I have a lot of faith in both the baz team and Martin Pool, simply because they've thought things through very well. Currently though, tla and baz are nothing but an exercise in pain for me to use, and bzr isn't ready yet. I'll keep checking on them, because I really want to like them, but they make it so hard on me.

Re:Wow, that's a lot of code for telnet poking aro

[Anonymous Coward]

Why not just:

wc -l `find . -name "*.[ch]"`

Because that'll give the line counts for every C and H file, as well as the total. The poster only wanted the total, so they concatenated the files first.

Re:Wow, that's a lot of code for telnet poking aro

[gdr]

Why not just: wc -l `find . -name "*.[ch]"`

Because this gives you the count for each file individually before the total (which is all you really want) and if find returns no files wc sits there waiting on the standard input.

Re:Why not GNU Arch?

[OrangeTide]

arch is vi unfriendly (you have to edit filenames that start with +), so that perhaps discourages a third of the potential users. Also arch enforces some rules by default that annoy people (like you can't keep any non-repository files in your directory, everything must be under version control at all times. which stinks if you like to keep little notes in your own shorthand). Also the jury is out as to if arch can handle development of a highly active project like Linux. But otherwise I think arch is a neat idea and a neat tool.

Waste of time

[bheading]

But there's already an open-source tool for pulling code out of BitKeeper. So what is the point in Tridge's release?

Re:Waste of time

[Fnkmaster]

Which was only released a month ago... something tells me they released that in response to Tridge's work which demonstrated the demand for such a tool to them.

Re:Waste of time

[Quixote]

Look at the date: 2005/03/17 (or March 17th 2005 for the rest of you). Obviously it was released to undercut Tridge.

Tridge had 2 options after that:
a) release his version (which is also just a client) and get the "is this it? But why did Tridge do that?" comments like the parent; or
b) not release it, and let the rumors fly around.

I'm glad Tridge chose to release his version and now we can all move on. Of course, the real loser in this is McVoy (he lost his biggest mouthpiece), followed by Linus (who has to now duplicate the functionality of BK as much as he can, and while he's mucking with the tools the kernel development takes a backseat).

Re:Waste of time

[BillKaos]

Sorry it's open source but the license is not free neither acceptable:

/*
* Copyright (c) 2005 BitMover, Inc.
*
* Licensed under the NWL - No Whining License.
*
* You may use this, modify this, redistribute this provided you agree to:
* - not whine about it;
* - the fact that there is no warranty of any kind;
* - retain this copyright in full.
*

Re:Waste of time

[Bananenrepublik]

Tridge's tool is about extracting the complete history out of a bitkeeper server, bitkeeper's open client is about providing an equivalent to cvs up. Fairly obvious if you RTFA and TFPYLT (the f*cking page you're linking to).

Re:Waste of time

[murdocj]

I don't recall the exact timeline, but the original article mentioned that the Bk folk and Linus tried to come to some sort of terms that would satisfy Tridge. Given that this tool was made available a month ago, I suspect that it wasn't created to "undercut Tridge" but rather in an attempt to placate him. Apparently that attempt wasn't successful.

Re:Waste of time

[Anonymous Coward]

I thought the free client was available to all, but under the 'look but don't touch' licence.

Anyway, there is little waste here - the OSS community will get a very good SCM tool, that may yet surpass the other ones (that'll be good, won't it?)

Re:Waste of time

[bheading]

Does Tridge's tool extract the code from the metadata or does it merely download the metadata from the server ?

kind of irrelevant now

[ahdeoz]

Isn't it kind of irrelevant now that nobody uses bitkeeper anymore and it will be broken in the next release that all paying customers will be forced to upgrade to?

Re:Wow, that's a lot of code for telnet poking aro

[Chyeburashka]

Because that provides more than the summary I wanted:

[gena@zoopark sp-0.1]$ wc -l `find . -name "*.[ch]"`
263 ./sp.c
128 ./inventory.c
94 ./releases.c
125 ./checksum.c
[ snipped similar output ]
151 ./annotate.c
119 ./getrev.c
14275 total

I suppose I could have done this:

[gena@zoopark sp-0.1]$ wc -l `find . -name "*.[ch]"` | tail -n 1
14275 total

But that 'total' there is superfluous, so:

[gena@zoopark sp-0.1]$ wc -l `find . -name "*.[ch]"` | tail -n 1 | awk '{print $1}'
14275

Thanks for the help. ;)

Re:Wow, that's a lot of code for telnet poking aro

[Wdomburg]

If I were you, I'd have done find . -name \*.c | xargs cat | wc -l. That'd avoid getting back "Argument list too long" if there'd been more files. :)

Summary of the this anywhere?

