Hans Reiser Speaks Freely About Free Software Development

Okay, here are Hans Reiser's answers to your questions about ReiserFS, starting and managing (and publicizing) a free software project, earning a living writing free software, and the good and bad points of being considered somewhat of a curmudgeon. As a free bonus, Hans adds a little insight into the politics of Linux kernel development, as in what gets accepted and what doesn't. Good stuff!

1) Guideposts? - by TopShelf

Having obtained financing for the project, how does that impact the future direction of development? How do you balance the interests of developers, users and sponsors to choose which updates to pursue?

Hans:

There will always be more features that I would like to implement than I can implement. Users and sponsors for the most part want good things to be added, and it is really not so bad to first implement the features that someone is willing to pay for, and hope that in a few years someone else will pay for other features. I take a 30 year perspective on the project, and I have as my final objective the elimination of reasons to not use the filesystem as the unifying namespace of the operating system. That makes for quite a lot of features that I am willing to add.;-)

There is an common situation though where pressure from sponsors is severely negative, and that is when they want a quick hack that meets their needs but lacks elegance. It is not usually the features they want that are wrong, it is the timeframe they want them in and the shortcuts they expect to be made to meet that timeframe. This happened with ACLs and extended attributes for instance. I turned away 4 different sponsors for ACLs before I was lucky enough to find DARPA.

All of the commercial sponsors wanted some quick hack that would not be consistent with the semantics I am evolving ReiserFS towards, and would leave us with unwanted additional primitives. To DARPA I proposed that filesystem designers were not providing security researchers with the right infrastructure that they needed, and this lack of the right infrastructure was leading to inelegant hacks. For instance, I argued that there was no need for extended attributes, that instead there was a need for *files* that

1. were efficiently accessed many at a time (new system call that reads and writes many files in one call)

2. were accessed atomically (transactions infrastructure)

3. had constraints on their allowed values not just who can write to them

4. inherit some contents/metadata from other files/directories (note that streams share a common metadata)

5. could be made invisible to readdir

6. could be both files and directories depending on whether you accessed them as files or directories

7. could be implemented via a rich plugin infrastructure that allowed one to compose new plugins by selecting methods from already existing plugins, and only writing from scratch that which was truly unique to the plugin

8. efficient storage of small files (V3 was suffering a performance loss when tails were turned on (that is, when small files are packed several to a block) that V4 cures)

You can probably imagine small appliance vendors not being willing to wait 18 months and spend lots of money when all they want is for ACLs to work well enough that they can sell a samba server, yes?

With DARPA, if you advance the field of security research by developing along an angle the proposal reviewers thought was interesting, it is enough to get the funding. It is easy to think that private industry is sufficiently well motivated to fund long-term research, but unfortunately long-term research tends to benefit a lot more than just its creators and funders, and only the government funds work whose benefits will be mostly diffused throughout society. Or at least that is one theory to explain what happens. The observed reality is that venture capital does not fund long term research, persons wanting to do long term research mostly must pursue government funding, and the readers are encouraged to suggest other explanations of that if they have them.

I would be curious to know if privately held companies without plans to go public (such as Namesys) generally tend towards more long term research. Surely society needs long term research.;-)

DARPA is really quite excellent to work for. Having the right customer is very important to the development of a product, and I learned an enormous amount about serious approaches to security that I would never have learned without participating in their environment. Their accounting requirements are exacting, but it all has a reason, and for each requirement you can easily imagine the taxpayer being abused by someone without it. Actually I was a bit reassured as a taxpayer about how the DoD spends our money by seeing them in action, and being a small company we learned a lot about generally good professional accounting practices by adopting their requirements.

2) Good business planning - by mao che minh

Did you embark on this project in hopes of making a profitable business? It certainly seems that way, considering that you went looking for sponsorship and even planned pay-per-incident support, showing that you were prepared to work the whole "support revenue" angle.

Now you just need to hire someone to desire a modern, more "commercially pleasing" website. =)

Hans:

We make enough from pay per incident at www.namesys.com to support 1/4 of a programmer, and monitoring all requests to ensure that a professional response is always given consumes some of my time. This revenue is steadily growing though, but since we make so little money from it we don't bother with trying to charge a lot ($25 currently), instead we just hope that word of mouth will make it continue to grow, and maybe in a few years it will become something more significant. I think we were key to deflating some of the excessive per incident support rates that were out there for Linux, and this is good, because who can afford $250 to be told how to make Xwindows work?

Where I had originally planned to make money was by using Linux as a market sampling methodology while selling to OS vendors. That a bunch of hackers who had written software I used myself would get to use it for free was ok with me, and a product that was used by the OS vendor's engineers at home would be an easier sell I thought.

Unfortunately, paying money just to catch up to Linux does not seem to excite OS vendors, even if in reality it is very important that they do so. It is theoretically irrational but reality real that the proprietary OS vendors have not yet bought a license.

Instead, I have sold to storage appliance startups who need a code base to start from, and this has been slightly more than half of our income.

We have only sold one priority support contract (this is where you have the right to wake us up and you pay in proportion to the amount of hardware you have), but we sold it to Lycos, and they keep doubling their usage of ReiserFS every year and increasing the contract accordingly. They have been very happy with the support we provided. This one contract is much larger than all our pay per incident fees combined, and thanks much to Lycos for being a customer.

It would be nice if I could sell some government a nice ReiserFS support contract....

Many users don't realize it, but we don't make a lot of money here at Namesys, the programmers are very much overdue for pay raises, and there are many good people I can't hire because we have no funds for it. Hopefully this will change though as ReiserFS increases its technical lead over other filesystems with time. Reiser4 is going to give us some very compelling performance and security advantages, and we will be the easiest filesystem to hack on thanks to the plugin infrastructure. That infrastructure is really going to accellerate development by a lot, and provide us with a compelling competitive advantage. It is interesting to watch Nikita casually toss together a couple of plugins in an afternoon when it suits his whimsy. We can already see how much easier it is going to make doing the semantic enhancements we have planned, and then once we have the semantic enhancements out there and in use, performance will no longer be the primary decision point for users.

As for the website, by "commercially pleasing" I assume you mean using bland and uninteresting graphics, and corporately styled, with lots of insincerity everywhere.;-) Forgive me if I read too much into your comment, but you would not be alone if I read you correctly.

It is important to be true to oneself. You should maybe understand that some years ago I put the suit my mother bought me into the fireplace along with all my ties. If a restaurant requires a suit and tie, I just don't go there, it is not for the likes of me. If you need some corporately commercial justification for our website design, then let it be that I am willing to be cool and appealing to the younger generation, etc., instead of bland. If a pedagogical justification will work for you, then let me point out that military manuals with their cartoon based approach are far more effective in engaging the reader than the pedagogical techniques employed by most college textbooks. The military is more advanced in its pedagogical technique than the university system, which is really rather amusing, and I think it is due to the greater pretentiousness of universities in this matter.

3) Versioning - by tjansen

Beside the finding and organizing files, the biggest problem for desktop users today is probably that changes on the file system are not recoverable. It is easy to accidentally overwrite a file and lose your work, and the only only sane way to solve these kinds of problems would be to make it possible to revert changes.

Several research systems have been created, like the Elephant File System, but none of them made it into the mainstream free and commercial operating systems. Are there any specific reasons why nobody offers recovery (high complexity in implementation, very bad effect on performance, etc) or is it just because FS designers don't see the need for it?

Hans:

Actually, there is a version control filesystem called Clearcase that costs thousands of dollars per seat. (If I wanted to make more money, I could return to working as a Clearcase sysadmin --- there are some jobs that pay very well because nobody wants to do them.;-) ) Clearcase was written a bit too quickly, and its performance sucks as a result (though I am told that has improved since I left that field). If it used Reiser4 as a backend it would be a lot quicker.

Version control definitely belongs in the filesystem. Clearcase may have a lot of implementation uglies, but as a concept it clearly works. Filesystems manage files, and that should include managing file versions. Our support for transactions and compression should make it easier to implement version control in Reiser4. As soon as someone offers to sponsor it, we will do it.

Larry McVoy makes a lot of arguments about how the economics of version control means it has to be expensive. I think this would not be true if three things changed simultaneously: 1) it was integrated into the filesystem, 2) it was free, 3) it was easy to understand for average users. For 3) to be true it has to become as easy as, say, .snapshot directories on a NetApp are, for the average user. It should also be as well integrated into apps as version control is in MS Word. I predict that in 20 years version control in filesystems will be standard and expected by all users as a basic feature.

4) Filesystems and metadata by androse

In your Future Vision white paper, last modified in January 2001, you outline several very interesting ideas about metadata.

Several developements have taken place since; the extensible attributes of BeFS has been buried with BeOS, the database-like metadata of Longhorn (aka Yukon) may actually be a separate layer from the filesystem altogether, and Apple is also moving all metadata out of the filesystem to XML files shared between applications (see iLife package).

My question: What is your current take on the metadata debate? Do you still think the filesystem is the right place to handle metadata? Any predictions?

Hans:

Reiser4 required some fundamental breakthroughs in tree balancing technology before small files could be combined into one block without adding additional seeks for typical usage patterns. In particular it required discarding the BLOB paradigm, recognizing that BLOBs unbalance the tree, and creating a new more height balanced tree. (See www.namesys.com/v4/v4.html) It does not surprise me that the other filesystems failed to find these techniques; until we had benchmarks no one but me on the team thought this new stuff would work. I think you can reasonably assume that MS abandoned its efforts to put the database into the filesystem because their algorithms failed to deliver good performance. Without these techniques, extensible metadata causes performance problems that constitute a market entry barrier.

I should be careful in my phrasing here: Reiser4 does not support attributes, it merely supports files that you can choose to use for metadata, and its files have all the functionality you need for doing the usual metadata tasks should you choose to use them that way.

Oh, and BeFS is probably not all that dead, Dominic Giampaolo the author is working for Apple now, and since he is very bright and capable there are likely to be interesting things coming out of Apple in the future (probably not called BeFS though).

5) Researching filesystems - by ProteusQ

I'm going back to school this fall, and in a year I hope to be admitted into a Masters of Computer Science program. I'd like my main research focus to be on filesystems.

I'm preparing by reading everything I can find: I'm working on Tanenbaum & Woodhull's "OS Design & Implementation"; I've read "Design and Implementation of the Second Extended Filesystem"; Steve Pate's "UNIX Filesystems" is waiting on my shelf; and of course, there's the FAQ and ReiserFS v.3 Whitepaper at www.namesys.com [namesys.com]. Specific questions: what branches of math are useful in this line of research? Any books, articles, etc., that I haven't listed that are a 'must read' or 'should read'? Those who have succeeded in building a better filesystem: what have they done that I should also do? Any mistakes I should avoid? Anything that no one told you about filesystems that you wish you had known up front? And are there any special tricks (above and beyond mastering your subject) to getting hired in this field once a degree is in hand?

Hans:

I was never able to get hired in this field, so I am probably not the one to ask about how to get hired.;-) Hmmm. Oh I know one! Don't tell your potential employer that you are working on your own file system nights and weekends, and you will retain all rights to it, and you won't stop work on it once they hire you.;-)

You should probably read about Plan 9, and about namespaces generally. The literature on namespaces seems to be just about hierarchical namespaces, but the notion present in that literature that they should be unified is a good one. I rather liked Gerard Salton's book on automatic text processing. Ted Nelson's Xanadu project was interesting reading, and you'll want to read Codd and Date about databases. Mikhail Gilula's book about set theoretic databases is a good one.

In regards to math, study the design of new mathematical models. Study closure, and its importance to various models ranging from algebra to relational algebra. Understand why mathematical models were designed to have the structure they have rather than learning what those structures are, so that you can learn to construct your own models. I don't know of any courses that teach that, but it is what is important to learn.

Are you sure that it wouldn't be better to hang out in cafes and bookstores for 4 years, and at the end of it write some piece of a filesystem? Cafes, bookstores, and attending random seminars will educate you better, and writing some piece of a filesystem will employ you better.

If only one could get student loans for the purpose of hanging out in cafes and bookstores for 4 years I would have been so happy....

6) where next? - by wfmcwalter

Hans,

Reiser FS is already a pretty mature, stable, usable product. Once V4 is done, is there really much work left to be done on ReiserFS proper? Do you have a giant to-do list that'll keep you and the guys occupied for years, or do you intent to work in a diffent direction (SAN, networkFS, databases, etc.)?

(or perhaps you'll just retire to Portugal and play lots and lots of golf)

Hans:

V4 is a local host storage layer. V5 will make it a distributed storage layer. V6 will enhance to the semantics to where one can do semi-structured data queries. Whether V5 or V6 comes first depends on funding.;-)

A new model for doing semi-structured data queries was my original goal, and it remains my primary goal, and the storage layer was just a necessary prerequisite to getting there. I describe the enhanced semantics I plan at www.namesys.com/whitepaper.html.

7) Starting Large Free Software Projects - by unsinged int

When you began a file system project as a free software project, you must have known that (assuming it worked) it had the potential to turn into a big project. How did you determine how long to work on it as your own project before making the first release? I imagine there must have been a strong temptation to just get it "out there" knowing its potential, yet certainly releasing too soon would make it look unprofessional and thrown together.

Hans:

When something becomes stable enough for you to use it yourself without it crashing, and you don't know how to make it crash, you should release it (with lots of warnings). Fortunately, there are people who like to play with technology, and they will help you find its bugs while understanding that it is still experimental code. Each order of magnitude increase in the number of users will find as many bugs as the previous order of magnitude. After some number of years, if you are kind to your users and only do new features on a new branch, the stable branch will get to where months go by and there are no bug reports at all even though there are millions of users. It was this way with V3, and it will soon start to happen with V4.

8) Raising Awareness - by blinder

One question I always have with regards to successful (meaning funded, wide acceptance, large user/developer community etc.) is how did you raise the awareness of your project to get it from just a side project to something that it is today?

Did you use traditional PR techniques, or just through a community of connections?

Hans:

There are two things that work, and not much else does:

1) Word of mouth from users brings people from seeming nowhere.

2) Being open and eager to make a bit of effort to work with people makes you friends or at least allies.

