| |||||||||||||
Bruce Perens: the Linux colonel talks (vnunet)Bruce Perens: the Linux colonel talks (vnunet)Posted Sep 9, 2004 0:17 UTC (Thu) by hppnq (guest, #14462)In reply to: Bruce Perens: the Linux colonel talks (vnunet) by BrucePerens Parent article: Bruce Perens: the Linux colonel talks (vnunet)
A lot of folks in the IT world only speak what marketing communications puts in their mouths. You can expect more plain talk from me, and when I think there's a problem that needs fixing, you can expect me to lay it at the doors of the people who could fix it. I'll talk about it in front of the rest of the community too, so that maybe someone there will solve the problem. There's nothing wrong with plain talk. I'm all for that. ;-)
Regarding RH and SuSE, I had talk with HSBC, the huge international bank, a few years ago, after RH went to its Enterprise Linux marketing. HSBC had at that time decided to place a Linux migration on the back burner, because they felt that now, Linux is as expensive as Windows. Ignoring the complaints of big potential customers is not a good way to promote Linux. In the interview you explicitly relate the success of Linux to the Open Source development process. I couldn't agree with you more. RedHat has managed to build a business on top of that process, while at the same time supporting it in a way that, I think, has yet to be rivalled. So, by bluntly dismissing their rather succesful business model, instead of pointing out possible flaws in the Big Plan, you are running the risk of being suspected of promoting your own business at the cost of your competitors. Which doesn't contribute anything to widespread adoption of Linux, or Open Source in general. The HSBC example is of course a quite weak one, for one because of an inconsistency that creeps up everywhere in the interview too: apparently RedHat is just not aiming for early adopters, so why would you or potential customers hold this against them? You realise of course that the two kinds of support you are (or seem to be) talking about are completely different beasts, each with their own organisational and financial requirements. You also seem to suggest that HSBC was in for a big migration. I am curious as to what exactly they had in mind regarding support, since it is highly unusual for huge companies to not entertain long term support contracts. When they said "Linux is as expensive as Windows", I hope and trust you said: "I'm very sorry, but you seem to miss the point. RedHat might be expensive, but Linux is not. Not in beer, nor in freedom." (I am just taking RedHat as an example here, of course. You can substitute any Linux vendor you like, although I do think that RedHat is ahead of others in some respects. I am not associated with them in any way -- in fact, I work for a top 10 global financial player and am involved in Unix management.)
(Log in to post comments)
Bruce Perens: the Linux colonel talks (vnunet) Posted Sep 9, 2004 1:03 UTC (Thu) by BrucePerens (guest, #2510) [Link] RedHat has managed to build a business on top of that process, while at the same time supporting it in a way that, I think, has yet to be rivalled.I am not convinced that the Red Hat model has long-term viability. The economics of Open Source work very worst for distribution creators, a fact that has become extremely clear to Red Hat. Thus, they have had to revert to a "bondage" model in which suport is bound to a subscription per seat and draconian things happen if you violate the subscription terms, for example, by putting the software on more systems than you've subscribed for or divulging support data to the general community. Of course, that sort of bondage flies in the face of all of the examples we claim for Open Source - it really makes a lie of it unless you can provide a viable alternative to Red Hat. I feel the Red Hat model will eventually fail because 1) everybody knows the software can be had for zero cost elsewhere, and once those sources have the necessary support, they'll go there and 2) the model decouples the service rendered (support) from what is being paid for (a software seat) in a way that simply cries out for a competitor to set it right Even Red Hat acknowledges this, or they would not have attempted to create a poor copy of Debian called Fedora. Fedora is supposed to be "Free Enough" but the certification and support is still with the virtually-proprietary Enterprise Server product and the governance of Fedora is firmly under RH management's direction. I don't believe that the customer has to settle for "Free Enough" in order to get excellent enterprise support. I propose to reform the support model. We shall do this by acknowledging that the software is not a profit center for anyone, and by giving away the supported and fully certified version of the software at a zero per-seat cost. We will develop that software in the public interest, within a democratic organization (Debian) that has already been established as a legal non-profit. Then, we will provide an organization of independent commercial support providers who compete with each other in providing service or differentiate from each other by specializing in a field or region, while cooperating on the overall system that makes all of their individual businesses possible. I submit that the economic benefit of competition and cooperation in the right places is of sufficient benefit to the customer that business shall be drawn from other distributions. The drive to differentiate among service providers will result in our eventually covering all bases of customer, large and small, many different vertical markets, etc. while the collaborative center in Debian will keep all of this together technologically. An ambitious vision, yes. Time will tell. Meanwhile I know where to start, which is not with the biggest customers. Sure, I told HSBC what the real score was, but I didn't have a boxed product with ready support to go with that tale. Now, I will have that. Thanks Bruce
Bruce Perens: the Linux colonel talks (vnunet) Posted Sep 9, 2004 1:05 UTC (Thu) by BrucePerens (guest, #2510) [Link] Of course, that sort of bondage flies in the face of all of the examples we claim for Open SourceOops. advantages, not examples.