[nolife]

Anyone have a link to a decent summary of this entire story as it was played out? I've browsed through the past /. articles and searched Goolge but nothing I've found gives a decent summary or big picture history.

Re:Summary of the this anywhere?

[cranos]

I have a brief Summary. Have a look if you like.

Re:Why not GNU Arch?

[SnowZero]

That's why I use darcs. Find out the differences with some surfing.

Shoulda named it...

[Cyn]

BitTaker

Or, for a more exacting description of what their relationship is... rename both tools:

BitKeeper -> BitPitcher
SourcePuller -> BitCatcher

Re:Shoulda named it...

[jwymanm]

I'd just save some time and name all of the tools BitCh

Re:Wow, that's a lot of code for telnet poking aro

[Chyeburashka]

Excellent point. That is actually necessary with a sizable project, like the kernel.

[gena@zoopark linux-2.6.12-rc3]$ find . -name "*.[chS]" | xargs cat | wc -l
6250930

Wow, that's a lot of code for a project that started out with:

Do you pine for the nice days of minix-1.1, when men were men and wrote their own device drivers? Are you without a nice project and just dying to cut your teeth on a OS you can try to modify for your needs? Are you finding it frustrating when everything works on minix? No more all- nighters to get a nifty program working? Then this post might be just for you :-)

Re:Wow, that's a lot of code for telnet poking aro

[Anonymous Coward]

Because that may not work for very large numbers of files, Mrs Smartypants.

Re:Why not GNU Arch?

[skoch]

Interesting way of measuring complexity.

$ diff --help | wc -l
81

Does that make svn twice as simple? Does it not have a good diff built in?

Probably not

[TheConfusedOne]

Unlike Samba, SourcePuller is completely useless without the proprietary product. So, you need to violate a commercial license in order to use it at all.

Samba provides both client and server capability (and PDC and AD functionality now too) for non-Windows OS's.

There really is no BK core and BK client like there is in the CIFS world. In order to replace the "BK core" someone will have to write from the ground-up a complete distributed SCM system. If that effort is undertaken then I really see no benefit whatsoever to having a client that speaks BK's client commands and writes files in their format? (Aside from potential IP violations in duplicating their format.)

So, honestly, what does something like SourcePuller get you aside from losing use of a tool?

Re:Probably not

[Master of Transhuman]

Ah, the first idiot posts...

First of all, as I understand it, the commercial license involved says you will not reverse engineer the product YOU licensed. SourcePuller doesn't do that if YOU use it with some other SCM - it simply interoperates. It may do that if Tridge wrote it that way, which he didn't since he wasn't a licensee in the first place.

Secondly, using SourcePuller does not "get you aside from losing a tool". That statement is just idiotic. SourcePuller gets you exactly what Tridge says - interoperability with BK.

If BK is replaced as the dominant SCM, SourcePuller will be obsolete only when no one uses BK at all. When no one uses BK at all because there is something better, what is the problem with "losing a tool"? The statement is nonsense.

As for Samba, does anybody believe Samba would exist in the form it does if Windows didn't exist? And referencing Samba's AD support as being independent of Windows is just laughable. Samba entire existence is due to the need (however temporary) for Windows interoperability.

Some of the posts on /. are just unbelievable...

Re:Probably not

[Moofie]

"Unlike Samba, SourcePuller is completely useless without the proprietary product."

Samba is useful without proprietary products, but it's also redundant. There are a number of other ways to scratch the same itch that Samba scratches. The reason to use Samba is to interoperate with Windows.

Your argument is silly.

Re:Probably not

[Anonymous Coward]

So, honestly, what does something like SourcePuller get you aside from losing use of a tool?

Apparently a better tool.

Re:Wow, that's a lot of code for telnet poking aro

[HeelToe]

Yeah, I guess I wasn't considering you were only interested in that summary - of course, the cat is easier in that circumstance.

I always find it interesting how different people use the text utilities on files in the shell. It often shows different thoughts and approaches to "set programming."

The REAL issue was overlooked

[Anonymous Coward]

Why is that NO ONE seems to point out that there was a really simple solution to Linus' problem. He could have bought a freakin' commercial license to BitKeeper. He got pissy because he couldn't keep using it for free, but everyone seems to have overlooked that he had the option to pay for it, he just didn't like that option. Maybe the real problem here is just that Linus is a cheapskate. I have to give Linus credit. He totally managed to deflect the issue from his unwillingness to pay to use the tool to make it about how evil Tridge was. Personally, I've got better things to do with my time than waste a month trying to get some new tool to halfway do what a tool I was using but don't want to pay for could do. I would just buy the license and get on with my life. Then again, I'm not a "cut off my nose to spite my face" kind of guy, but it seems that Linus is.