Heinz wasn't able to get the time of day from the ext2 team, and he needed a resizer for LVM to happen. We said yes, sure, we'd be happy to do it. He introduced us to SuSE, and SuSE paid us $5k to write a resizer. That led to SuSE becoming a sponsor, and then once we were stable they made reiserfs the default filesystem.

Word of mouth from the users is the most powerful tool free software has. If you combine that with being open, willing to work with other people, and not being an exclusivist, things happen randomly out of nowhere that keep your business alive.

The only times I am not eager to work with people are when I think their technical direction is wrong (e.g. extended attributes), and this is sometimes the significant short term price paid for having clean design.

9) Rules of thumb - by realnowhereman

In your future visions paper early on you talk about Reiser's Rule of Thumb #2. However, I can't find Reiser's Rule of Thumb #1 -- what is it? Is it a secret? Does it contain the sum of all human knowledge?

Hans:

I think it was this, and when I write a book in a few years, I think I had better go looking through some of my old versions of that paper, because it used to have a lot more design principles in it, and I cut them in some ill-conceived effort to reduce the length for some reason I no longer clearly remember. For the convenience of readers I have included the entire context of it.

The Little Inconveniences Dominate What We Do

Do the small inconveniences caused by fragmentation of name spaces really add up to something that should dominate the design concerns for the name spaces of an OS?

Information Diffusion Rule of Thumb: The extent of information diffusion is divided by the effort required to navigate the name space.

Information owners tend to think of the cost of access as only subtracting from the value of their information. but it does much worse, it divides it. The economics are little different from what they would be if money rather than time were the cost, and since we all know that halving the monetary cost of a silicon chip more than doubles its usage, it should seem reasonable to the reader that halving the time cost of a piece of information would double its usage. Three seconds rather than 1/3 second of access time means that the same information will spread to an order of magnitude fewer people, be used by them an order of magnitude fewer times, and be an order of magnitude less useful to the organization as a whole. A common mistake by authors of information is to not realize that most of the total utility of their piece of information will be felt by those to whom its utility is either rather small, or for which its value is speculative to the person considering accessing it. The other common mistake is to not realize or care how much harm will be caused by others expending the time cost of accessing their information only to find it irrelevant. Since we all have limited lifespans in which to do our research, time spent accessing rather than reading information detracts from our ability to wander speculatively after information that might be useful. The quality of the name space design determines these costs.

Example of the evils of name space fragmentation at work: Every employee must create a job description, and then store it in the dreaded Hyped Document Storage System for easy access by all. Mr. B. Bizy just enters HDSS, types Bizy duties, and it pops onto his screen inside a full blown Hyped Editor with lots of features. Easy. Unfortunately the only editor he knows how to use is emacs. Emacs doesn't know how to navigate or edit Hyped Indexes. He takes a moment to swear at the paucity of features possessed by emacs, and then he spends fifteen minutes paging through the Hyped manual. Finally he figures out how to put the document into a file. He edits it using emacs masterfully to achieve a new level of job description ambiguity, and puts it back in HDSS.

His boss comes by, sees him working on his job description, and tells him to compare the list of employees, as entered into the REGRES payroll relational database, to the list of those who have entered a job description into HDSS, to make sure everyone is complying with the "describe your job" directive recently issued. Unfortunately, HDSS is a keyword system and REGRES is a relational system. Neither HDSS nor REGRES can use each others' indexes, and while he could write a program to compare the output from the two applications, he can eyeball the output from the two faster than he could write the program. His eyeballs grow tired.

In general, whenever Mr. B. Bizy wants to act on information namable within one application and operate on it with another application, he must extract it from the first application into a file (at best a pipe), sometimes he must hand massage it into a form that the other application can enter into its own database, then he must put it into the second application, do his work, re-extract it, possibly massage it some more, and finally put it back into its original application indexes. It is never a thoughtless read or save, though it is always tedious. Mr. B. Bizy would prefer to spend his thoughts on other activities.

This is why most of the time the employees of Mr. Bizy's company store most of their data in flat files in the semantically impoverished filesystem: the greater connectivity pulls them there.

10) On being one of those "outspoken" people - by salmo

Mr. Reiser, first off I have no complaints about ReiserFS (which is a high compliment), I use it on almost all my machines, except a couple are running EXT3 because they're not heavily used and I'm lazy at times. But thats neither here nor there.

You fall into an interesting subcategory of project managers or whatever you want to call them. I'll call it the "outspoken genius" category (even though the first word might be understated and the last is probably hyperbole). Basicly your work is technically interesting, applicable, etc. That's a give in. But there are quite a few people who have personal issues with you and your manner and usually cite some exchange or another. Sometimes this is the basis of an argument to reject the use of your work, which I think is somewhat silly. You're not the only one, and certainly not the first to be interviewed here.

So what do you think about this? ie. Do you think you made interpersonal mistakes that landed you here or do you think you've been misunderstood? Does it bother you? Why do you think people enjoy egging on folks such as yourself and then citing the moment you get annoyed with them? Do you think this question ever has a prayer of being moderated higher than someone following the method of the previous question?

Jeeze, I realize I just wrote an essay question in the style of one of my old Philosophy professors. You know the kind, here's a statement now write some stuff (I guess I'll give you a few ideas of where to go).

Hans:

I am not a genius, I am just never satisfied and very very persistent. I approach science like a blind man with a stick who is determined to fully understand what is going on. The difference between me and my competition is that I poke more than they do. I observe, find something to be unsatisfied with, try something to fix it, most of the time it fails and I try again. You don't see the failures because they don't get released. Why haven't other people already fixed the traditional balanced tree algorithms and made them effective enough for storing files? Because it was too much work, and they were smart enough to avoid the work, that is all. We simply rewrite more times and more deeply than others do, and that is how we get our results in our admittedly obscure field.

Now if you think about it, who wants to be around a blind man with a stick, someone who keeps insisting things aren't good enough and they need rewriting?

There is yet another way of looking at it though.

Linux is an ecosystem, and in this ecosystem there is fast growth vegetation and slow growth vegetation. The fast growth vegetation are the people who took what had already been done by Unix, and without changing its design they copied it while making coding improvements.

Then there are those who look at Unix, err, Linux, and see something just barely begun that needs a complete overhaul. These are the slow growth vegetation. Namesys is slow growth vegetation that got started a long time ago.

Now it is human nature that however a human being is, he is inclined to think that is the right way to be. There are those who think that design does not matter, and one should just make incremental coding improvements. There are also those who think that just coding without introducing fundamental new ideas is unimportant. Both of these sets of persons are fools. To say that one approach is better than another is like saying that grass is better than trees, or trees are better than grass. For Linux to prosper as an ecosystem it must have both.

Unfortunately the fast growth vegetation is actually developing a culture of exclusion, kind of like grass working to strangle the tree seedlings. Linux is developing more and more of an insider circle. Those who cannot code well enough to survive on the merits, must politick to exclude the threats....

A sad thing about this is that the most talented young security researcher I know doesn't want to develop for Linux because of the attitude of the inner circle to new people, and I can't really blame him, it is why I didn't develop ReiserFS for BSD back when BSD was....

Almost certainly he is not alone....

It is all very fine to discuss the sociology of herd formation, exclusion, and prejudice in the abstract, but one should never say that particular persons are making particular decisions on the basis of their herd instincts unless one wants to be truly hated by them and all of their numerous friends, and this was my mistake over and above the choice of what sub-herd to be part of.

I don't think anyone "eggs me on" though. I press released benchmarks of ReiserFS vs. ext3 the day ext3 was formally released at a conference, before ReiserFS had been included, and is it a surprise one of them was pissed at me? My competitors didn't and don't want ReiserFS in the kernel, and I wanted and want it in, and the result has all the dignity of a food fight. Filesystems that are less threatening nobody cares enough about to seek to exclude them. Many thanks to Linus, who chooses to allow healthy competition among the filesystems in his kernel.

If only the largest distro was so permissive....

You do all understand that while the GPL doesn't permit tying by license, distros have now moved to using threats of invalidating support contracts to achieve the market leverage they need to exclude competitors, yes? By doing this they can exclude mainstream official kernels from being used, exclude rival filesystems, exclude whatever might lead to less customer lockin.....

This is why you should try to avoid buying support contracts from distros and only buy support from those who agree to support you the customer doing whatever you choose to do, even if it is something fringe like using a kernel from Linus.;-)

They will tell you all this nonsense about how they can't support whatever software you choose to use. Buy better support from an independent and you won't hear this nonsense (www.Namesys.com/support.html is $25 a question and there are plenty of others). Most independents will support you using whatever distro you want, using whatever configuration you want, and they have the skill to cope with that. Sure, they will tell you that such and such gcc release on such and such distro was a lemon, or maybe even that the only reasonable fix for your bug is to upgrade to a recent release, but your support provider should never be telling you that you can only use what they sold you.

I am trying to convince the GSA that they should avoid procuring free software support that constrains the government's choice of what software to use, and they are at least considering the point of view. Why bother to have the GPL if you accept this loss of freedom?

Ummh, maybe these sorts of statements are why I am not so popular....;-) Well, glad to have answered your question!

We should all keep in mind though that there aren't any hard core greedy evil people in our industry. They are all basically good hearted people who chose trying to create a better society as their life's work at a substantial cost in personal income. Petty, bickering, overly impressed with ourselves, flaming, yes that describes most of us Linux kernel developers, but there isn't enough money floating around to attract any genuinely bad folks into our industry.

Not yet....;-)

--

Hans

Comments

WHAT?

[jabbadabbadoo]

"Linux is an ecosystem, and in this ecosystem there is fast growth vegetation and slow growth vegetation.

The fast growth vegetation are the people who took what had already been done by Unix, and without changing its design they copied it while making coding improvements. "

Is this guy working for SCO, or what?

Re:WHAT?

[cwernli]

Why should he ? SCO goes on about code copying, whilst Hans makes the point of design copying - and that's what GNU's all about, and Linux is almost the same (remember, it started out as a Un*x for x86 ?)

Re:WHAT?

[jabbadabbadoo]

I know, I know. It was just the wording; "... and making code improvements..."

Re:WHAT?

[pclminion]

Is this guy working for SCO, or what?

I think by "copy" he meant "duplicate the functionality and semantics." Chill, dude, that was harsh.

Re:WHAT?

[Anonymous Coward]

No, he's telling it like it is.

Re:WHAT?

[Anonymous Coward]

Actually, he's in Moscow, Russia presently. I have no clue why he's there, probably because things in the USA are so fucked, and he has more freedom there.

Re:WHAT?

[Anonymous Coward]

but it's in Soviet Russia. Wouldn't the freedom have him instead?

Well done...

[TopShelf]

I can't recall an interview that came back with such well thought out answers. Major kudos!

Re:Well done...

[EZmagz]

I can't recall an interview that came back with such well thought out answers. Major kudos!

Oh, you're just saying that because your question got answered and the rest of ours didn't... ;)

1) Guideposts? - by TopShelf Having obtained financing for the project, how does that impact the future direction of development? How do you balance the interests of developers, users and sponsors to choose which updates to pursue?

Honestly though, I agree. This was a good interview, and I found most of the answers satisfying (with the exception of the whole Linux = ecosystem skewed analogy). I'm an ext3 man myself...mostly by convenience, since I run RH on about 1/2 of my linux boxen and it's enabled by default. However, I definitely plan on giving Reiserfs a shot one of these days though even though I don't expect to see a huge difference (vs. ext3 on a single-user, low-access system). Maybe if I was running a huge db farm, it'd be different.

An interesting point he brought up was in answer to the fella who is contemplating going back to get his MSci and focus on file systems. Hans told him to basically save his money and hang out in bookstores instead for 4 years, spending his time developing his own gig (a shitty parapharse, I apologize). Hans' answer especially hit home to me, because I've been contemplating grad school as well as of late. However, I know for a fact that I've done way more learning (at least CS-related) since I graduated and got OUT of undergrad than I ever did slaving through a Programming Languages course (Sorry Prof. Allen, Scheme will never be fun no matter what you say). Does anyone else feel like they've learned more on their own than in school?

Just a thought...

Re:Well done...

[red floyd]

Oh, come now. Hans' answers are short and sweet compared to this guy's! :-)

Re:Well done...

[TopShelf]

Oh, you're just saying that because your question got answered and the rest of ours didn't... ;)

That's why I said the answers rocked. Although, it was pretty neat to get my question included (a first).

Re:Well done...

[Anonymous Coward]

Here! Here!

That interview was a good read, and reminds me why I like slashdot.

Re:Well done...

[lars_stefan_axelsson]

Hans' answer especially hit home to me, because I've been contemplating grad school as well as of late. However, I know for a fact that I've done way more learning (at least CS-related) since I graduated and got OUT of undergrad than I ever did slaving through a Programming Languages course (Sorry Prof. Allen, Scheme will never be fun no matter what you say). Does anyone else feel like they've learned more on their own than in school?

Well, two points. The first is that grad school is a bit different than undergrad in that you decide yourself to a much higher degree what to work on. I.e. it's not so much an education as time to pursue your own interests. (And time is scarce when you have to work for a living). This is perhaps more true here in Sweden where a PhD position (a Master's degree is at undergrad level in Europe) is a salaried position which pays decently. (You USians make it up in spades when you graduate though).

The other point is that while you'll always learn more when you're really interested (especially when you're interested enough to spend every waking hour learning more about your subject) you'll tend not to bother at all with the stuff that does not interest you. That's where a University education comes in handy; in forcing you to learn about the areas of your chosen field that you aren't interested in, or even don't agree much with.

This brings perspective, which IMHO is much lacking in industry. In your case even though you still might hate Scheme you now know something about functional programming, which is an important paradigm. Having made the switch myself (from hating SML which is was we studied) to embracing the functional programming paradigm as superior (compared to OO) in my field, that perspective was important, even though I didn't much care for it at the time. The same was also true when I took psychology, I don't care for Freud one iota, and I still don't, but now I at least know what I don't like about his theories, something I would never had studied with out the pressure to do so.

Perhaps ironically, when taking the grad course in operating systems, I had to learn something about filesystems, something I couldn't be bothered with when hacking Minix in the late eighties since filesystems are boooring. Little did I know, and I've since mended my ways. :-)

P.S. If you're ready for some practical industrial strength functional programing check out Erlang.

Re:Well done...

[_Sharp'r_]

I had the same comment hit home to me and am also considering grabbing another piece of paper to hang on the wall and list on a resume.

