We'll see | Matt Zimmerman

a potpourri of mirth and madness

Greg Kroah-Hartman’s Linux Ecosystem

As the opening keynote at the Linux Plumbers Conference, Greg Kroah-Hartman delivered a talk entitled “The Linux Ecosystem, where do you fit in it?

There were, let’s say, a few elements of it which I found objectionable.

The central issue, of course, was that he devoted a large portion of the talk to showing that Canonical contributes fewer patches to the Linux kernel than many other companies.  While it’s somewhat flattering for Canonical’s role in the community to be a headline topic for a conference like this, the message that he chose to deliver was a negative one.  He presented list after list of contributors, ranked by number of patches, and pointed out how low Canonical was on each one.

I approached him immediately after his talk to suggest that we have a conversation about the topics he raised, but he wasn’t interested in talking with me at that time.  I made a standing offer to talk with him at any time during the three-day conference, and hope that we can get to the bottom of this.  Until then, I’m not sure what exactly he’s trying to accomplish.

Meanwhile, I’d like to make a few points.  I’ll start with a disclaimer, something that Greg chose to omit from his presentation: I work for Canonical, and am one of the founding members of the Ubuntu project.

Greg’s view of the ecosystem is odd

Greg considers the “Linux ecosystem” to be GCC, binutils, the Linux kernel, X.org, and a handful of other projects.  He disregards most of the desktop stack (including GNOME and KDE), all desktop and server applications, and most anything else that is recognizable to an end user as “Linux”.

Some members of the audience picked up on this and commented.  His justification for this was that these other components are not specific to Linux.  “Any contribution to GNOME also benefits OpenSolaris”, says Greg.  Apparently, this means that GNOME isn’t an important part of Greg’s Linux ecosystem.  “I had to draw the line somewhere”, he says.

This is not the Linux ecosystem that I use and contribute to.

Greg’s figures are wrong

The first slide in his presentation acknowledged that he had miscounted Canonical’s contributions to the Linux kernel.  As he freely admits, his method is not an exact science and there were many other errors.  However, given that he intends to use these statistics to attack Canonical, he should take more care in compiling them.

His original claim, given at a Google tech talk in June 2008, was “Canonical doesn’t give back to the community”.  He supported this by saying that “Canonical made six changes to the kernel in the last five years.”

Greg now states that Canonical has in fact contributed in excess of 100 patches.  This means that his raw data for the kernel was incorrect by two orders of magnitude.

His LPC presentation also put forth some new claims regarding contributions to other projects.  In particular, he lambasted Canonical for not contributing to binutils at all (zero patches).  What that actually means in terms of our position in the ecosystem is debatable, but numbers are not.  It’s true that Canonical does not contribute as much to binutils as Red Hat does (and more on that later), but as Kees Cook pointed out after the presentation, he personally contributed a patch (now merged) which credited Canonical as his employer, which fell within the date range of Greg’s analysis.

Of course, none of these errors impact his fundamental conclusion, which is that Canonical engages in relatively small-scale development compared to Red Hat and Novell.  No one is disputing that.  However, the fact remains that his data is inaccurate.

Greg is failing to disclose his bias

Greg is, of course, a well respected contributor to the Linux kernel, having sustained a significant level of contribution over a period of several years.  I’m grateful to him for his technical contributions, which of course benefit Ubuntu as a consumer of the Linux kernel.  However, his contribution to the public dialog about the Linux ecosystem leaves much to be desired.

We all have bias, and the best that we can do is to disclose it so that others can take it into account when hearing our ideas.  Unlike the presentations given by other Novell employees at this and other conferences, Greg’s slides omitted the Novell logo.

Novell is, of course, a competitor of Canonical, being an operating system vendor (and a large one at that).  To attack his corporate competitors without acknowledging his affiliation is in poor taste.

Greg’s comparisons are bogus

The fundamental argument he makes is that Canonical doesn’t contribute as much as Red Hat, Novell and many other organizations which he names.  This is absolutely true.  He says that it’s unethical to claim more contributions than one has made, and he’s absolutely right there as well.

However, no one, certainly not Canonical, has ever claimed that Canonical does as much Linux development as Red Hat or Novell.  He’s refuting a claim which has, quite simply, never been made.

Canonical is primarily a consumer of the Linux kernel.  It is one of the building blocks we need in order to fulfill our primary mission, which is to provide an operating system that end users want to use.  It is, on the whole, a good piece of software which meets our needs well.  We routinely backport patches from newer kernels, and fix bugs which are particularly relevant to us, but our kernel consists almost entirely of code we receive from upstream.

Why, then, does Greg feel that Canonical should be expected to make more changes to the Linux kernel?

Is it because Ubuntu is a very popular system, with a lot of users?  It is that, but most people who use Linux aren’t kernel developers, so a large user population doesn’t translate to a lot of Linux kernel patches.