Re:Wow, that's a lot of code for telnet poking aro

[Anonymous Coward]

It's broken for filenames that contain spaces.

Re:Wow, that's a lot of code for telnet poking aro

[Anonymous Coward]

Then again, so is the original.

Re:ATTENTION: THOSE CONSIDERING LINUX AS A MARKET

[Master of Transhuman]

ATTENTION: THOSE CONSIDERING WINDOWS AS A MARKET

This is what always happens. Bring your product to Microsoft and they love you, then they kill you.

Lesson: don't bother.

why not LicensePuller?

[Pollardito]

more people would know what it's describing :)

Yay for OSS!

[black mariah]

We have yet another half-assed implementation of proprietary software for douchebags to mentally masturbate over.

Used already?

[jefu]

Isn't that Wendy's new Logo?

Re:Used already?

[Anonymous Coward]

Isn't that Wendy's new Logo?

Their middle finger isn't attached to the rest of the hand, though.

The original UNIX code

[btarval]

Heh. This reminds me of some of the original UNIX code, from the V6/V7 days.

The startup code was all done in DEC assembly (of course), in a file called locore.s (or something like that - it's been years since I've seen it). Either Brian Kernigham or Dennis Ritchie must have written it; I think it was the former.

Utterly obtuse assembly code, unless perhaps you've spent a good deal of time programming in DEC assembly. Then, no doubt, it was quite clear.

Anyway, about halfway down (after about 50 lines or so), was this priceless comment:
"Here's the tricky part"

Heh. Thanks guys.

sccs

[Anonymous Coward]

the files pulled from bk seem to be in sccs
format, one of the oldest rcs. In this light linus remarks about ip and stuff seem really strange. Besides that, Linus's argument that Trigdes tool is useless are void.

Insightful

[Anonymous Coward]

Linus was just keeping the boat afloat. The Linux and FOSS communities are sailing around on the oceans with very powerful allies and foes all around us. IBM has turned out to be one of our strongest allies, by providing us protection. The War on Intellectual Property is coming soon.

By doing what he did, Tridge is further defining the ground rules that the *real* hackers agree too. As has already been said, he did the reverse engineering on his own time, without ever having used the BitKeeper program. Lessons Learned: Reverse engineering on your own time is absolutely an ethical and allowed activity.

The more I think about it, the biggest winner in all this is the GPL license. It will now have a custom code management application built around working with GPLed code.

*BZZZZZZT* Wrong!

[Anonymous Coward]

Nope, Larry McVoy refuses to sell licences to OSDL-employees, since they didn't do anything against Tridge. So no, Linus can't buy a licence unless he quits OSDL. Check your facts, idiot.

Re:Wow, that's a lot of code for telnet poking aro

[einhverfr]

wc -l `find . -name "*.[ch]"` | tail -n 1

How many ways can you skin a cat in UNIX anyway?

Re:thanks spacktard!

[Anonymous Coward]

The page cannot be found
The page you are looking for might have been removed, had its name changed, or is temporarily unavailable.

Please try the following:

* Make sure that the Web site address displayed in the address bar of your browser is spelled and formatted correctly.
* If you reached this page by clicking a link, contact the Web site administrator to alert them that the link is incorrectly formatted.
* Click the Back button to try another link.

HTTP Error 404 - File or directory not found.
Internet Information Services (IIS)

Technical Information (for support personnel)

* Go to Microsoft Product Support Services and perform a title search for the words HTTP and 404.
* Open IIS Help, which is accessible in IIS Manager (inetmgr), and search for topics titled Web Site Setup, Common Administrative Tasks, and About Custom Error Messages.

Tough Love

[Anonymous Coward]

This will be a tough lesson to the software industry not to do business with Linux, and with Linus.

Work with Linus, have your software eventually copied and stolen by FOS pirate/programmers. Yarg!

I would like a t-shirt with the source a la DECSS.

[Anonymous Coward]

$ telnet thunk.org 5000
help
$ echo clone | nc thunk.org 5000 > e2fsprogs.dat

I'm even more generous

[talaphid]

You can tell how early in some revision of the program a variable first appeared by which letter in the alphabet it is. That is, my first variable is "a", my second is "b", and when I run out of letters, I just repeat incrementing a prepended letter, that is, "aa", "ab"... "ba".