However, I think I've hit on the "solution". It's Western Governors University, which is fully regionally accredited and does competency based classes.

In other words, if you already know it or want to learn it from a book or doing it, you can then just take a test and get the credit.

Reviewing the requirements, I figure I can handle one semester of tuition to learn the two or three things that I don't already know in their program and graduate.

Re:Well done...

[Hard_Code]

"Does anyone else feel like they've learned more on their own than in school?"

Amen. I'm not regretting avoiding $X0,000 (hm, at an ivy league rate that would be what, $120,000 total for 4 years?), debt also.

Re:Well done...

[Angst Badger]

I know for a fact that I've done way more learning (at least CS-related) since I graduated and got OUT of undergrad than I ever did slaving through a Programming Languages course (Sorry Prof. Allen, Scheme will never be fun no matter what you say). Does anyone else feel like they've learned more on their own than in school?

Well, in my own case, certainly, as my formal CS education consists of exactly one course on C taken at a local community college more than a decade ago. On the weight of totally unrelated skills, I sneaked into the business chiefly as a designer, writing a little code here and there, until I do nothing but write code ten years later.

I was programming for about half that time before I sat down with several textbooks on algorithms and design methodologies and finally flashed on the fact that knowing a programming language is probably the least part of being a programmer, just as knowing a natural language doesn't make you a good novelist. I suppose I could have learned that in school, but the impression I got from friends who were students was that it wasn't a major focus. Of course, my friends weren't attending MIT, and I suspect it's different there.

The biggest part of being a "real programmer" can't be taught in school, and that's the reality of programming in a business environment. No book will adequately prepare you for office politics or time management. Despite my lack of formal training, I've been considered an excellent programmer everywhere I've worked. When I entered the field, I thought that was all that mattered. The big secret, I learned, was knowing where and when to write shitty, sloppy, hastily conceived code in order to get a project out the door on time and under budget, and when to push back to get more time to write good code, because shitty code in that particular case will come back to bite you in the ass and wreck schedules further down the road. That is the job skill par excellence of programming, and you can only learn it from experience.

OTOH, if you're basically perfectionist, the shame of writing corporate code will lead you to spend your spare time writing textbook-perfect open source projects. Occupational hazard, I guess.

You won't learn the fine arts of ass kissing or blame shifting in books, either, but that's something everyone but the receptionist and the mailboy needs to know, not just programmers.

Re:Well done...

[Feztaa]

Does anyone else feel like they've learned more on their own than in school?

Is there even anybody that feels like they learned more in school?!?

Re:Well done...

[Hellkitten]

If you're ready for some practical industrial strength functional programing check out Erlang.

Ericsson employee or ex Ericsson employee?

Erlang was really different from anything I had seen before. Certainly boosted my understanding of how to use recursion.

Re:Well done...

[The Cydonian]

I usually hate posting fanboy comments, but must compliment you on your post. Insightful and definitely helpful for me.

Just curious, but you must be either in KTH or Uppsala? :-)

Re:Well done...

[The Cydonian]

Alright, alright, so I didnt check your homepage. :-)

Bah, this replying-to-my-own-posts nonsense is happening for the second time in two days...

Re:Well done...

[Master of Transhuman]

God, was that ever a good post! MOD PARENT UP!

That was MY biggest mistake - not learning when to write crap to satisfy the boss! And then getting pissed off about it and quitting - which bites you later when you're looking for a job...

Re:Well done...

[lars_stefan_axelsson]

Ericsson employee or ex Ericsson employee?

I'm still there, but only a day a week. The rest of the week I'm at Chalmers Computer Science department, which is quite famous in functional programming circles. (That's not my field though; I'm doing computer security).

And while we're on the subject of functional programming with a practical bent, be sure to check out O'Caml as well. It has static typing with automatic type inference, what static typing should always be like.

Re:Well done...

[Angst Badger]

I had a coworker who was very, very good at kissing ass. He was also one of the crassest and most irreverent people I have ever met. One day I asked him how he could, given his character, kiss ass the way he did.

He looked around to see if anyone was listening, and whispered to me, "Don't think of it as ass kissing. Think of it as patronizing."

Re:Well done...

[drfreak]

>I'm an ext3 man myself...mostly by convenience,
>since I run RH on about 1/2 of my linux boxen and
>it's enabled by default.

pssst! just type 'linux reiserfs' in at the boot prompt for your redhat install. you are welcome :)

It's kinda sad

[Anonymous Coward]

This one little interview puts last night's WinFS story to complete shame.

Reiser latency

[Rosco P. Coltrane]

If ReiserFS has good disk access latency, Hans doesn't : boy did he take a long time to answer the questions, I thought it would never come. The guy must be real busy with his work ...

Re:Reiser latency

[Anonymous Coward]

Perhaps he was waiting for slashdot to send him $250..

Re:Reiser latency

[PetiePooo]

He says right in his interview that he's "slow-growth vegetation." Don't be surprised about the delay... He probably rewrote some of his responses before releasing them!

Re:Reiser latency

[3rd_Floo]

Didnt you see the ReiserCVS tree he built to answer them????

Wow.

[bwhaley]

You do all understand that while the GPL doesn't permit tying by license, distros have now moved to using threats of invalidating support contracts to achieve the market leverage they need to exclude competitors, yes?

I wasn't aware of this at all. I'd like to see this expanded, i.e. what distros are doing this? Does it violate any GPL issues? Why are these distros undermining the glue that holds Linux together? What does Linux think?

Competition is important. Hans is exactly right when he says that no support contract should tie a customer to a specific piece of software. Free software is all about choice!

As The Dude would say, "this is a bummer, man..."

Ben

Re:Wow.

[bwhaley]

What does Linux think?

Linus*

Damned slashcode. Why can't I go fix it?

Re:Wow.

[usotsuki]

Red Hat for one is known to tie use of their server software to a support contract.

Looks Evil, Bad and Worng (deliberate spelling error) to me.

-uso.

Re:Wow.

[IamTheRealMike]

Competition is important. Hans is exactly right when he says that no support contract should tie a customer to a specific piece of software. Free software is all about choice!

So if a company doesn't support your software, go find somebody who will. I'm sure various trolls will turn this into a "Red Hat are EVILLLLL!" issue, but of course the reality isn't like that.

Support is hard, because software has a nearly infinite number of combinations. If you're going to provide reliable, accurate support, you can only have expertise in a small subset of those configurations.

Can YOU imagine being on the end of the phone, trying to help somebody recover a server and every few minutes find that they were running with experimental kernel patches, or ancient/buggy software, or that a fault seemed to be caused by a random frob off SourceForge that you'd never heard of? Total nightmare! You have to draw the line somewhere.

This is why CodeWeavers only support a subset of the available Win32 applications, and don't support you if you hack CrossOver to use a pre-existing Windows installation etc, the number of unknowns gets so high that it's not only not profitable, but you run the risk of giving the customer bad advice.

Re:Wow.

[Telastyn]

Umm why? A distro doesn't have to include GPLd software in their distro; and they certainly don't need to support it. If you don't like it, pick another distro/support contract. The GPL will ensure that the distro provides the source so that you may do your own support if needs be, and that others can provide competative support [unlike microsoft for example because they're the only ones that *can* provide support/fixes on certain issues]

Re:Wow.

[nadadogg]

Well, the distros would undermine the glue for one reason. Money. The biggest distros couldn't exist without their big service contracts and the techs that handle it. If a move to a newer file system would require them to spend money training, they obviously don't want to spend more money, thereby requiring you to stick with the old file system.

Re:Wow.

[kinnell]

Competition is important. Hans is exactly right when he says that no support contract should tie a customer to a specific piece of software. Free software is all about choice!

To play devils advocate, when a support contract is negotiated, it has to be costed by the company providing the support. Supporting a uniform system is much easier/cheaper than supporting a system which not uniform, and could have all sorts of unanticipated problems. Allowing the customer to mess around with the distro in whatever will almost certainly reduce the profit margins of the supporting company.

While this is very un-linuxy, it allows the big distro makers to thrive as commercial entities, and on the whole, they seem to contirbute a fair bit back to the open source community. It also has a side effect in opening up a market for small businesses/consultants who will support anything. Which is good.

Re:Wow.

[bwhaley]

Support is hard, because software has a nearly infinite number of combinations. If you're going to provide reliable, accurate support, you can only have expertise in a small subset of those configurations.

Agreed, support IS hard. But I disagree that "you can only have expertise in a small subset of those configurations." Take, for instance, a typical University or ISP helpdesk. There is a HUGE variety of issues that must be supported, almost all due to user error. The point of the helpdesk is to be helpful in as wide a variety of areas as possible. Well-organized helpdesks accomplish this invaluable service with just a few weeks of training.

Your examples are not valid here. "experimental kernel patches, or ancient/buggy software, or that a fault seemed to be caused by a random frob off SourceForge that you'd never heard of" are not the software in question. ReiserFS is a stable and mature filesystem in use by millions (read: made up statistic) of people. It is not fair for a distribution, who should be promoting competition rather than inhibiting it, to disallow use of software because of personal issues.

I'm glad Linus has enough foresight to include it in the kernel.

Re:Wow.

[norweigiantroll]

Umm why? A distro doesn't have to include GPLd software in their distro;
A Linux distro doesn't have to include Linux?

Re:Wow.

[PetWolverine]

Can YOU imagine being on the end of the phone, trying to help somebody recover a server and every few minutes find that they were running with experimental kernel patches, or ancient/buggy software, or that a fault seemed to be caused by a random frob off SourceForge that you'd never heard of?

No, I can't imagine Hell. I can, however, imagine a company claiming to support Linux at least making an attempt to support kernels that are fully official and straight from Linus. This is what Hans was complaining about: that some distros don't support official releases of the kernel because they don't like this or that feature.

If you say you support Linux, then support Linux.

Re:Wow.

[justins]

So if a company doesn't support your software, go find somebody who will. I'm sure various trolls will turn this into a "Red Hat are EVILLLLL!" issue, but of course the reality isn't like that.

Support is hard, because software has a nearly infinite number of combinations. If you're going to provide reliable, accurate support, you can only have expertise in a small subset of those configurations.

And yet, SuSE and even Mandrake manage to package good, stable, working Reiserfs (not to mention XFS) code, and have for years. It's not unreasonable to ask, what is Redhat's problem?

Re:Wow.

[Telastyn]

Could be a BSD distro I suppose :P

Re:Wow.

[IamTheRealMike]

Take, for instance, a typical University or ISP helpdesk.

The kind of support Red Hat provide is not the same kind as the kind a helpdesk provides.

Red Hat are there to support the help desk. An answer of, "Umm, dunno" is not acceptable. If they don't know the answer, they have to be able to say, "we'll get our engineers right on it, you'll have an answer soon". They can't do that for a piece of code they have no expertise on.

Support in this context also means patches and updates. Obviously if you install your own stuff, and then get rooted because of a bug in that, you can't expect Red Hat to take the fall for it, hence, no support for you. Maybe they could be more flexible about it.

Your examples are not valid here.

Why not? If RH have no expertise on a piece of code, what difference does it make where it came from?

It is not fair for a distribution, who should be promoting competition rather than inhibiting it, to disallow use of software because of personal issues.

Personal issues? What personal issues? Yeah, some RH employees have (technical) doubts about the reliability of ReiserFS. That's not personal, if they don't trust Hans' work for whatever reason, he should try and convince them (or customers choose somebody less conservative).

I'm glad Linus has enough foresight to include it in the kernel.

If Linus includes it in the kernel and people use it and it goes wrong, they have nothing on Linus. On the other hand, Red Hat is there partially to take the blame for things like that. It's a whole different set of rules.

There was another reply about Mandrake and SuSE, which I'll answer here as well. Red Hat make ALL their money from support, as (virtually) all their software is GPLd and always has been. Mandrake and SuSE on the other hand both sell their distro and are far more reliant upon support from home users. And I think you'll find that they are similarly inflexible with respect to "bending" your support contract, if they choose different components to do so then the market will decide.

Now, I agree it'd be nice to include ReiserFS in RH Linux, ie the unsupported version meant for hackers and hobbyists. The relevant bug is here, so why not go and vote for it (but try not to spam the bug with "me too" comments). The reason given there for not including it is lack of time to integrate with the installer. BS or the truth? You decide, nobody forces you to use Red Hat.

Re:Wow.

[Wolfrider]

--If $DISTRO doesn't run/support a stock Linus kernel, I won't use it. Period. About the 1st thing I did when installing SuSE 7.3 (years ago) is upgrade the outdated (2.4.10?) kernel to the latest Linus 2.4 source code. Everything still worked.

Re:Wow.

[Requiem]

Truly. I've been using SuSE 8.0 for almost a year now, and have had *zero* problems with it what-so-ever.

Re:Wow.

[Anonymous Coward]

It does rather suggest that the "We can't support all this variety of software" is, in actual fact, a smokescreen for "We don't like Reiser, so we don't touch it".

Re:Wow.

[SlashDread]

At risk of being modded a troll.. but how much support does Red-hat generally give on "Non-US"?

Hey!? If US and EU patent law will be sufficiently different, would that validate a "Non-EU" catagory in debian?

Ok.. ok im being silly..

Gr /Dread

Re:Wow.

[Ringlord]

You decide, nobody forces you to use Red Hat. That's right, I use Slackware. One of the reasons are that ReiserFS is included by default. I have used ReiserFS for several years and have found it to be very stable and reliable.

Re:Wow.

[ajs]

Red Hat ships with Reiser.

$ rpm -qi kernel
Name : kernel Relocations: (not relocateable)
Version : 2.4.20 Vendor: Red Hat, Inc.
Release : 8 Build Date: Thu Mar 13 18:01:52 2003Install Date: Wed Jun 11 17:18:16 2003 Build Host: porky.devel.redhat.com
Group : System Environment/Kernel Source RPM: kernel-2.4.20-8.src.rpm
Size : 31954258 License: GPL
Signature : DSA/SHA1, Thu Mar 13 18:20:14 2003, Key ID 219180cddb42a60e
Packager : Red Hat, Inc. <http://bugzilla.redhat.com/bugzilla>
Summary&nbs p; : The Linux kernel (the core of the Linux operating system)
Description :
The kernel package contains the Linux kernel (vmlinuz), the core of your
Red Hat Linux operating system. The kernel handles the basic functions
of the operating system: memory allocation, process allocation, device
input and output, etc.
$ rpm -ql kernel | grep reiser
/lib/modules/2.4.20-8/kernel/fs/reiserfs
/lib/modules/2.4.20-8/kernel/fs/reiserfs/reiserfs. o

Nuff said?