Bruce Perens: the Linux colonel talks (vnunet) Posted Sep 9, 2004 2:41 UTC (Thu) by hppnq (guest, #14462) [Link] First of all: had you put it a bit more like this in the interview, I probably wouldn't have posted my original remarks at all. I'd also like to mention that I do appreciate your efforts in the promotion of Open Source. ;-)However.
... a subscription per seat and draconian things happen if you violate the subscription terms, for example, by putting the software on more systems than you've subscribed for ... If this is not a misinterpretation of the actual subscription license, then somebody had better sue RedHat now. Without knowing the actual text, I suppose the last bit should read something like: "by allowing more systems to benefit from the subscription than you have paid for". Which would be perfectly legal, of course.
I feel the Red Hat model will eventually fail because 1) everybody knows the software can be had for zero cost elsewhere, and once those sources have the necessary support, they'll go there and 2) the model decouples the service rendered (support) from what is being paid for (a software seat) in a way that simply cries out for a competitor to set it right Both points are common practice, and no vendor seems to be suffering from it. The conclusion in 1) is too easy: of course there will be some form of competition, but RedHat will not automatically lose customers. 2) is just the unit of IT support, if you like; although one might argue about its usefulness, the number of seats (or CPUs) is nothing more than a way to quantise an amount of support.
Fedora is supposed to be "Free Enough" but the certification and support is still with the virtually-proprietary Enterprise Server product and the governance of Fedora is firmly under RH management's direction. Correct me if I'm wrong, but I think Fedora is a testbed for RedHat's Enterprise products -- it has community support, it has no certification. Its goals are stated clearly on the Fedora webpage. Sure, one might argue that RedHat envisioned and promised a bigger, more active role for its community than currently is the case, but this is exactly why Open Source will win in the end: you can always pick up the pieces and make a better product; you don't have to whine about it. RedHat's involvement in Fedora is not fundamentally different than things are in the Debian or SuSE development processes. There is no evil plan, you might call it quality control if you like. Besides, I very much appreciate RedHat's clear stance on the GPL and the philosophy of Open Source (resulting in the most GPL conforming major distribution), whereas the Debian solution is a half-hearted attempt to ignore the problem of non-free software by putting it just outside of the main tree where it is still distributed. (Boy, am I going to be flamed for that. ;-) In other words, your description of "Free Enough" doesn't accurately describe RedHat's adherence to the GPL. Nor does it make any sense (it is in fact misleading) to switch from a description of Fedora to a product that requires "excellent enterprise support". As for your plans with Debian (what happened to UserLinux?): I sincerely wish you good luck and I am looking forward to seeing how things will unravel. Your project is ambitious, if not IMHO naive on the independency front, but sympathetic none the less!