Is it because he thinks Canonical is making a lot of money off of the Linux kernel, and should give some of that back?  We make no secret of the fact that Canonical as a company is not yet earning a profit.  We make a promise to our community that we will never charge money for Ubuntu.

Is it because he thinks Canonical developers are writing a lot of patches and not contributing them?  If he does, he hasn’t compared our kernel tree with Linus’.

Why then?

Greg’s approach is not constructive

If we give Greg the benefit of doubt, and assume that he doesn’t merely have an axe to grind, then there must be some genuine concern behind all of this.  He must feel that Canonical is somehow not doing the right thing.

If that’s true, why hasn’t he ever talked to me about it?  He has my email address, and we’ve exchanged mail and spoken on the phone before.  Why am I hearing about this through presentations given to Google employees, posted on YouTube, or delivered to audiences of kernel developers?

To present his commentary in this way is indefensible.  LPC is promoted as a productive community event aimed at solving problems, and Greg has used his voice as a speaker to promote a much less honorable agenda.

When this sort of thing happens on mailing lists, it’s called trolling.


Written by Matt Zimmerman

September 17, 2008 at 22:10

Posted in Uncategorized

Tagged with ,

101 Responses

Subscribe to comments with RSS.

  1. thank you Matt!
    As a volunteer Ubuntu developer I also dislike that people link Canonical’s “give back” to Ubuntu as a whole. Canonical just isn’t that big and much of Ubuntu’s contributions aren’t via @canonical.com addresses.


    September 17, 2008 at 23:02

  2. Well .. most kernel developers are probably way too clever than to believe fairly lame Novell FUD. So don’t worry about it too much.

    I also saw the Google Video. There he did not attack Canonical from the start but only when people guessed that Canonical was the one of the biggest contributors… maybe that made him snap or something .. who knows ..

    I follow Ubuntu and Canonical quite closely and I have never heared from anyone that you guys do a lot of low level development.

    You do however do a great job at integrating things and at the higher level of the stack. And the best community work etc. Thanks again for that.

    And it is Free Software! Don’t contribute with an attitude like GregKH. That talk sounds utterly pointless. PCLinuxOS was the most popular distro for some time. How much low level development did they do? I guess not a lot. But only weird people would point their finger at them. Really strange ..

    Maybe we are witnessing the birth of a new form of zealot. The “contribute everywhere or I will troll” zealot.

    Let’s call those “Greghks” B)


    September 18, 2008 at 00:22

  3. How many bug-reports and fixes have been assisted from Canonical initiatives such a Launchpad?

    You cannot fix a bug without knowing about it first. Launchpad lowers the bar to reporting bugs and collecting sufficient information to fix them. This type of activity is just as valuable (if not more so) as the final coding which fixes the bugs.

    I really think that Launchpad is the best thing we’ve (FOSS) had in a while because of it’s interoperability with upstream bug trackers.


    September 18, 2008 at 00:27

  4. This talk is now on LWN: http://lwn.net/Articles/298864/


    September 18, 2008 at 01:34

  5. “Canonical is primarily a consumer of the Linux kernel.”

    Maybe you should move from being primary a contributor instead?

    “However, no one, certainly not Canonical, has ever claimed that Canonical does as much Linux development as Red Hat or Novell”

    How about when Mark shuttleworth lied about having a better security record than Red Hat.


    Would you care to retract those claims and admit that Red Hat does a much better job in security and contributions to pretty much everywhere?

    “I really think that Launchpad is the best thing we’ve (FOSS) had in a while because of it’s interoperability with upstream bug trackers.”

    Bull shit. Launchpad is a Canonical controlled proprietary service. Proper interoperability would come from using aggregation between distributed instances. For example, bugzilla uses xml-rpc for this.


    September 18, 2008 at 03:09

  6. That’s funny… Launchpad is the single biggest reason that I don’t contribute more to Ubuntu. It’s gigantic, slow, hard to use, and totally closed. I find Bugzilla to be easier to use! There are many wonderful things about Canonical and Ubuntu but I don’t think that Launchpad can be considered among them.


    September 18, 2008 at 04:09

  7. @neo:

    Maybe you’d like to post the chapter and verse that say contribution is required in order to use or distribute GPL code?

    James A.

    September 18, 2008 at 08:22

  8. @neo
    I did not lie – I was quoting a genuine study which looked at patches published over time. Thanks for the link to the LWN article, I will ask our security team to respond to that, I hadn’t seen it before.

    W.r.t. Launchpad vs bugzilla, you may not be aware that we have funded a fair amount of work on bugzilla’s xml-rpc interfaces, as well as new interfaces for Trac, and are working on those for other trackers.