Re:Wow.

[ajs]

In the example that you give, the college helpdesk a) does not lose customers if they can't solve your problem b) does not have to QA everything that they support.

Those two factors should be answer enough to the question of why Red Hat doesn't just support "stuff that runs on my box".

Re:Wow.

[amorsen]

Depending on official kernels is hopeless. What will you do if the official kernel leaves a local root exploit open for months? If you patch it yourself, you are not using the official kernel anymore...

Re:Wow.

[4of12]

Can YOU imagine being on the end of the phone, trying to help somebody recover a server and every few minutes find that they were running with experimental kernel patches, or ancient/buggy software, or that a fault seemed to be caused by a random frob off SourceForge that you'd never heard of?

Yes, unfortunately.

Re:Wow.

[Richard_at_work]

He said distro, that could include the BSDs as well.

Re:Wow.

[IamTheRealMike]

According to bugzilla it's a resource/man hours issue (for installer integration).

I hope it gets in soon, but I'm not going to crucify Red Hat for this omission.

Re:Wow.

[kbielefe]

I agree completely. There is definitely room in the market for both the big commercial distros and the independent support. However, my personal opinion is if you have enough in house expertise to make a non-standard system, that you probably have enough in house expertise to support the non-standard stuff (or at least you should).

For example, I really wanted the usb-storage updates in the new 2.4.21 kernel. However, I use Win4Lin and netraverse hadn't released a kernel patch for 2.4.21 yet (they have since, but I had no idea how long it was going to take). So, I applied the 2.4.20 patch (which succeeded with hunks) and hoped for the best.

Did I have a right to ask netraverse for support if Win4Lin didn't work under 2.4.21? Of course not. If I didn't have the personal expertise to make it work I should have backed out to 2.4.20 and waited for the official patch or hired someone who did have the expertise. That was a simple example, but you get the general idea.

Re:Wow.

[Hellkitten]

if the official kernel leaves a local root exploit open for months

You either fix it yourself, wait for someone else to fix it or pay them to fix it. Most likely the exploit becomes known simultaneously with the fix for it, since the one that finds the problam is likely to provide a patch

If you patch it yourself, you are not using the official kernel anymore

If you patch a root exploit and it really fixes the problem without introducing new ones you will be using the official kernel, because it will become part of the official kernel faster than you can say "recompile"

Re:Wow.

[amorsen]

If you patch a root exploit and it really fixes the problem without introducing new ones you will be using the official kernel, because it will become part of the official kernel faster than you can say "recompile"

Bzzt. Thanks for playing. I picked this scenario precisely because it was decided to not release 2.4.21 early even though 2.4.20 had a local root exploit. If you were running the official kernels, you were vulnerable between 17th of March at the latest (this is the time of the official announcement with patch, it seems) and 13th of June.

Official kernels are not useful for production on boxes that I have the responsibility for.

Re:Wow.

[Anonymous Coward]

And yet, SuSE and even Mandrake manage to package good, stable, working Reiserfs (not to mention XFS) code, and have for years. It's not unreasonable to ask, what is Redhat's problem?

In order to create Reiserfs partitions in Redhat's install, type "linux reiserfs" at the boot prompt.

Re:Wow.

[justins]

Nuff said?

Not really. Why can't I install to it?

Re:Wow.

[spacemky]

to us reiserfs on RedHat:

step 1: boot from CD
step 2: at the boot prompt type: linux reiserfs
step 3: ?????
step 4: profit!!

Re:Wow.

[Narcissus]

Nobody said you had to rely on it. What was said is that it should be supported.

It would be kind of like Microsoft saying "Sorry, we only support Windows if you've installed 14 different pieces of spyware and hand edited your registry."

Obviously, you don't want to use Windows in it's pure state, but it should be supported.

Re:Wow.

[Anonymous Coward]

I'm running Redhat 8, with reiserfs on all my drives (except /boot), but I cannot install with reiserfs as a choice. I had to install with ext3, then move everything off, make file systems reiserfs, move data back.

That said there's another problem. Redhat assumes that your not running reiserfs for /. What happens when you try to reboot after you installed yet another updated kernel from redhat? Your system won't boot? wtf?!

Well, the initrd they build doesn't load reiserfs kernel module. What I did was un-compress the initrd, mount it via loopback, delete the ext3 module and copy over the reiserfs module from the new kernel, compress it back up and put it into my /boot. I suppose I could use the mkinitrd util but thought that my way was easier.

I wish Redhat included reiserfs as an install option...

Re:Wow.

[Anonymous Coward]

He said distro, that could include the BSDs as well.

Which include stuff like gcc.

Re:Wow.

[strobert]

Not only do I agree, but have an additional point to make. If Hans was refering to RedHat (and it seems pretty obvious he was) then he is mistaken. I have NEVER been threatened by RedHat to have a support contract invalidated. The basic effective policy I have experienced is if you install something outside of what they ship they will try to help but you may end up on your on. Or at least showing the problem also exists in the stock verision. I don't tap RedHat support very often (generally internal staff and basic net research solves the problem) but when we have we have gotten good results. Far better than I have gotten from other commercial support contracts (think companies like Cisco, Microsoft and Sun).

Re:Wow.

[amorsen]

Why should it be supported?

The support is useless, since the next security vulnerability forces you off into patch-land again, and then the support is gone. Linus' kernel is a reference implementation these days, the distro kernels are actually useful.

Re:Wow.

[Error27]

It's red hat advanced server we're talking about not regular RedHat.

It's pretty stupid to pay $2500 for a well tested kernel with a support contract and then go and install a random Linus kernel that might not even compile.

Re:Wow.

[Richard_at_work]

They dont have to, infact Its quite plausable that I can install a BSD base system with no GPLed applications.

Re:Wow.

[eugene_roux]

What does Linux think?

Linux thinks that Linus is god and that Tux is cute...

Oh, you meant "what does Linus think?"

So sorry... :-)

Re:Wow.

[ajs]

Yeah, you can install reiserfs as boot under RH9, AFAICT from the documentation. Baby steps... a company that puts out a new filesystem the day it's released is not the company that I want to do business with. The reason that RH released ext3 when they did is that they had been using it in production for a year. Same goes for reiser. I'm pleased with that level of dedication to QA, and a little dissapointed at the distros that released reiser ASAP only to have massive problems with the FS (all of which are fixed now, as far as I know, so that's good).

Re:Wow.

[drfreak]

I'm not going to say Red Hat is evil as I run it myself. But I will tell you one thing: support *is* hard, but nine out of ten support people are *lazy* and saying "I can only support a limited subset of software" really says a lot about your lack of skill, including your lack of willingness to acquire any. Sounds like an MCSE if I ever heard one..

Re:Wow.

[Anonymous Coward]

I doubt it's an installer thing - the hooks are
already there, in fact I think you can install on
reiserfs with the appropriate s3kr3t boot argument.

Re:Wow.

[salimma]

The reason given there for not including it is lack of time to integrate with the installer. BS or the truth? You decide

Neither - it is actually possible out-of-the-box if you type 'linux reiserfs' when booting the install CD.

The actual reason is that, as mentioned before, ReiserFS is not supported by Red Hat. Not that they can't integrate it with the installer - after all, the XFS team has been shipping their modified installer with XFS support for ages; adding a filesystem to the installer can't be that hard.

Revision control in the filesystem..

[lpontiac]

I thought it not being in the filesystem was adequate until I saw a demo of Katie, which lets you mount the repository as an NFS filesystem and access snapshots of the repository at certain points as different directories, take diffs just by running the standard diff between these directories, etc.

If only there was a stable (Katie's author describes it as "pre-alpha") and free piece of software to do it...

Re:Revision control in the filesystem..

[rpeppe]

If only there was a stable [...] and free piece of software to do it...

Well, there's always Plan 9's fossil. Oh... you mean Linux software?

Re:Revision control in the filesystem..

[Anonymous Coward]

I thought it not being in the filesystem was adequate until I saw a demo of Katie, which lets you mount the repository as an NFS filesystem and access snapshots of the repository at certain points as different directories, take diffs just by running the standard diff between these directories, etc.

Actually this is Hans' point. This is exactly what clearcase does. In fact, the author of Katie was trying to create a Clearcase clone.

-- ac

Re:Revision control in the filesystem..

[lpontiac]

Well, there's always Plan 9's fossil [bell-labs.com]. Oh... you mean Linux software?

Hmm, interesting, thanks for that.

We're actually a FreeBSD/MacOS X/Win32 shop, and the nice thing about CVS is that we can run the client (and indeed the server, although we've never served from anything but BSD) on each of the machines. I haven't looked closely but in principle I suppose it should be possible to NFS export a fossil repository from Plan 9..

explanation pls

[an_mo]

>If only the largest distro was so permissive....

can somebody explain this sentence to me? I am not sure what he is talking about.

Re:explanation pls

[sheddd]

Redhat doesn't 'easily' support ReiserFS.

Re:explanation pls

[kalimar]

I believe he's talking about the support contracts offered by the largest distributor of Linux. I'm not sure which distributor that would be, though at a guess I'd say RedHat.

Re:explanation pls

[Feztaa]

>If only the largest distro was so permissive....

can somebody explain this sentence to me? I am not sure what he is talking about.

If you've ever done a RedHat install recently, you might have noticed that while it will let you use a pre-existing ReiserFS partition, it won't let you format a partition as ReiserFS before using it.

So, if you really, really, really want to use ReiserFS with RedHat, you have to preformat the partition before starting the installer, then tell the installer not to format the partition for you.

Re:explanation pls

[AKnightCowboy]

So, if you really, really, really want to use ReiserFS with RedHat, you have to preformat the partition before starting the installer, then tell the installer not to format the partition for you.

Really now? What's their explanation for that? I'm going to have to install a couple dozen new servers in a few months and I have to use Red Hat (the software is only supported under Red Hat). I'd hate to be stuck with ext3 as a filesystem because I find it extremely lacking compared to reiserfs under Debian and Mandrake. So that would suck if I have to boot off a Debian boot disk to format the partition before I installed Red Hat. *sigh*

Re:explanation pls

[Feztaa]

Really now? What's their explanation for that?

Well, not long ago I had a /home partition that was ReiserFS, and I didn't really want to format it just to install RedHat, so redhat let me use the partition "as-is". I haven't tried using a root partition that's formatted as ReiserFS, but I don't see why it wouldn't work.

RedHat probably just lets you use a ReiserFS partition if you already have one, because they wouldn't win many friends if they forced everybody to format their ReiserFS partitions as ext3, making them lose all their data. If you're going to be formatting a partition during the install _anyway_, then they want you to be using ext3, and don't give you much choice.

Re:explanation pls

[Anonymous Coward]

Redhat doesn't 'easily' support ReiserFS.

Wrong. ReiserFS comes with Red Hat.

Re:explanation pls

[AKnightCowboy]

If you're going to be formatting a partition during the install _anyway_, then they want you to be using ext3, and don't give you much choice.

But, like I said, ext3 is horrible. It's just ext2 with a journal and ext2 is horrible at many tasks. Take a squid proxy for instance. It can take minutes to create a large cache directory for Squid under ext2 or to add or remove a large amount of directories. Under Reiser it takes less than 10 seconds on my system (creating 45GB+ cache). I sincerely hope Red Hat adds Reiserfs, XFS, and JFS as choices to install as the boot partition in future versions.

Re:explanation pls

[42forty-two42]

I use Gentoo, which recommends ReiserFS:

ReiserFS is the filesystem we recommend by default for all non-boot partitions.

Re:explanation pls

[sheddd]

As do I :) Ext3 for me, though; none of that newfangled stuff.

Re:explanation pls

[Anonymous Coward]

Reiserfs isn't that new. It was actually included in the kernel before ext3 was.

I know what you mean, though. ext3 descends from ext2. It's compatible with ext2. I like that. Unfortunately I switched all my partitions to ReiserFS when ext3 was still an unmerged patch, and ReiserFS was in 2.4.x. :(

Re:explanation pls

[Anonymous Coward]

When you boot your RedHat 7.2 or 7.3 CD, type

"linux reiserfs" instead of just hitting Enter and ReiserFS will become a partition option in the disk druid.

I have never ever though I'd see this

[Bombcar]

He takes a moment to swear at the paucity of features possessed by emacs,

Never have I expected to see this comment.
Pigs must be a flying.

Re:I have never ever though I'd see this

[haapi]

So Hans has a sense of humor!

Re:I have never ever though I'd see this

[Anonymous Coward]

Here's your flamebait for the day:

A vi user can look at vi, with its fairly minimal set of features, and not wish for much more at all. This is because the vi user expects his editor to just be an editor and nothing else.

The emacs user, however, can actually look at emacs, which has far more features than vi will ever have, and see it as lacking features. This is because the emacs user expects his editor to do everything, and he is disappointed if he runs into a situation where it doesn't do everything.

Re:I have never ever though I'd see this

[Anonymous Coward]

Pigs must be a flying.

Maybe you can make a ReiserFS plugin of this.

Re:I have never ever though I'd see this

[dododge]

Caveat: emacs is usually the first thing I start when I log in and the last thing I quit before logging out, but I do occasionally use vi. Friends who are vi-diehards have mentioned this same issue...

A vi user can look at vi, with its fairly minimal set of features, and not wish for much more at all. This is because the vi user expects his editor to just be an editor and nothing else.

If only that were true. Unfortunately there have been occasions, usually with major Linux distributions, where I've run "vi" and found that it was really some uber-enhanced monstrosity pretending to be vi. For example "vi foo.html" has sometimes resulted in interpreted HTML with bolding, coloring, and no visible tags.

When I ask for "vi", I want "vi", dammit.

Re:I have never ever though I'd see this

[sgtrock]

The times that I've seen this it's almost always been someone who renamed vim to vi, or just created a symlink between the two. :)

Versioning

[marktoml]

Automatic versioning of files is one thing I actually miss from VMS. Certainly not most of the rest of it :)

Re:Versioning

[j3110]

Haha, I was going to post the same thing. I'm suprised many people even remember VMS versioning, but it was done pretty good I would say. Maybe listing all the versions with an ls was a bad idea, that should probably be a different command. Besides, I hate seeing file~ whenever an editor wants to save a backup copy.