Bruce Perens: the Linux colonel talks (vnunet) Posted Sep 9, 2004 4:41 UTC (Thu) by BrucePerens (guest, #2510) [Link] First of all: had you put it a bit more like this in the interview, I probably wouldn't have posted my original remarks at all.I did talk with the reporter for two solid hours. I don't control which bits get in the article, but conflict makes for an exciting story and those bits are most likely to get in. I'd also like to mention that I do appreciate your efforts in the promotion of Open Source. ;-) Thanks! Without knowing the actual text, I suppose the last bit should read something like: "by allowing more systems to benefit from the subscription than you have paid for". Which would be perfectly legal, of course. Using more than one attorney, various non-profit and commercial entities have determined that the RH enterprise business model is within the letter but not the spirit of the GPL. Whereas the Debian solution is a half-hearted attempt to ignore the problem of non-free software by putting it just outside of the main tree where it is still distributed. I do support splitting non-free off of the Debian organization into its own entity. RH avoids the problem of having to deal with non-free software by not attempting to put everything in one repository. The non-free stuff exists, it's just elsewhere. But that introduces all sorts of dependency hell and we don't want to go there. Nor does it make any sense (it is in fact misleading) to switch from a description of Fedora to a product that requires "excellent enterprise support". I think Fedora was intended as "you make the distribution for us, enthusiasts, and we'll let you have the non-certified version for free and keep the certified one for ourselves". It didn't get as much willing outside participation as they wanted, which is right. As for your plans with Debian (what happened to UserLinux?) UserLinux is Debian with support and certification added. The stuff that belongs in a non-profit is in Debian. Which is pretty much all development. Support is outside in an organization of commercial providers. Thanks Bruce
Bruce Perens: the Linux colonel talks (vnunet) Posted Sep 9, 2004 5:54 UTC (Thu) by hppnq (guest, #14462) [Link] The non-free stuff exists, it's just elsewhere. Now you're pulling my leg, right?! ;-)
I think Fedora was intended as "you make the distribution for us, enthusiasts, and we'll let you have the non-certified version for free and keep the certified one for ourselves". It didn't get as much willing outside participation as they wanted, which is right. I really don't want to come across as a RedHat zealot -- but why the funny angle? To me, taken at face value, it sounds like such a great and innovative idea: a true cooperation between the community and a company, together creating a unique win-win situation. Like I said, the objectives are quite clear. Now, I am as cynical as the next guy, but I like a fair share of realism as well. Everybody has an agenda, and I am glad that Fedora's is out in the open. Of course you can always make it sound bad like you do here, it's just so easy. If you first suggest that your Joe Home User OS really should be certified (against what?!) and then complain that it isn't, and that you are in fact dealing with an inferior product, you are making up your own rules and you always win. So, cut the crap and give us the real stuff. ;-)
Bruce Perens: the Linux colonel talks (vnunet) Posted Sep 9, 2004 7:26 UTC (Thu) by gvy (guest, #11981) [Link] > sounds like such a great and innovative ideaGreat, maybe. Innovative, no: ALT Linux was two years into this model before RH and SuSE started scratching their heads in a visible manner. The commercial firm (ALT Linux Ltd) produces the distributions (which are also published on FTP and widely mirrored) based on its public unstable tree, ALT Linux Sisyphus, which is considered as a product of ALT Linux Team (some 20% being fulltimers in commercial firms who benefit from the development directly and some 80% volunteers). ALT Linux Ltd is providing the infrastructure, QA (mostly) and publishing (until recently; now professional publishers got in). Partners and community also do maintain packages, create additional infrastrusture and resources, and provide specialized and public support. So far so good -- I believe it's somewhere in between of RH/SuSE/Mdk, Debian and Owl which is quite unique. Weak points being currently low level of manageability using GUI tools (the framework is being revamped) and lacking PR/marketing, strong points being considerable security out-of-box [*], most packages being readily available for immediate use right after installation with apt-get, and very nice community to communicate within. [*] from the last talk security advisory (translate.ru'ed): "In ALT Linux distributions, talk service is off by default, remote users don't have access to it, and local unprivileged users can't form UDP packets of such special form needed for DoS attack; nevertheless we recommend talk server users to update the package with the fixed version". Last glibc troubles also enjoyed "fixed 3 years ago" status statement. Go figure :-)
Bruce Perens: the Linux colonel talks (vnunet) Posted Sep 9, 2004 8:03 UTC (Thu) by hppnq (guest, #14462) [Link] Great, maybe. Innovative, no: ALT Linux was two years into this model before RH and SuSE started scratching their heads in a visible manner. Damn! ;-) I humbly bow to ALT Linux, of whom I must admit I had never heard before. Cheers for the pointer, mate. It's a weak defense, but I was really only thinking of the major distributions, although I am aware that there is a vast number of unique Linux distributions, each of them with their own merits, as well as Linux companies that are developing business models that hopefully serve them well. Don't you just love Open Source? ;-)
Bruce Perens: the Linux colonel talks (vnunet) Posted Sep 9, 2004 10:26 UTC (Thu) by gvy (guest, #11981) [Link] > of whom I must admit I had never heard before Heck, that's what I'm regarding as "weak publicity" :-) btw: the pointer, DistroWatch, LWN, 2.4 beta5 (seems like release's on the way to the publisher) [skip: sure] Yeah, and I hope that business model patches will get at least as agile between companies as software patches are to get the overall result better yet. :-) [err... even though the latter aren't nearly as agile as could be] > Don't you just love Open Source? ;-) Exactly. :-) --
Bruce Perens: the Linux colonel talks (vnunet) Posted Sep 15, 2004 7:51 UTC (Wed) by angdraug (subscriber, #7487) [Link] ALT Linux, of whom I must admit I had never heard before No wonder: ALT Linux is really big in Russian-speaking countries (here in Minsk it is second only to Debian), but has zero marketing outside of Russia.