Huh? You mean I can have a variable named ResultsTable? No, no.. "c" is clearly a better name. After all, you wouldn't know where to find the first reference otherwise! (hint: after "b")

Re:Why not GNU Arch?

[cduffy]

This is beyond annoying, it makes arch unusable for my fairly simple needs.

Unless, that is, you make the version part of your branch name.

Re:Why not GNU Arch?

[pHDNgell]

arch is vi unfriendly (you have to edit filenames that start with +), so that perhaps discourages a third of the potential users.

This is not entirely true. I use vi and arch and everything's cool. ``vi ./+blah'' I don't try to edit those directly anyway, I just have an alias that looks like this:

vi `tla make-log`

Also arch enforces some rules by default that annoy people (like you can't keep any non-repository files in your directory, everything must be under version control at all times. which stinks if you like to keep little notes in your own shorthand).

It's perfectly happy ignoring my build directories and other random crap I use for testing. I just need to tell it what kinds of things it should ignore.

More importantly, if something looks like source, I want my revision control system to let me know if I'm forgetting it on a checkin (or worse, if someone else is forgetting it on their checkin and I sync and attempt a build only to find that they're gone and one little piece they added is missing).

I'm sorry to disappoint

[TheConfusedOne]

Not really though.

Referencing Samba's AD support is actually a big point. The point is that Samba can replace an AD system, just like you could use a box running Samba to be a PDC before.

SourcePuller can do no such thing. SourcePuller's one and only function is to pull stuff out of BitKeeper. There's no interoperability. What SCM software can SourcePuller put that data into? I know, let's write a CVS plug-in for it. Oh wait, BitMover already provided that!

Re:I'm sorry to disappoint

[Anonymous Coward]

Referencing Samba's AD support is actually a big point. The point is that Samba can replace an AD system, just like you could use a box running Samba to be a PDC before.

I believe Master of Transhuman's point is that the AD feature of Samba is useless unless you have a (proprietary) MS Windows network. This was in response to TheConfusedOne's point that, "Unlike Samba, SourcePuller is completely useless without the proprietary product."

Re:I'm sorry to disappoint

[keMen]

You seem to miss the point of interoperability. Pulling data out of a proprietary system into a format that can be used by any other system is the foundation for interoperability.

Re:I'm sorry to disappoint

[Master of Transhuman]

Your handle of TheConfusedOne is appropriate as the other respondents have demonstrated.

Think a second

[TheConfusedOne]

One of the more useful functions of Samba where I work is to allow Windows boxes to access Unix file systems.

The point is that Samba acts as both a client and a server. It contains the complete functionality.

SourcePuller only works to get data out of BitKeeper. It doesn't do anything to put data into another SCM system. It can't replace the BK repository's functionality. It basically becomes useless without someone running a BK repository.

Paranoic

[QuestionsNotAnswers]

The journalism on this issue has me wondering if there is not an evil company pushing PR splinters into Linus or Linux from behind the scenes...

A couple of articles appear to have gone a long way towards blowing up an issue - with the open source community being dragged into a non-existent 'fight'!

Thank you /.ers for not joining in and posting days-of-our-lives comments.

Re:Paranoic

[child_of_mercy]

You should have seen the double standing ovation Eben Moglen got at linux.conf.au this morning.

He's planning on making badges saying "STALLMAN WAS RIGHT"

FSF has the high ground in the community now, linus has blown himself out of the water.

I don't get it

[Anonymous Coward]

When Apple releases some new plastics and pays Slashdot PR to trumpet it, every Apple Fan Boy comes out to give thanks and praises. And company never get reprimanded for foul tactics. Ever. It's like people pretend that Apple writes Open Source Software.

When some other propriety software outfit comes out to play, it's "Up Yours".

Can somebody say "double standards"?

Linus = Denethor

[child_of_mercy]

Stared too long into the abyss, gave way to despair, betrayed his realm.

Lunckily Gandalf/Stallman is going to install a new king for us...

Close but not quite

[TheConfusedOne]

The point is that Samba can act as either Server or Client. SourcePuller can only ever be a client.

Samba actually provides a migration path away from the MS Windows network by removing the dependency on MS software to provide server functionality. Think, you could replace all file and print servers on a network using Linux and Samba and not affect a single Windows client running on the network.

This may seem like a minor distinction to you but to me it is the crux of the matter and jibes with Linus' complaint about SourcePuller. It's a tool that will only function as long as BitKeeper is being used. The irony is that the use of the tool almost guaranteed that BitKeeper wouldn't be used.