I, for one, will be the first idiot to jump on the versioning train when it gets here in a pre-alpha form that is gauranteed to crash. It's such a useful concept, and implementing it at application level is just not as good in a lot of respects.

Re:Versioning

[Anonymous Coward]

while useful, vms versions were whole files. what they are talking about is sort of a cvs filesystem where file changes are tracked, and you could roll back to any given version of a file. a bit higher overhead, but lower storage requirements

Re: Versioning

[plsuh]

Version control definitely belongs in the filesystem. Clearcase may have a lot of implementation uglies, but as a concept it clearly works. Filesystems manage files, and that should include managing file versions.

One place to look for (sort of) file system versioning support is Subversion, an Apache-licensed replacement for CVS. You can mount this via webdav_fs, and you get automatic versioning that works with just about any application, although right now you need to use a separate client to get access to older versions of the files. Still in alpha, but very stable and worth checking out.

--Paul

Re:Versioning

[Anonymous Coward]

I thought of it as well. VMS versioning actually has pretty simple semantics. I bet you could write a user-space c library add-on pretty easily. Folks need to realize tthat this is not the same as CVS/RCS/bitkeeper so don't have those expectations. It IS a useful technology (ask any VMS-head).

Re:Versioning

[crmartin]

With hard-drive costs less than $1 a GIGABYTE (you younger folks have NO IDEA how impossible that sounds) I'm not sure that keeping file deltas is worth the effort any longer. VMS versioning was fairly straightforward and you didn't HAVE to look at all the back versions: if I recall the syntax correctly, 'dir foo.*;-3' would show you the three most recent versions (don't bother to flame me if I'm wrong, it's been 20 years) and a DCL macro would let you make it easy. Add to that a way to transparently archive the files to WORM, like Plan/9 does, and you're in fat city.

Re:Versioning

[Skyshadow]

With hard-drive costs less than $1 a GIGABYTE (you younger folks have NO IDEA how impossible that sounds) I'm not sure that keeping file deltas is worth the effort any longer.

I gotta disagree -- while the hard drive price to size thing is absolutely jaw-dropping (I remember looking on my first 100 MB drive with absolute awe), it doesn't make sense to waste space if you can come up with a reasonably efficiant alternative.

I don't know what strategy Reiser has in mind, but even after just a few minutes of mulling it over I was able to come up with a couple of promising-seeming methods for keeping your data intelligently without slowing down filesystem access -- you're always going to use more CPU, but unless your processers are pegged 100% of the time there's no reason that the versioning can't be run as a background job.

The easiest way I can think of would be to have the user work with a "snapshot" of his latest files, which would be updated as he worked and user to create a delta'ed version in the repository. This seems much more effective than trying to constantly recreate each file dynamically like MVFS does...

Re:Versioning

[RevMike]

Quick roundup - by a current user of VMS (on Vax!):

-- standard commands, using dir for examples --
dir foo.txt;* operate on all versions of foo.txt
dir foo.txt;0 operate on the most recent version
dir foo.txt;n operate on version n
dir foo.txt;-n operate on the nth most recent version

-- The defaults --
dir foo.txt - show me all versions
delete foo.txt - doesn't work, needs explicit ;*
edit/tpu foo.txt - open the most recent version

-- other interesting commands --
purge foo.txt /keep=n - keep the n most recent versions
copy bar.txt foo.txt - create new version of foo.txt with version number incrmented
copy bar.txt foo.txt;n - overwrite foo.txt version n

-- remember most recent version has largest num --
$ dir foo.txt
foo.txt;1 - earliest
foo.txt;2
foo.txt;3 - latest

$ copy foo.txt;1 foo.txt
foo.txt;1 - earliest
foo.txt;2
foo.txt;3
foo.txt;4 - latest - identical to foo.txt;1

$purge foo.txt /keep=3
foo.txt;2 - earliest
foo.txt;3
foo.txt;4 - latest

Versioning is great. VMS has other cool features, as well. My favorite is the "environment" (LOGICALS in VMS speak) spaces that are shared at different system levels. To put it into Unix speak, imagine that an appropriately privileged user could export an environment variable to every environment on the system. The logicals can be used for lightweight IPC, better than writing flag files to the filesystem. If I recall there are four namespaces of logicals: SYSTEM - shared by everyone; USER - shared by all processes for a given USER; PROCESS - shared by a process and all its subprocesses; and LOCAL - isolated to a process.

The worst parts of VMS are the bizarre device:[dir.subdir]filename.ext;version format and the lack of real pipes.

Re:Versioning

[Crispy Critters]

" Automatic versioning of files is one thing I actually miss from VMS. Certainly not most of the rest of it :)"

I agree completely. As an aside, when you want this kind of versioning, you can get emacs to write out versions as somestuff.~1~, somestuff.~2~,...with the most recent copy named just somestuff. The biggest reason I don't use this is because I can't get ls to sort them the way I would like, especially when it gets to version ~10~.

Re:Versioning

[Roadkills-R-Us]

That was my thought on reading that quote.

There are several things I miss about VMS, including automatic versioning, delete permissions on files, ACLs (finally showing up outside "DARPA-classed" *nix), the help system, and system vs world privs, and user, group and system (world? I forget) variables.

It has always boggled my mind that the vast majority of [outspoken?] *nix users have nothing good to say about VMS.

While *nix has kept a great deal of the elegance and simplicity of the early days, in many ways it has grown far more complex than VMS. It's well past time that we admit to ourselves there *were* good tings in VMS, and embrace them within our OS of choice.

Re:Versioning

[Anonymous Coward]

*sigh*...I'm still using VMS...not by choice, mind you.

Re:Versioning

[anthony_dipierro]

it doesn't make sense to waste space if you can come up with a reasonably efficiant alternative.

Well that comes back to the slow growth/fast growth argument. Some of us would prefer to waste space and have a solution rather than spend years and years solving the problem.

Of course, in this case a medium growth solution would probably be even better, such as the one you presented in your last paragraph.

Re:Versioning

[Nevyn]

The biggest reason I don't use this is because I can't get ls to sort them the way I would like, especially when it gets to version ~10~.

ls --sort=version

Re:Versioning

[yerricde]

With hard-drive costs less than $1 a GIGABYTE (you younger folks have NO IDEA how impossible that sounds) I'm not sure that keeping file deltas is worth the effort any longer.

Unless of course each version of e.g. an audio or video file is multi-megabytes in size. Yes, I know that the current version of CVS doesn't grok audio or video, but the point remains that there exist files that are larger than text files. Besides, if the files are compressed as diffs, many more projects can be hosted on a single set of servers. Without compression, SourceForge.net probably wouldn't be able to offer its services at no charge.

Re:Versioning

[kbielefe]

I hate seeing file~ whenever an editor wants to save a backup copy.

That's why I have
set backupdir=$HOME/.vim_backups
in my .vimrc. The backups are still there in case I ever do something stupid, but I don't have to look at them all the time. I'm sure emacs has something similar. I liked VMS versioning too, but my directories did fill up really fast.

Re:Versioning

[listen]

Most of the LOGICALS stuff can handily be done using the filesystem. IE writing stuff to /tmp ~/.tmp ~/.tmp/pid/

Of course, it is also completely dodgy in terms of security, we all know about the ultra fun of /tmp symlink races. I can't see how VMS could do better here, is it not a vulnerabilty of any shared namespace?

This will hopefully become slightly more elegant with true per process namespaces ala plan 9, I'm hoping environment variables themselves can be implemented in terms of a stack of ramfs instances, ie all the parents environments. Need some copy on write logic to maintain old env vars when the parent overwrites them though...

Anyway, the main random namespaces that annoy me in Linux atm are: network devices, tcp/udp/etc ports, hostnames, device numbers, SQL servers / XPath based xml things / other sub-file addressing schemes / dodgy all consuming namespace things - kio/ gnomevfs.

Hopefully this can all be integrated in to one posixesque-but-sure-as-fuck-not-posix like namespace with reiserfs and a few other tricks.

Re:Versioning

[Anonymous Coward]

Which of you kiddies remembers the automatic file versioning built into the Univac Exec 8 OS where you simply specified "filename(-3)" to get the third previous version of a text file? Some day, PC software will catch up to the standards of mainframe software from the 60's

[OT] Re:Wow.

[cwernli]

Damned slashcode. Why can't I go fix it?

A hundred obvious reasons.

The first one is: use your imagination. When it says "Use the preview button" it acutally means "re-read your comment, and watch out for logical inconsistencies, spelling errors and ECHOLON-triggering expressions". :)

Re:[OT] Re:Wow.

[revery]

ECHOLON-triggering expressions". :)

What is this E-COLON thing? Tracking information through the bowels of the internet? Sounds like something the government would fund ;)

Re:[OT] Re:Wow.

[Master of Transhuman]

bowels of the Internet...

Never heard /. described so concisely before...

High colonic - I think that's a hiccuping drunk US Army colonel in Iraq?

Re:[OT] Re:Wow.

[Anonymous Coward]

Redhat obviously, those guys are Microsoft all day, oure amrketing and now pure legal BS. They only sell "Entitlements" which constrain you to per server per year use, and they have the right to send in the jackbooted Redhatted nazi sw police to make you back pay.

Specifically RH dissed ReiserFS when it was stable and still does. I had a customer that wanted to run it and RH told them no way under thier support. They also dissed LVM for thier childish SW RAID. What UNix does not have an LVM? RH.

Re:Licensing is a shame

[Blasphemy]

If you'd read the article more closely, you'd see that the filesytstem can also be licensed for a fee (ie a non-free license). If others want to incorporate his work and not make the result freely available, they will have to pay for the privilege.

This prevents M$ from stealing his work, making it incompatible with the main branch and calling it WinFS. This is the beauty of the GPL and why I consider it to be a "truely free" license, while I consider the mBSD license to be just a half-step above the public domain. Just look what happened (is happening) to Kerberos.

Typo.

[Anonymous Coward]

...for paucity of relevant features. No pyschotherapy for me today, thanks. I just want to edit this file....

Re:Licensing is a shame

[Anonymous Coward]

Of course, if those other operating systems were licensed under the GPL...

Say it ain't so!

[Anonymous Coward]

Red Hat for one is known to tie use of their server software to a support contract.


who chooses to allow healthy competition among the filesystems in his kernel.

If only the largest distro was so permissive....

I thought he was talking about Yggdrasil.

Berlin

[kyoko21]

After reading this interview, all I can say is quoting from the lyrics of Berlin (80's Band) "you take my breath away." Wow, what a read. :-) One of the best interviews I have read.

Would you rather have...?

[phorm]

(a) Short but quick, not overly useful answers

(b) Long-thought-out, rather details, quite intelligent and possibly useful answers?

I'll choose (b), as I've noticed that it is notoriousl difficult to release a post-release patch for slashdot comments...

Linux attitude towars newbies....

[botzi]

I know doesn't want to develop for Linux because of the attitude of the inner circle to new people,

I recall that in an interview a few months ago, Eric Raymond said that on the Linux kernel list there're some dinosaures, that have a hard time acception anything/anyone new.
Back at that time, I didn't take him seriously(because the reason of his statement was that his kernel configuration tool was not accepted, and he was really affected by it....) and I marked the phrase as an act of anger.
However, now a second (respected) personality is saying almost a similar thing and I'm starting to wonder. Is there anybody out there who has had similar experience when he wanted to join some Open source project?? Do you think that OS developpers guard their "territory" even more than at the workplace??? Finally, when you're new somewhere, the integration in the team is somehow easier due to the fact that it's the boss who put you on that job...

PS: For the Eric Raymond interview, google something about his last book and it'll come out....

Re:Linux attitude towars newbies....

[Pegasus]

This is common in any dynammicaly evolving system. As it grows, it's becoming more and more complex, which leads it into being more and more conservative.

So this 'inner circle' thing is a natural way of developement ...

Re:Linux attitude towars newbies....

[jmv]

I can't comment on the lkml specifically, but I think I've got an idea to explain part of the bad attitude towards newbies. I think it comes from the fact that many start by commenting about how they would do/change things without even understanding the software first. This is the best way to annoy developers and after a while some of the developers might get a "bias against newbies".

Re:Linux attitude towars newbies....

[Lennie]

As no one else answers it, so I will (but without much of an explanation): I think it's completely the same.

There is no difference, maybe it's easier to misunderstand via e-mail, but that's it.

Re:Linux attitude towars newbies....

[Anonymous Coward]

Developers have to "guard their territory", or the project may mutate into something with horrible problems. You can't just accept every patch that comes along. Doing that leads to chaos.

I once had someone send me a patch for my project that would have added dependencies in the Makefile for every .h file that was referenced somehow. I'm not talking about the usual headers that are part of the project. Oh no. This guy also included all of the *system* stuff, including the OS and arch-specific stuff, like asm-i386/foo.h.

I ignored it, and the project hasn't imploded yet, so I guess that was the right thing to do.

One of the strengths of free software is that anyone with a C compiler, an editor, a copy of 'diff', and a mailer can send patches to the author(s). This is also one of its weaknesses.

Re:Linux attitude towars newbies....

[Darth Daver]

I still hear people complain that they tried to get help learning Linux, and someone was mean and insulting to them on a forum or mailing list. Of course the last person that told me that is an inveterate liar (Oops. I mean he exaggerates.) so I took it with a grain of salt.

It still concerns me that elitist attitudes will turn talented people away from Linux and thus inevitably undermine its growth potential. Almost all of the really cool things in and out of security are being done on Linux, and I want it to stay that way. So be nice to everyone! You don't know if you are insulting the next Alan Cox.

Re:Linux attitude towars newbies....

[macshit]

Naw, Eric is completely full of shit on this particular issue.

The LKML is actually amazingly meritocratic, and really quite friendly to new faces -- if you have your shit together, and have done your homework.

However, like many programming lists, there's little facade of politeness, and if they look at your proposal and see problems, they'll call you on them, and in no uncertain terms. This sort of thing unnerves many people, especially those (like Eric) that want to develop their own little fiefdom and have it all merged with no questions asked and no changes made. It ain't going to happen, people are going to go over your code with a fine tooth comb, and questions will be asked.