ALT Linux Posted Sep 15, 2004 7:41 UTC (Wed) by angdraug (subscriber, #7487) [Link] ALT Linux was two years into this model before RH and SuSE started scratching their heads in a visible manner. There is significant difference, though. In Fedora's case, most of the work is done by Red Hat, while company tries to present it as a community project. In ALT's case, most of the development is done by the community, while company tries to present the product as its own achievement. Ironically, ALT Linux has only recently run into the same problem with this model as Red Hat: a product that is developed by a community, but controlled by a single commercial entity. It took two years and an unpleasant trademark controversy for community to notice the conflict inherent in such situation. Exactly the problem UserLinux is attempting to solve. Last glibc troubles also enjoyed "fixed 3 years ago" status statement. Not a fact to be proud of, actually. This particular fix (IIRC borrowed from OWL) received a well-deserved bashing from LWN: keeping important fixes to yourself and not propagating them upstream where they belong is not considered a good practice in the FLOSS world.
Bruce Perens: the Linux colonel talks (vnunet) Posted Sep 14, 2004 21:13 UTC (Tue) by zenaan (subscriber, #3778) [Link] > I really don't want to come across as a RedHat zealot... > [Fedora] sounds like such a great and innovative idea: a true cooperation between the community and a company, together creating a unique win-win situation
Seriously, I can't stop smiling over that one.
Well, if you aren't in some way related to Red Hat, I ... don't know what to say to be honest. You _are not_ as cynical as the next guy in your writing, you're a Red Hat zealot. And your suffusively harsh email ends with a smiley, as though that somehow makes your attack all alright. Go away, shill, for a shill you are!
Bruce's point: Red Hat is using lock-in tactics that go against the community spirit.
If this isn't obvious to you, you are willfully blind, painfully ignorant, or simply marketing Red Hat (which is what I believe and is so bluntly obvious, Red Hat should get better marketing reps in my mind). No point posting a disagreement to this, as I'm sure we'll agree to disagree based on your consistent <cough>marketing</cough> message.
The UserLinux project is Debian. It is a Custom Debian Distribution:
It is also a community of service and support providers, and a marketing organization for these providers.
_This_ is how to do service and support, in a manner consistent with community spirit.
It's about a Free Market with actual competition - competition between service and support providers, while sharing the cost of certification, development of the base distribution/ testing, as well as marketing.
This is the most logical venture I have ever seen.
Bemused
Bruce Perens: the Linux colonel talks (vnunet) Posted Sep 14, 2004 21:21 UTC (Tue) by zenaan (subscriber, #3778) [Link] >> 2) the model decouples the service rendered (support) from what is being paid for (a software seat) in a way that simply cries out for a competitor to set it right
> Both points are common practice, and no vendor seems to be suffering from it.
This is a classic example of the beliefs you belie: your perspective is of the Vendor (Red Hat in your personal case), not the customer.
And this is the very point the Bruce and the UserLinux project are addressing. Red Hat is structured to maximize benefit to Red Hat.
Bruce didn't disagree with you where you say "no vendor seems to be suffering from it" - of course Red Hat doesn't suffer (yet) from this model, because it has no serious competition _from a different model_.