    Mark Shuttleworth

    September 18, 2008 at 08:28

  9. Well put.


    September 18, 2008 at 09:31

  10. > Is it because he thinks Canonical developers are writing a lot of patches and not contributing them? …. our kernel consists almost entirely of code we receive from upstream.

    While technically true, it’s misleading. Ubuntu ships a goatload of out-of-tree drivers in the kernel, but does no work whatsoever in cleaning them and submitting them to be included in mainline. This leads to two problems:

    1) ubuntu appears to support more hardware than Linus’s kernel, yet none of these new drivers are flowing from ubuntu to Linus’s
    2) The drivers are typically out of mainline either because they are still beta, or otherwise low quality. Thus ubuntu’s kernel is more buggy than mainline kernel, creating a general impression that “linux is buggy”.

    Psst. here’s a low hanging fruit: lm-sensors subsystem is in the kernel simple, has many drivers not merged yet mainline and is currently without a maintainer.


    September 18, 2008 at 10:13

  11. Matt, can you explain ‘two orders of magnitude’? I count it as 6 v. 100 would be rounded (when logarithm’d in base 10) to only one order of mangitude: Log(100)-Log(6)<1.5 so rounds down to 1. (This may be pedantry because I can see the other argument: there’s three digits in 100 and only one in 6.)

    That aside: I think that it is reasonable to presume a disconnect between the SABDFL’s orations about upstream integration and Ubuntu community members’ interactions with upstream. Is that right or fair?

    Thanks for being available that I can ask you these questions.


    September 18, 2008 at 10:37

  12. When you are talking about Novell as “competition” you have no Idea about what Linux should be. “Cooperation instead of competition” guys. Thats the only way to achieve Linux’s goal. Which is: Convince users to use free software. Whether it is Ubuntu or Fedora or OpenSuSe.
    And yes, I also noticed that Ubuntu does not contribute enough. RedHat for example employs the core developers of ConsoleKit, PolicyKit, PackageKit and PulseAudio. No wonder that Ubuntus integration of PulseAudio has problems.
    I would love if Canonical would employ a few developers whose job it would be to make sure that such core components are well integrated. Plus one developer that makes sure that Ubuntus own creation, upstart, is fully integrated and its possibilities exploited to the max.
    Well it’s good that Canonical hires developers to build programs that make integrating proprietary drivers easy (jockey). But it would be even better if they hired developers who work full time on open drivers such as nouveau, radeon, Gallium3D and kernel mode-setting. Just as RedHat does. That would really make a difference.

    Arpad Borsos

    September 18, 2008 at 11:11

  13. From a *normal user* perspective LaunchPad is definitely the best the FLOSS world has to offer as an entry contribution tool. It is in the process of becoming open source so we can’t complain about that.

    If you are an avid GNU zealot and know your way around of course you will find bugzilla and the like to be great. It really is! But if you deny the importance of LP for the moment of Ubuntu and the FLOSS adoption rate it has impulsed among *normal users*, then you really do not understand the way Ubuntu works.

    As for Ubuntu not contributing to the kernel userland and low-level utilities: Come on! It’s time people get over the fact that Ubuntu have the biggest community and see it as a threat to stablished contributors. Ubuntu and its community is here to deliver FLOSS to the masses, not just technical zealots that have been using FLOSS for years anyway. Now what it has achieved in less than four years is not based on luck, but hard community and technical work.

    Give it some credit and let’s continue contributing in what we can, the best way we can.


    September 18, 2008 at 11:20

  14. For the record, Launchpad is the reason why I’ve filed lots of bug reports for Ubuntu (which I don’t even use at home) while for Debian I’ve filed maybe three bug reports in four years; or put the other way: if Debian had a bug reporting tool as simple and easy as Launchpad, I would certainly report more bugs for Debian (in comparison with Launchpad, I find it a real hassle to make a bug report for Debian, even when using tools like reportbug-ng).


    September 18, 2008 at 11:30

  15. “We make a promise to our community that we will never charge money for Ubuntu.”

    Where that it’s true that you can always freely download it, it’s worth a mention that it’s in Best-Buy for $20: Is it worth re-wording this in the future?


    September 18, 2008 at 11:33

  16. Focusing more on the Ubuntu project, without it. I think we would not have mayor hardware vendors preinstalling Linux. please do consider BUG#1 whether we like it or not we are in the same big boat ( If you read Bug#1, I don’t know much about Novell).

    How old is SuSE? How old is Ubuntu? projects that are born under different circumstances. give ubuntu a few more years and probably it will provide with more significant input for this specific area of contribution (Kernel Work).

    Ubuntu has been focusing on delivering the best of libre and opensource software experience for the desktop and has succeded. It made an impact. Ubuntu releases are every six months and it is a pretty wild ride. and maybe and just maybe it contributes more to it’s direct upstream. with time we will see more contribution headed to where everyone likes. but please also consider that Ubuntu builds upon foundations of Free Software to provide a great desktop (and server )experience, sometimes it may not include much kernel work :(.