On balance, I'd say it's actually much better than other similar lists I've observed, and I think the reason comes down to one thing -- Linus. If he likes what you've done, he'll merge it, even if everybody else disagrees, and by-and-large he's pretty good at paying attention to principals rather than personalities. This has a rather astonishing levelling effect, and it seems to prevent exactly the sort of ossification that Eric complains about.

A few pictures of Hans...

[pen]

In case you wonder what he looks like, here are a few pictures of Hans giving a lecture over at LUGOD.

Re:A few pictures of Hans...

[Anonymous Coward]

He's kinda hot. Maybe somebody should have asked him if he's single for this interview. It might even start a new trend in "/. interviews as dating service".

Re:A few pictures of Hans...

[the morgawr]

Better picture here

Were these pictures taken at different times or what? These don't even look like the same guy.

Re:A few pictures of Hans...

[Anonymous Coward]

That site was updated in 1997 according to the footer. He's probably changed a lot in the past 6 years. Definitely for the better (what was up with that hair?).

Re:A few pictures of Hans...

[Anonymous Coward]

Maybe you can make a ReiserFS plugin of this...

Re:A few pictures of Hans...

[Anonymous Coward]

Is it just me, or does he kinda look like Steamboat Willy from Saving Private Ryan?

Totally agree.

[ashitaka]

I got flamed before for pointing out the flat, thoughtless answers submitted by previous respondents. (Mr. Shatner for example)

Typical justification where: "He's a busy man", "What did you expect deep insight?", etc.

Here we had someone who is no doubt busy but provided, in-depth, insightful and complete answers.

Re:Totally agree.

[boresicle]

Yeah, the only people that could defend Shatner's interview are hopless geek fans. There's nothing wrong with being a hopeless geek fan, mind you. But that's what was going on there. If he was too busy to answer properly his manager would have rejected the interview in the first place and he wouldn't have even heard about slashdot. This interview was great. I've never used ReiserFS, never even looked into it. I'll at least be going as far as looking into it now.

Re:Totally agree.

[Dr. Spork]

And don't forget to mention that his answers were written very well. I teach at the university level, and even humanities majors are rarely as articulate in their writing. I especially liked the "dignity of a foodfight" image.

Re:Totally agree.

[geekoid]

yes, but he is not william shatner. I mean what did you expect from Bill? after all, he did write a script where he fights God and wins.. ;)

Re:Totally agree.

[mepr]

Don't blame Mr. Shatner for lame answers given the dumb, crackhead and insulting questions that were delivered to him.

Re:Totally agree.

[Anonymous Coward]

I must say that one of the primary things I took away from this interview was an incessant awe at how Hans manages to communicate both his vision and methods.

Hans' work has always appealed to me because he seems to actually think about it, and not just do it. I'm sure that he's done many many versions of ReiserFS features that have failed (as he claims) but to have conceived of and attempted all those concepts is what impresses me.

His whitepapers have always been a good read, his future visions page is one of the first that I send to old Unix-heads considering using Linux and if you really want good reading, check out his and a few of his cronies' writings on the linux-kernel mailing list archives (look for Jedi w.r.t. Squid as well for interesting ReiserFS/Application interaction discussions).

Re:Totally agree.

[LeftOfCentre]

In addition, I think he is based in Germany and English wouldn't be his mother language. That makes it even more impressive.

Re:Licensing is a shame

[Ender Ryan]

Exactly! Despite the claims made by self-loving people like Bill Gates, the GPL is actually better for business than the BSD license. The GPL actually allows the author of the work more freedom to make money as (s)he sees fit. With the BSD license, everyone just "steals" it and never contributes anything back.

Microsoft et al just want a "free lunch". Well they can eat @$#%; if they don't want to share their work freely then they will damn well pay for my work if they want to use it. The GPL ensures that.

At the same time, anything licensed under the GPL can be freely looked at and reimplemented, so innovation isn't bottled up and kept secret à la proprietary software.

What about SpamFS from Hormel Systems?

[RimmerExperience]

Reiser4 seems pretty nice, but you still have to pay for the disk space in the first place.

That's why I'm excited about SpamFS - the first POP3/SMTP-based file system, leveraging the preponderance of free email accounts available on the internet.

Need more disk space? Sign up for some more email accounts. One advantage ReiserFS has is that it will work a lot better if you have an internet connection problem ... props on that.

However, Hormel Systems is really taking a revolutionary approach that I expect Reiser5, Longhorn, and OSX will be forced to incorporate.

Re:What about SpamFS from Hormel Systems?

[xchino]

There are alot (and I mean ALOT) of downfalls to using this as a type of fs. First and most obvious is the latency involved with network communication. If it takes as long to copy a divx movie from partition A on drive A to partition B on drive B as it did to download it, I don't really see this as being any more viable than sftp, or a basic VPN.

Also, what happens if one of your email addresses goes kaput? Was it housing important data? And since most free email providers give you like 5 megs of space, you'd need several hundred accounts before you had any sort of decent storage space. Of course, now you have to clean the spam out of several hundred mail boxes to acheive the level of data storage you were looking for. If a server where you have several email accounts goes down, your data will be nearly unretreivable. Your almost better off having a crashed HDD sent to a recovery lab.

It's an interesting and innovative idea, but I don't think it has a whole lot of real world implementations.

Re:What about SpamFS from Hormel Systems?

[bobv-pillars-net]

You sir, have just been (oh, so deliciously) trolled.

Re:What about SpamFS from Hormel Systems?

[justins]

Maybe you can make a ReiserFS plugin of this.

Re:What about SpamFS from Hormel Systems?

[nuntius]

Actually, it shouldn't be too hard and would be a neat exploit to show off...

Of course, you'd need to write an automated client, but that's where the fame comes from.

Since free email accounts are _free_, signing up for a thousand (~5 Gigs of space) doesn't cost anything. Since the only mail which belongs in these accounts is mail which you sent, removing spam would be easy. As for an account going down, use a RAID-style redundancy pattern to allow easy recovery. Splitting large files into smaller ones and md5'ing them are easily obtained functions.

The only hard thing I forsee is making sure the hosts dont (a) shut down your accounts, (b) filter your messages as being spam, and (c) get very pissed at you.

Viability is not the point - bragging rights are.

Re:What about SpamFS from Hormel Systems?

[Anonymous Coward]

Why would Slashdot be rude enough to allow this? Hormel has been very good about not causing a stink about using SPAM as a word to describe something evil, even though they do not think their product is evil. Posting a link to them is very bad form. Mod this down.

Re:What about SpamFS from Hormel Systems?

[Anonymous Coward]

You know, you keep talking, but all I can hear is "I'm a whiny bitch having her first period".

STFU Sophie.

Re:What about SpamFS from Hormel Systems?

[Anonymous Coward]

Fuck, even the TROLLS don't seem to understand the word troll any more.

There was no trolling going on - that post was just a joke, plain and simple. The guy who replied is simply humor impaired (while you're 'definition impaired').

Re:What about SpamFS from Hormel Systems?

[bobv-pillars-net]

Perhaps "baited" would have been a better term. Anyway, I greatly appreciated the joke.

Re:What about SpamFS from Hormel Systems?

[MyHair]

myhair@darkstar$ps

init
klogd
inetd
smtpd
pop3d
imapd
spamfsd
viagrad
celeb-photosd
college-degreed
nigeria -expatd
apache
bash

myhair@darkstar$

Hmmm, I think I have some configs to tweak.

Excellent interview

[Harbinjer]

Wow! What a great interview. Sometimes you get basically yes/no answers, but I'm really, really impressed with this one. I never really looked much into filesystems, but boy, this was interesting. I would say hurray for SuSe and Mandrake who use ReiserFS. Moreso for Mandrake because they are completely OSS.

It really seems like ReiserFS is going places. I wonder is there much going on in the developement of the other filysystems? (JFS, XFS, Ext3fs?). I haven't heard any long term plans from any other filesystem guys, but maybe its worth looking into. I'd be curious of other [filesystem] people are actually developing long term, or just fixing stuff.

I definitely agree with his point about long term research. You just can't do something in a couple weeks or even months. The really great achievements are usually done over a period of years, and as we (humans) learn more and more, the great discoveries will take longer and longer. While it does make sense for buisinesses to want short term research, the government should do things long term, as should the larger companies. I'm especially impressed with IBM, because you sometimes hear of fun stuff(writing IBM in atoms) that is definitely long term being done there, which is cool.

Again kudos to Hans for this great interview.

Re:Excellent interview

[jadavis]

ext3 can be mounted ext2, and ext2 can be converted to ext3 without having to move any large amount of data.

JFS is an IBM filesystem, which makes me think it serves a similar compatibility issue.

Compatibility is especially important in a filesystem, because it's so difficult to upgrade a filesystem.

I don't think that people make a lot of blank disks into JFS/Ext3 unless you have tools (i.e. recovery tools) that only work with JFS/EXT3. I used to use ext3, but then debian put reiserfs support on the 3.0 ISO so I could recover from whatever problems I caused without worrying about whether my boot disk could mount my root filesystem.

As for XFS, I think it's supposed to have better performance in some areas (mostly dealing with large files) and it's also been around for a while and is stable. So, it's supposed to be stable like ext2, but perform well and have journaling. Also, I'm sure SGI mainly did that so that their customers would have better access to their existing data.

You're right though I think: ReiserFS is taking over as a technically superior FS.

The $25 question...

[moriya]

"Is it done yet?"
"Is what done? Can you rephrase?"
"I can't... otherwise, you'd charge me for two questions."

Version Control

[phorm]

I'm wondering how exactly filesystem-based version-control would work? Would that mean that a versioned backup of each file is made, or have it optional as to which files are backed up.

The only version control I've used is MS SourceSafe (insert boos and hisses here, no I don't like it) and also when editing COBOL files on a VAX. Now, it seems to me that the versioning was a property of the VAX and not the editor (as it automatically incremented versions with up to 2 backups) regardless of editor, and/or implemented the newest version on an edit.

Re:Version Control

[waldo2020]

Actually, VMS kept around all versions of a file until you did a PURGE command on a directory. Accessing the file without the version got you the latest files. But imagine a log file.. Needless to say this was very wasteful of disk space.

Re:Version Control

[JoeBuck]

The most well-known filesystem-based version control system is Clearcase. Essentially, the user sets up a view, which has a config spec. The config spec specifies what versions of each directory and file the user wants to see. Checkins and checkouts are done as in CVS, except that directories and file renames are tracked as well.

Re:Version Control

[Skyshadow]

It's worth mentioning that both ClearCase and a couple of other version control systems (like Perforce) have had "second gen" features for a few years now.

Basically, it revolves around using the tools to encourage/compell best practices, hopefully avoiding a lot of the classic pitfalls that people encounter when they try using a version control system on a large-scale basis (hopelessly fragmented file trees, users conflicting with each other's checkouts unnecessarily, an overcomplexity in the use of the version selection tools, etc).

The big difference between those tools and the Reiser filesystem, at least as I understand it, is that Reiser is straight versioning, while the VC tools include powerful functionality like branching, aimed at allowing several users to work on several versions of the same file at the same time. This extra functionality is what brought around the need for second-gen version control practices.

Re:Version Control

[Gekko]

Actually it was configurable by the admin. The first VMS system I used kept 3, others 5.

Re:Version Control

[Anonymous Coward]

To clarify: It's admin configurable on a per-file (or per directory) basis. To let users have 2-3 versions of their files is reasonable, but you might want more for development files, or logs, or whatever.

Re:Version Control

[Anonymous Coward]

Record Management Service (RMS) defaults to keeping 32,767 versions. When you issue the commands:

create/directory/version_limit= n

or the command:

set directory/version_limit= n

you could specifiy a number n less than 32,767. It is also possible to limit the versions retained for an individual file with:

set file/version_limit= n

Rather few production VMS systems allow for the default of 32,767 to obtain for created directories (hence the preponderance of mkdir.com DCL wrappers around the $ CREATE/DIR command).

Re:Version Control

[hansreiser]

It should be optional which files are backed up. Now the interesting question is, what should the interface for that be? For instance, should only one version be maintained until you do the first commit, then two versions until you do the second commit?

Or, alternatively, should one specify that for a particular file every write creates a new version, or that every close creates a new version, or is there a place for all three approaches, depending on the user's needs?

Re:Version Control

[hachete]

It seems to me like there's "version control" for file-versioning (VMS - which would be great for automatic rollover of log-files) and there's version control for the big projects - a couple hundred+ branches, checkpoints, merge-control (merge control in Clear Case seems less than optional - particularly for the people forced to do it), parallel lines of development etc etc. Then on top of that, you've got distributed working, product release etc etc. Pushing that stuff down to the filesystem is definitely a Big Target and, quite probably, only *half* the job. Managing the complexity is, imo, the other half of Version Control.

I'm thinking, it would be nice if the hooks were in there for the larger legacy Source Code Control Systems. For example, something which would allow Webdav to interact at an API level with this funky file system.

h.

Reiser and the GPL

[lspd]

You do all understand that while the GPL doesn't permit tying by license, distros have now moved to using threats of invalidating support contracts to achieve the market leverage they need to exclude competitors, yes?

Although I agree with Reiser's statement that tying support contracts to a GPL product is/should be illegal, some of his other statements about the GPL have been questionable.

RMS responded to a question about some of Reiser's statements about the GPL v3 indicating that Reiser was incorrect about GPL3 including ad-removal restrictions.

The entire thread is an interesting read about the GFDL, GPL, and possible crossover between the two. Take a look at the author index for some other interesting tidbits.

Re:Reiser and the GPL

[Dunedain]

Whoah! The sort of tying which should violate the GPL already does: if I said "You must pay me for a support contract to get the source to this binary I'm giving you," that would violate the GPL. But there's nothing wrong at all with what RedHat's doing: they give you the binary, and they give you the source. Then they offer a separate deal, in which they'll support the software they gave you in exchange for money. They choose not to offer yet another deal, in which they'll support any other software for gobs of money.

How is this different from what, say, Debian does? They do not warrant that they will provide support; nevertheless, they'll try to help if you ask on their mailing lists. But they really can't help much with your Gentoo/RedHat/OpenBSD problems.