Suse, Xandros, or any other.
UserLinux.com is the real competition here. Red Hat is yet to feel its effects. Red Hat may have to alter its business model in line with this new model, just as Proprietary software has in very large ways given way to Open Source and Free Software.
Sincerely
Bruce Perens: the Linux colonel talks (vnunet) Posted Sep 9, 2004 3:08 UTC (Thu) by bojan (subscriber, #14302) [Link] > Even Red Hat acknowledges this, or they would not have attempted to create a poor copy of Debian called Fedora.
I'm not sure why you think Fedora is a copy of Debian, poor or not. Debian is a collection of over 12,000 packages, Fedora is not. Debian is a true voluneer project, Fedora is not. Fedora is release often, Debian is not. And so on and so forth.
I'm guessing you are referring to the "development model", but not even that is the same. Outside contributors get more chance, but Red Hat does most of the heavy lifting. No doubt about that at all. In all honesty to Red Hat, they had the courage to introduce some really interesting stuff into their disto, way ahead of other folks. For example, NTPL support (without this threading on Linux - read Java containers - really sucked), O(1) scheduler for 2.4 kernels, exec-shield to minimise danger of buffer overflows etc. In essence, they are attempting to assert themselves as an "engineering force" when it comes to Linux. So far, it seems to have worked.
Don't get me wrong, RHEL will never run on any of my systems for a very simple reason - it is too expensive. But, I give Red Hat credit when credit is due. They have done a lot for open source. Whether they are doing a good thing for their own business is their own problem, IMHO.
Bruce Perens: the Linux colonel talks (vnunet) Posted Sep 9, 2004 4:49 UTC (Thu) by BrucePerens (guest, #2510) [Link] I'm not sure why you think Fedora is a copy of Debian, poor or not.It may not have worked out that way, but when it was introduced, Fedora was very clearly aimed to attract the sort of community that already existed for Debian. I think it was also intended to increase the number of packages in a single repository that would install on RHEL. And we want to help Debian get to annual releases. Debian actually releases more often, though, if you count "sid". Having run sid for years and years, I have always been impressed that something that is updated every day can be so solid. I do not discount that RH has contributed a lot of good Free Software and has done an interesting job with keeping 2.4 viable for business people. I am sort of concerned that now that Linus is not going to make stable kernels, the RH way of handling 2.4 might be the only way to do stable kernels. I would prefer to see more collective stability engineering take place. Bruce
Bruce Perens: the Linux colonel talks (vnunet) Posted Sep 9, 2004 7:32 UTC (Thu) by gvy (guest, #11981) [Link] > but when it was introduced, Fedora waswhich one? ;-)
Bruce Perens: the Linux colonel talks (vnunet) Posted Sep 9, 2004 7:54 UTC (Thu) by bojan (subscriber, #14302) [Link] Is "sid" actually "released", in the sense Fedora is? My understanding is that sid is "always in flux", which is not the case for Fedora. With Fedora, you get milestone releases and then a set of updates for that particular release. Then, later on, Fedora Legacy (this is pure volunteer bit) takes over and makes sure security issues are fixed for the next year and a half or so.
Just one example as to who actually does the work on Fedora. A few days back I bumped into a strange memory leak in PostgreSQL libraries shipped with FC2. Posted a bug on Red Hat's Bugzilla for FC2. In the matter of hours, a Red Hat employee got on it and spent several hours attempting to duplicate the problem on his setup (admittedly with no success) and we left it off where I have to give it a try on an unspoilt FC2 machine (my suggestion) to verify that it isn't my own box that's causing the issue. So, this person at Red Hat actually provided several hours of unpaid support for me. I know that one cannot count on those things, but it nevertheless happens. It is Red Hat that make the majority of Fedora going, there is no doubt about that at all. Of course, other people help, but even if that help wasn't there, Fedora would still be released, IMHO. Possibly with more bugs, but released all the same.
The value of Fedora is in the delivery of the snapshots of a selection of state-of-the-art free software. Maybe not that great for production quality servers, but just fine for making sure you get a reasonably recent cut of binaries that will get timely updates when you need them. So far, Fedora has been relatively easy to keep upgraded, with minor issues related to software choice changes.