Re:Reiser and the GPL

[lspd]

But there's nothing wrong at all with what RedHat's doing: they give you the binary, and they give you the source. Then they offer a separate deal, in which they'll support the software they gave you in exchange for money.

I didn't mention RedHat, but if you take a look at the licensing terms you'll see that their support contract wraps around the GPL software by claiming that if you accept any support contract you give up the right to freely use, and copy the software on any machines you want. In fact, their terms claim that you can't use the software at all without a support contract, but that's patently false so it's not even worth worrying about.

Debian, Gentoo, Slack, etc offer community based support which does not require any seperate agreement. They also don't make claims that their trademarks must be removed if you install any non-distro specific software on your machine. Check out RedHat's trademark guidelines page for even more fun. If you compile a custom kernel on your RedHat machine you have to remove all the RedHat logo's and trademarks even if the changes are only for personal use.

Anyway, I don't think RedHat is any worse than...say...Mandrake or Lindows in wrapping GPL software with additional restrictions. I just hate to see companies trying to make money this way. Sooner or later the restrictions will go a bit too far and we'll see the house of cards come tumbling down. Some enterprising developer will decide to start suing everyone that's wrapping their GPL work with additional terms. The GPL does not allow for placing additional terms on end users. Any support contract, download agreement, EULA, trademark restrictions which restrict the rights listed in the GPL clearly cannot apply to GPL software. I'd hate to see that proven in a court of law since it will only serve as a nasty example for GPL+Business doubters.

Re:Licensing is a shame

[shaitand]

ok, since when can microsoft not afford to pay a "fee", create a proprietary FS, and release jack back to the community?

Re: What books to read

[Alderete]

Specific questions: what branches of math are useful in this line of research? Any books, articles, etc., that I haven't listed that are a 'must read' or 'should read'? Those who have succeeded in building a better filesystem: what have they done that I should also do? Any mistakes I should avoid? Anything that no one told you about filesystems that you wish you had known up front?

You might find Dominic Giampaolo's book about implementing the BeOS File System (BeFS) interesting:

Practical File System Design with the Be File System
by Dominic Giampaolo

* Paperback: 237 pages
* Publisher: Morgan Kaufmann (January 15, 1999)
* Average Customer Review: 4 out of 5 stars

tax $ - me = (:

[Anonymous Coward]

Actually I was a bit reassured as a taxpayer about how the DoD spends our money by seeing them in action

I'm not.

Re:Licensing is a shame

[jlrobins_uncc]

> Just look what happened (is happening) to Kerberos.

I'd say that Microsoft's 'extension' to Kerberos is ultimately perfectly fine and valid, just unfortunate that they did not make it completely valid to reimplement. They used a field in the ticket exactly how it was meant to be used -- for vendor-specific authorization data. Remember, Krb5 was designed to be an authentication protocol, and you can use either AD for authentication of plain-*NIX-boxen just fine or can use, say, MIT or Heimdal K5 for authentication of users on Win2K / XP client boxes w/o too much (relative) trouble.

AD's implementation of Krb5 genuinely strives to be per-spec, and the spec defines the authorization field to be used as the vendor sees fit.

Re:Licensing is a shame

[Blasphemy]

But by paying the "fee", they are giving back to the community. The developer is rewarded and is able to continue work on the free version too (assuming a benevolent[sp?] developer).

I see no problem with this.

that is the funniest thing

[Anonymous Coward]

I have read in a looooong time. O' to be J-Dog for a day.

B man

[cheeseSource]

Now if you think about it, who wants to be around a blind man with a stick,

I am a blind man without a stick. I swing code around instead...

The answers were excellent though.

Re:Licensing is a shame

[Blasphemy]

From what I understand, they are using the vendor specific field to ensure that windows hosts can't be authenticated by unix servers. I don't believe this is what the developers intended.

IMO it's anti-competitive and illegal, but IANAL.

On the other hand

[roystgnr]

Is this guy working for SCO, or what?

I think by "copy" he meant "duplicate the functionality and semantics."

Sadly, based on their mishmash of contradictory claims, I think that's what SCO means by "copy" too.

Since he mentions Clearcase...

[Skyshadow]

...I would imagine that he pictures versions being kept until explicitly dealt with.

This isn't actually as big a deal as it might sound -- really, at least from a user level (as opposed to system), most people have a ton of static files and a few that are updated with any regularity. Assuming that the evolving versions are stored in a halfway intelligent manner (such as ClearCase's file deltas), the amount of "extra" space such a filesystem would consume should be reasonable when put in the context of the rate of size expansion in modern drives.

If they can create a (much) more efficiant version of ClearCase's MVFS, it would obviously be tremendously valuable on several fronts other than simple backups, such as tracking overall FS changes, metrics gathering (from a bid'ness client POV), etc.

Re:Licensing is a shame

[Ded Bob]

I believe Microsoft did not use the MIT-licensed code for their version of Kerberos: post

It is interesting that Microsoft wrote their own Kerberos even when code already existed with a BSD-style license.

This prevents M$ from stealing his work, ...

With the direction Microsoft is taking with file systems, I strongly doubt they will want his work.

#1 is THE answer

[jafac]

The question, much asked, is;
Why does software SUCK?

The answer is Commercial Constraints on the development process.
Yes - business pays the bills. But rarely looks to the future, in favor of making a quick buck.

Can purely academic software suck? Yes. Ivory-towerism leads to a steadfast refusal to meet user requirements.

Is there a middle-ground approach that might work?
Certainly.
But when it becomes a contest of egos between an Academic Founder of a company, and the MBA Suits to prove who's "better" or who has the "power" - guess who wins out more often than not? The middle ground suffers. After all, we're only human, aren't we?

Re:#1 is THE answer

[dasmegabyte]

Actually, when I look at the software I find most satisfying, nearly all of it is software I paid for. My top three favorites are: iTunes, .NET studio and Safari.

Of course, what's interesting is that they're all three of them TINY PARTS of a larger system. Meaning that somewhere in MS and Apple, there's a group whose developments are considered important enough to the final product to exist, but not essential enough to sell and market on their own (possible exception being the new iTunes, but I don't like the newest release as much as 3). This probably means that, in order to justify their existance, these divisions have had to constantly reinvent themselves to come up with new ways to make their software brilliant. They can't really be fired because they're supporting a core product. New development is therefore limited to stuff that's cool enough and stable enough to release.

I've noticed in the past that when a developer becomes a "fixture" of a company -- that is, his work is so deeply entwined in the company's success and is so poorly understood by others -- that he basically works on what pleases him in between fixing time critical bugs that could bring the company crashing to its knees. I expect this is what's going on with these.

I felt the same way about WinAmp as well, but AOL's soul crushing corporate environment did away with all the good innovation coming out of Nullsoft.

Debian

[Anonymous Coward]

My question is, why does Debian consider ReiserFS to be a "less mature, less stable" filesystem than EXT3? ReiserFS was stable in the kernel long before EXT3 had the "experimental" tag removed.

Re:Debian

[yanestra]

My question is, why does Debian consider ReiserFS to be a "less mature, less stable" filesystem than EXT3? ReiserFS was stable in the kernel long before EXT3 had the "experimental" tag removed.

There seem to be different philosophies in the use of the "experimental" tag. In fact, ReiserFS kept to be fragile under certain circumstance until about 2.4.16, much later then when it was officially incorporated into the Linux kernel.

Re:Debian

[rweir]

What are you babbling on about? The bf2.4 install disks (based on kernel 2.4.18) includes *both* ext3 and reiserfs support. When you ask it to create a new filesystem, they BOTH show up as option, and you can choose whichever you want.

Indeed, the 2.2-based disks don't, but they don't include ReiserFS either.

So, are you a troll or a moron?

Re:Debian

[mepr]

duh... If you read the text of the options given on the installer, you'll notice that it makes the exact suggestion that reiserfs is less mature than ext3

Re:Licensing is a shame

[tjansen]

Sure. And Reiser would be helping commercial OSs to adopt it and customize it for their needs without giving anything back. Why is this desirable? MS integrated Kerberos. What did MS give back? Nothing afaik. Releasing stuff under a BSD-like license gives a short-term benefit to society: it makes closed-source software cheaper. But the long-term disadvantage is that this diminishes several competitive advantages of free software, especially the lower cost.

What about support tiers?

[emil]

Look at RedHat's hardware compatibility lists... They have tier 1, 2, 3 hardware. RedHat itself has tacitly stated that their support will be less accurate if you are not using tier 1 hardware. The users always have a role to play in the odds of a successful support call.

I am happy that RedHat made a profit this quarter, and I wish them well.

However, I think that their changes to support will hurt them in a number of ways:

• They are placing restrictions on the distribution and use of GPL software, which will eventually annoy the FSF.
• Linux has a moderately good reputation for security. Their actions with up2date threaten to ruin this reputation; they have taken their own user base as hostages.
• Keeping a RedHat system patched and secure is now more inconvenient.
• Their user base is now expected to pay more for less, which is corporate philosophy at its worst.

Because of these problems, I have been dumping RedHat from my customers' systems, and replacing it mostly with OpenBSD. While the necessity of compiling from source for all security updates is in itself inconvenient, there is 1/10 of the traffic in OpenBSD errata, so there is less work in the end.

p.s. What RedHat really ought to do is package a Linux kernel with the OpenBSD userland, and sell this as RedHat/Secure BSD Edition. Run this for two years and see how the takeup and errata rate compares with the standard and advanced server product. Yes, charge $40 for it.

Re:What about support tiers?

[araemo]

p.s. What RedHat really ought to do is package a Linux kernel with the OpenBSD userland, and sell this as RedHat/Secure BSD Edition. Run this for two years and see how the takeup and errata rate compares with the standard and advanced server product. Yes, charge $40 for it.

Would this even work? Doesn't OpenBSD alter their userland apps, like apache, to use their more secure system calls? Even if they don't alter it.. you wouldn't have the more secure system calls.. it'd be like putting a bigass padlock on a door made of rice paper...(Ok, not exactly THAT weak, but you get the idea.)

Learned different

[Jerf]

Sure, I've "learned" more out of class. It's almost impossible for class to keep up with a motivated learner; it's only a handful of hours a week.

But I've learned different getting my grad degree. I would not have my familiarity with graph theory, algorithmic complexity, internals of compilers, etc. While perhaps none of these things were as in depth as you'd need to actually do work, the framework I've acquired to handle these in two years in grad school would easily have taken me 10 years of my own to learn... because on your own you get very focused on the task(s) at hand and never learn anything more.

Also, unless you're a rare genius with math, having accountability to a professor to really learn the material, and not just skim a book and fool myself into thinking I understood intensely mathematical material, is invaluable. It's a rare person that can truly force themselves to learn material like that... and for things like graph theory that can be important.

Each of those things is paying off, too, in my work. Graph theory in particular, though it's hard to point at a useless class.

(Of course, if you go into the classes assuming that you'll never get anything useful out of them, you won't.)

Note that I did a Masters, and I did not do a thesis; IMHO two years is too small for a thesis, so I actively chose to take the classes instead, which were more valuable to me. (I feel like this was my thesis, since I was writing it the whole time I was in grad school, but they'd never give me a masters in computer science for that.)

Re:Learned different

[Anonymous Coward]

But I've learned different

Too bad, apparently, you didn't learn about adverbs.

Migration and maturity.

[emil]

I seem to remember two reasons offered by RedHat for resistance to ReiserFS:

• ext2 filesystems can be easily upgraded to ext3
• The fsck_ext(2|3) userland utility is very mature, and anything beyond a journal replay under ReiserFS has some probably of a corrupt filesystem (which may have improved in the years since this claim was made)

I think that Bero made these arguments here on slashdot sometime before he and RedHat parted company. Still, there is no excuse for refusing the tremendous gift of XFS from SGI.

Don't take it as two data points

[Anonymous Coward]

You were right to dismiss Eric. On the other hand, Hans may have a point.

I didn't follow Hans' experiences, but he's a technically competent guy. Eric, in his complaints didn't. I followed Eric's work, and his configuration tool was poorly designed, atrociously implemented, and generally technically incompetent. He was unable or unwilling to play by the rules by which features were submitted into the kernel. As a result, there was no way his code would ever make it into the Linux kernel. This is very different from the opposition Hans met, when he came in with a technically solid bit of code (which, as a result, Linus eventually included).

Unlike Hans, Eric's goal was not to improve the kernel, but to get his name on some reputable software project, to lend credibility to his primary job as the self-proclaimed open source/Linux spokesperson. As a result of this, and previous issues I won't bring up, a lot of people on the list also disliked Eric. As a result, when they excluded his code, they also gave him shit about it. But this isn't general to all newcomers; just Eric.

Is Plan 9 free?

[yerricde]

Grandparent wanted free software. Plan 9 doesn't count. It prohibits private modification. It requires people outside the USA to obey USA export controls. It seems to prohibit distributing some component programs without distributing the whole thing.

In addition, the license's retaliation clause effectively grants all Plan 9 contributors a license to every copyright that a user owns. For example, if an employee of The Walt Disney Company were to install Plan 9, Lucent would be allowed to hold a Disney DVD copying party. This may seem like a good thing at first, but imagine this: If an employee of the Free Software Foundation were to install Plan 9, Lucent would be able to distribute every GNU program as proprietary software.

Re:Is Plan 9 free?

[pyr0]

Hmm...I could have sworn I just saw a story on the front page of slashdot yesterday stating that plan 9 is now officially distributed under and OSI approved license.

Re:Licensing is a shame

[Anonymous Coward]

MS integrated Kerberos. What did MS give back? Nothing afaik. Releasing stuff under a BSD-like license gives a short-term benefit to society: it makes closed-source software cheaper. But the long-term disadvantage is that this diminishes several competitive advantages of free software, especially the lower cost.


Could you please post examples of things that you have done to "give back"? Your website lists no software contributions that you have made, nor does it supply any source code...

Re:Licensing is a shame

[anthony_dipierro]

Releasing stuff under a BSD-like license gives a short-term benefit to society

So release stuff under the BSD license and the GPL! Then you get a long-term benefit and a short-term benefit, right?

Re:Licensing is a shame

[Anonymous Coward]

This prevents M$ from stealing his work, making it incompatible with the main branch and calling it WinFS.

Why is that a bad thing?