As for the stable kernel, I don't see how Red Hat's "stability" (or better, "backporting and enhancement") work cannot be incorporated into other distros. They only need to choose to do it, that's all. The problem is that Red Hat traditionally push it a little bit harder, mostly because they have clients that want to run Linux as a replacement for the big iron. Not to everyone's liking, but there is nothing stopping anyone from doing the same thing with Red Hat's code. There are absolutely no secrets there. In essence, this seems to be a political more than a technical issue.
I have to admit, I was sceptical about Fedora in the beginning, but it turned out better than expected, IMHO.
Stable kernels Posted Sep 9, 2004 12:50 UTC (Thu) by corbet (editor, #1) [Link] "I am sort of concerned that now that Linus is not going to make stable kernels, the RH way of handling 2.4 might be the only way to do stable kernels. I would prefer to see more collective stability engineering take place." Wow, Bruce, you understand things rather differently than I do. The whole point of the new kernel development model is to allow the distributors to stop backporting tons of stuff into older kernels, and to cut way back on the huge piles of distributor kernel patches in general. We are trying to get away from the "RH way". Linus very much intends to release stable kernels; it's just that the way to get there has changed. The "don't allow substantive changes for a year" approach doesn't work very well. It's better to get the patches into the mainline and get them used. The result has been a lot more functionality in the hands of the users more quickly, and some quite stable kernels. 2.6.8.1 has not been the best example, but 2.6.7 was quite solid, and 2.6.9 should be as well.
Stable kernels Posted Sep 9, 2004 15:21 UTC (Thu) by BrucePerens (guest, #2510) [Link] Jon,Maybe you should publish an article about this, because I don't understand what is going on with stable kernels, and if I don't get it, probably a lot of other people don't either. I agree that 2.6.8 is solid, I am running it on my net-facing systems, including the system at Serverbeach for which I don't have a remote console. But I don't see yet how Linus is going to have his cake and eat it. Bruce
Stable kernels Posted Sep 12, 2004 1:35 UTC (Sun) by hppnq (guest, #14462) [Link] Jon, of course, has already written an excellent article about the new kernel development process. I was too lazy to look it up before, but now, with the new, shiny Kernel Index, it took me two seconds. ;-)
Bruce Perens: the Linux colonel talks (vnunet) Posted Sep 9, 2004 14:10 UTC (Thu) by rahulsundaram (subscriber, #21946) [Link] Hi
I think Userlinux is a valid and needed alternative. I appreciate your efforts on that but I am not pleased with it as a largely reactionary effort. It would be helpful to stop pinpointing Redhat's efforts as not viable and just do your own thing.
Even when I am not using redhat/fedora I appreciate their commitment towards Free software and I wish them luck in their enterprise focus
regards
Linux == Redhat in US? Posted Sep 9, 2004 6:24 UTC (Thu) by gvy (guest, #11981) [Link] > "I'm very sorry, but you seem to miss the point. RedHat might be expensive,> but Linux is not. Not in beer, nor in freedom."
Errr... I believe in US Linux and RedHat are largely synonyms for "corporate minds".
Another question is that "beer" and "freedom" don't hold as arguments too: "TCO" (as fuzzy as it is) and "strategic independence" may do.
Just my EUR.02
Linux == Redhat in US? Posted Sep 9, 2004 7:31 UTC (Thu) by hppnq (guest, #14462) [Link] > "I'm very sorry, but you seem to miss the point. RedHat might be expensive, I meant that there is a clear distinction between Linux and Open Source on the one hand, and products and services like RedHat's on the other hand, and that this is one of the points that Linux advocates should bring forth when dealing with people and companies who don't know the difference. I'm quite sure people in the US know that Linux != RedHat. ;-)
Linux == Redhat in US? Posted Sep 9, 2004 14:50 UTC (Thu) by pflugstad (subscriber, #224) [Link] > I'm quite sure people in the US know that Linux != RedHat. ;-)
Well, no, they don't. Especiallly not in corporate IT departments and the like where their ONLY introduction to Linux is Red Hat. There may be clueful people around, but the vast majority think RedHat == Linux.
|
|||||||||||||