Re:Licensing is a shame

[tjansen]

scan the KDE CVS for stuff by tjansen...

Re:Licensing is a shame

[tjansen]

Of course not. If you release under BSD anybody can take your source to improve closed software. GPL doesnt matter then, BSD is GPL-compatible (so anybody is allowed to re-license BSD to GPL).

Re:Licensing is a shame

[anthony_dipierro]

So wait a second. Giving something away is a bad thing? If I write something and let others use it that's bad, because some of those people might use it for bad purposes? I don't buy it.

If you release under BSD anybody can take your source to improve closed software.

Or they could take it to improve open source software which just happens to not be GPLed.

Re:Linux attitude towars newbies.... ;-)

[Puu]

This piece in Hans' answer puzzled me:

It is all very fine to discuss the sociology of herd formation, exclusion, and prejudice in the abstract, but one should never say that particular persons are making particular decisions on the basis of their herd instincts unless one wants to be truly hated by them and all of their numerous friends, and this was my mistake over and above the choice of what sub- herd to be part of.

And all along I thought it was spelled "the HURD"!

Re:Licensing is a shame

[jlrobins_uncc]

No, non-AD Krb5 servers can be used for authentication of users on Win2k / XP boxes just fine. Just won't be able to emit the AD-specific user *authorization* data -- essentially the network-group/role memberships ala the NTLM doman-wide group membership info.

You can sortof work around that using machine-local groups.

Check out this paper from M$FT [microsoft.com) for more info. This was helpful in getting my old uni's Windows XP realm to defer to the exising MIT K5 for authentication (go UNCC Mosaic!)

Attitude

[ralphclark]

We should all keep in mind though that there aren't any hard core greedy evil people in our industry. They are all basically good hearted people who chose trying to create a better society as their life's work at a substantial cost in personal income. Petty, bickering, overly impressed with ourselves, flaming, yes that describes most of us Linux kernel developers, but there isn't enough money floating around to attract any genuinely bad folks into our industry.

Ha ha, bravo, bravo! I like this guy, he seems like a genuinely good person. If only more people had his attitude the world would be a better place.

Re:Licensing is a shame

[abigor]

The guy wrote KDE desktop sharing, for one thing, which I use all the time at work. Thanks, Tim Jansen!

How about you, AC?

Re:Licensing is a shame

[geekoid]

Why does that prevent MS, or anybody, from stealing it? I mean, if they are willing to steal it when it is available for free, why would they stop because there is a fee attached to it?

I used MS because the post I'm replying to did, I am not implyngthat thay have or have not stolen code. Naturally by stolen I mean infringed on a copy right.

Hans sez

[Arslan ibn Da'ud]

"I am not a genius, I am just never satisfied and very very persistent."

Glad I read that article! Now I have a new quote for my sig!

Re:Licensing is a shame

[tjansen]

> So wait a second. Giving something away is a bad
> thing?

It's not bad, but your contribution to change the world is not as big as it could be. If you put software under GPL, only the free software community can use it. If you put software under BSD, you give it to vendors of proprietary applications as well.

So if your goal is to help free software to replace proprietary software (and thus to lead to a world where the software infrastructure is free), the GPL is the right choice. Putting software under BSD won't help you to achieve this goal. It has other effects, like decreasing the value of software in general and providing *one* free implementation. But only software under GPL, especially if there is no proprietary equvalent, gives the free software world a technical advantage over the proprietary competition.

Version Control and other attributes

[hughk]

The VMS file system allowed you to have up to 32767 versions hangin around. If you created a file (which the editor always did by default), the system would create a new version automatically. Version numbers could be limited by directory so only the last n versions are kept. The problem is that 10 versions of a file wiould take a lot of space.

Diff based versioning allows older versions to be kept as a series of differences from the current version. This is compact, but I feel it is pushing too much down to the file system level. A file system should just store data and provide ways of tagging the data to facilitate retrieval (whether directories or whatever).

For real revision control across a project on VMS, there was something called CMS (a little like SCCS/RCS). This would only store diffences but allow you to form groups of elements and enter version sets into classes. It cost money but was very good and could be used via the command line or a GUI. What was nice was the whole thing could be embedded inside scripts or called from other packages so you could really roll your own.

What was really nice on VMS was RMS, the record management system that sat on top of the file system. It meant that all programs could share files, even indexed files. Then there were the use of logical names and search lists.

If I defined a logical, SRC to be a comma separated list of directories, it would work like the Unix Path and the system would look through each directory specified until it found the file. Typically, I would have the working directory, staging area and the base level directory in the list so if something wasn't checked out, it would be automatically pulled from the base level. By redefining the logical, I can switch projects and environments easily.

The thing is that this wasn't the file system and it was an optional layer but it came as a standard system library and all the normal opens went through this layer. However only relational databases tended to bypass it (they did their own locking and had their own ideas about internal file structure). As with the underlying I/O subsystem, this layer above could be used totally asynchronously.

Personally whenever I use a Unix system, it isn't so much the underlying file system that becomes an issue so much as the layer above it. What is inside the file and how I can describe those contents (file metadata or resouce forks). When I open a VMS file, I immediately know things like max record size, blocking factor and so on. I don't always need the in fo, but I can really speed things up when I do.

Re:Licensing is a shame

[Aapje]

That's a load of bull. The GPL and the BSD license have different liberties and obligations. In some cases the GPL will be a superior choice and in other cases it will be the BSD license (or other non-Free licenses). But you shouldn't equate the (relative) lack of restrictions in the BSD license with leeching/stealing. History has shown that many individuals and companies are willing to contribute to BSD/MIT/Apache licensed code. People will contribute back, if only out of self-interest (merging custom code costs $$$ and that code will not be tested (for free!) by other users).

Furthermore, the BSD license allows companies and individuals that want to contribute to use code without necessarily having to open source all other code that touches the open source stuff. This is important to companies if only to prevent accidental GPL violations. You also get into trouble if you want to assemble and distribute a system with closed and open source components (you might want to create the best of breed solution). IMHO, a more liberal approach is often better than forcing people to comply to fairly arbitrary rules.

Last and most importantly, my interest is in promoting the reuse of code, resulting in better software. If someone wants to leech, that's fine with me. I believe that open source is one of the strongest methods of cooperation between companies. In the long term, the companies that leech will be at a disadvantage to those that actively cooperate and share knowledge.

PS. I'm getting sick and tired of the "M$ will steal my work" argument that gets thrown around everytime the BSD license is mentioned. It seems to me that MS goes out of its way not to reuse existing open source code (resulting in mediocre products). I understand that you want to use the bogeyman of the open source community for an easy ad hominem attack, but its really not much better than those Hitler references that supposedly end a Usenet thread. FUD is easy, insightful comments are challenging. I dare you ;)

So it's closer to free

[yerricde]

I missed that story. Apparently, either RMS missed it as well, or he's busy.

The revised license looks much better, but it still has a couple problems. For one thing, it still requires adherence to the US export controls, which can approach extraterritoriality. Changing it to refer to something along the lines of "the export control laws of Distributor's jurisdiction" may have served better. And haven't the US laws changed so that export of open-source crypto (except to Cuba, Libya, etc.) doesn't need a license?

This part worries me: "Each party waives its rights to a jury trial in any resulting litigation."

Other questions are more practical: 1. Does Plan 9 support computer hardware commonly available at Best Buy? 2. Does Plan 9 support enough widely used desktop software to make migration from Windows or Linux not a pain?

Don't forget Slackware!

[SaDan]

They've had ReiserFS support at install time for a while now.

Re:Licensing is a shame

[shaitand]

Nope, nothing wrong with it at all... so long as I haven't contributed code.

Version control in filesystems

[Anonymous Coward]

"I predict that in 20 years version control in filesystems will be standard and expected by all users as a basic feature."

Try these numbers:

4-5 years in MS XP
15-20 years in FreeBSD
2-3 years in ReiserFS

Re:Licensing is a shame

[anthony_dipierro]

It's not bad, but your contribution to change the world is not as big as it could be.

I disagree. You contribution to change the world is bigger if you give away your contribution to more people, not fewer.

If you put software under GPL, only the free software community can use it.

Only part of the free software community can use it.

If you put software under BSD, you give it to vendors of proprietary applications as well.

Why is this a bad thing?

So if your goal is to help free software to replace proprietary software (and thus to lead to a world where the software infrastructure is free), the GPL is the right choice.

But what if I don't think people should be forced to release their source code? Then the GPL is the wrong choice.

Putting software under BSD won't help you to achieve this goal.

Sure it will.

It has other effects, like decreasing the value of software in general and providing *one* free implementation.

Why only *one* free implementation? The BSD license allows all the free implementations that the GPL allows, plus it allows other free implementations that the GPL doesn't allow.

But only software under GPL, especially if there is no proprietary equvalent, gives the free software world a technical advantage over the proprietary competition.

The free software world always has a technical advantage over the proprietary competition. That's why we want free software in the first place, remember?

Re:Licensing is a shame

[Ender Ryan]

PS. I'm getting sick and tired of the "M$ will steal my work" argument that gets thrown around everytime the BSD license is mentioned. It seems to me that MS goes out of its way not to reuse existing open source code (resulting in mediocre products). I understand that you want to use the bogeyman of the open source community for an easy ad hominem attack, but its really not much better than those Hitler references that supposedly end a Usenet thread. FUD is easy, insightful comments are challenging. I dare you ;)

Oh please! What kind of garbage is this? I was taking you seriously until this last paragraph.

I did not resort to ad hominem attack; Microsoft has stated several times that they prefer the BSD license because they can use your code, whereas with the GPL they can't. In a business case, why would I want to have a one way street flowing towards my competitors?

BTW, I was not attacking the BSD license. I have no problem with it at all and have contributed BSD code and have released work under similar licenses, but I was making a case for business.

Granted, there has been excellent collaboration with projects like Apache, and that's great. I'm not telling you to use the GPL, just pointing out that it allows for doing business in ways the BSD license can't. Use what you like.

RedHat reiserfs support

[spacemky]

I've seen a couple people commenting on RedHat's support for reiserfs. To install a redhat box on reiserfs do this:

1) boot from RedHat cd/floppy/whatever
2) at the boot prompt, type: linux reiserfs

That's it. Many people think that RedHat doesn't support reiserfs befause it isn't available by default.

Re:Licensing is a shame

[tjansen]

I disagree. You contribution to change the world is bigger if you give away your contribution to more people, not fewer.

If your focus is to let other people use your software, yes. If your focus is that all people have access to their software's source code, no.

But what if I don't think people should be forced to release their source code? Then the GPL is the wrong choice.

Sure. If you don't care whether a few proprietary vendors control all information, then BSD is certainly the right license for you.

The free software world always has a technical advantage over the proprietary competition.

No, it doesn't. There are many area where the proprietary competition is still years ahead. For example high-end scalability (compare Solaris to Linux on >=64 CPU machines), VMs (compare MS's .NET CLR or Sun's Java VM to Mono, Portable.Net, Kaffee and all the others), usability (compare MacOS to KDE or Gnome)...

Re:Licensing is a shame

[Aapje]

I can't help but read "MS will steal your work if you use the BSD license" in your post. IMHO that is FUD. MS might have stated that they prefer the BSD license, but I'm quite sure that it is an attempt to split the open source community. DotNET for BSD can be seen in the same light. It's just a weak attempt to lure people away from their main competitor, Linux. It's pretty clear that they aren't interested in leeching from the BSD community, simply because they don't. They much prefer to either write everything themselves or to buy a company that already has a half-decent product.

In short, your statement is not supported by the facts and puts the BSD license is a bad daylight. In my book, that is FUD. You might not have intended it, but that doesn't change the fact that your post will be interpreted this way by some people. If I'm misunderstood, I usually reread my post and see if I can express myself clearer in the future. There is always something to improve and the best way to find it is to answer the question why I'm misunderstood.

Re:Licensing is a shame

[Ender Ryan]

You're reading things that are not there. I'm not particularly worried about MS, just using them as an example, there are sure to be plenty other companies that will, and do, take code and never contribute back in any way. In a business sense, that could be thought of as bad, and the GPL is a way to prevent that.

What have I said that is not fact?

1. The BSD license allows companies to use code in proprietary products without ever contributing back.

2. The GPL does not allow that. A project, such as ReiserFS, can be GPLed and sold to proprietary vendors at the same time. The proprietary vendor can either purchase it, or use the GPL version and open their own code.

Look, I'm not saying that the BSD license is bad for business, just that the GPL has certain advantages. If your business is based on support, then in that case the license may not make much difference. But as is the case with MySQL and ReiserFS, their businesses may not work if their code was BSD licensed.

Whatever... I'm not going to argue anymore.

Cheers.

Re:Licensing is a shame

[anthony_dipierro]

Sure. If you don't care whether a few proprietary vendors control all information, then BSD is certainly the right license for you.

I didn't say anything about proprietary vendors. I said vendors who don't release their source code. There's a big difference.

Re:Licensing is a shame

[Aapje]

What have I said that is not fact?

"With the BSD license, everyone just "steals" it and never contributes anything back."
"Microsoft et al just want a "free lunch"."

The first is patently false since companies do contribute back. The second is arguably false since MS rarely (if ever) uses BSD software and because there are good business reasons for companies to dislike the GPL, even if they want to contribute.

If you can't understand the difference between "it is possible that" and "everyone will", you have a big problem communicating. You can keep wondering why you get flamed, but it might be more productive if you learn to distinguish between having the freedom to do something and actually doing it.

Re:Licensing is a shame

[vadim_t]

The second is arguably false since MS rarely (if ever) uses BSD software

root@alice root # cd /mnt/w2k/WINNT/system32
root@alice system32 # strings ftp.exe | grep -i regents
@(#) Copyright (c) 1983 The Regents of the University of California.

There you go. BSD code in Windows 2000.

Re:Licensing is a shame

[Aapje]

I know. They bought their TCP/IP stack from a third party who used some BSD code for related utilities. Note that the TCP/IP stack in Windows is not derived from the BSD version, even when that was the best one around and MS needed one badly. So my point still stands, they seem to go out of their way not to use open source solutions.

Re:Licensing is a shame

[k98sven]

Releasing stuff under a BSD-like license gives a short-term benefit to society: it makes closed-source software cheaper.

Wrong. It makes closed-source software development cheaper!