    I really wouldn’t mind comments made in order to make a point stomping on somebody else. I appreciate the work effort for the guys at Novell. but I am not interested in their community offer. And much less their dubious tie-in with proprietary vendors. To each its own.


    September 18, 2008 at 11:40

  17. Well put, Matt.

    To encourage folks to participate in Linux kernel development, Greg K-H should definitely share the [correct] numbers about who contributes the most. But he’s way off base for singling out Canonical in this spiteful, negative light.

    Adam Monsen

    September 18, 2008 at 12:23

  18. It is true that companies such as Novell and RedHat have invested a lot into Linux development. On the other hand, they also failed to do what Canonical is doing now, namely, delivering a clean, polished, well-integrated distribution that end users can and, above all, want to use. So, we have a funny situation now, where the company enjoying the popularity (and the potential profits down the line) is a relative newcomer that has made comparatively small contributions to overall Linux development.

    Of course, this situation is difficult to swallow for some people. One should not forget, however, that, first, Canonical has been playing by the rules from the start, and, second, RedHat, Suse, Debian and many others actually had the opportunity to deliver a good end-user system, but, for whatever reasons, they missed it. I’d rather ask those who disregard Canonical’s contributions as “just putting together a few software bits” a simple question: if it was really that easy, why didn’t you guys do it first?

    Martín Soto

    September 18, 2008 at 13:00

  19. Just to add: I don’t think Greg’s view of the ecosystem (though it’s arguable whether it’s a reasonable view) would really change his point much at all. It’s not as if there is a great discrepancy with the desktop elements in terms of contributions from Canonical: Red Hat and Novell are still the runaway biggest contributors even when you consider just the desktop elements of Linux (KDE, GNOME, OO.o, X.org).

    That aside, it seems that the main complaint in this blog post is just the way in which Greg has gone about his complaint, which is fair enough, but it still doesn’t address the issue (but rather: side-steps it) of why so many FOSS developers have the impression that Canonical takes a lot and doesn’t give a lot back. It of course doesn’t help that one of Canonical’s most prominent and flagship products — Launchpad — is just entirely proprietary. Contrast this to the openSUSE Build Service which has always been published under the GPL.

    The frustration of FOSS developers is of course increased just because it seems that in some cases Ubuntu aficionados tend to encourage the idea that Ubuntu is the innovator behind given technologies — I recall seeing several people believe that Ubuntu created Compiz, for example. So in this case, not only are Red Hat and Novell spending considerable resources on improving the Linux desktop, but they’re not getting the credit for it. The hype around “Bullet-Proof-X” is not too dissimilar to this.

    The comments above remarking that Launchpad is apparently the only way to file bug reports (again, see Bugzilla) are a little lost on me I’m afraid.


    September 18, 2008 at 13:03

  20. there are simply three types of bugs that are reported
    to distros:
    1) it was already fixed upstream. backport is done, fixed.
    2) it was not fixed upstream but doesn’t makes sense anymore
    because the upstream changed too much. more frequent in
    enterprise distros that have to freeze versions for long time.
    3) it was not fixed upstream

    I think what Greg is talking about is the 3rd type. You see,
    Ubuntu has a lot of users and probably are exposed to a
    lot more scenarios and environments than other distros.
    And in this case, it’d be natural to see much more patches
    or even more detailed bug reports for the kernel.
    So, that may let one to think that either you don’t care
    about kernel bugs (the fix will be available when *they*
    fix it upstream) or you you don’t have enough people
    to do it, and in the last case, a better integration of
    whatever you use to track bugs with lkml would be nice.
    Note: I’m just trying to get something constructive out
    of this. Instead of doing the classic “let’s find arguments
    to refute his statements” why not try to find ways to
    improve (even more) what you already do today?


    September 18, 2008 at 13:58

  21. “Any contribution to GNOME also benefits OpenSolaris”

    hah! By the same argument, he shouldn’t have picked GCC, as contributions to it also benefit the BSDs and MacOS X. And x.org, Jesus! Contributions to it benefit pretty much every Unix under the sun.


    September 18, 2008 at 13:59

  22. Multiple times I have found kernel bugs or regressions fixed by Ubuntu devs with no reference to the patch or statement that the patch was sent upstream.

    When I email the author, I have gotten statements from them that they have no plans or no idea if the patch will be submitted upstream.

    This is not being a good community member. And I’m sure it happens less often than I have personally experienced. If the kernel devs could at least link to the patches then others could submit on their behalf.


    September 18, 2008 at 14:21

  23. Was leistet Canonical…

    Nette Diskussionen über die Hilfe von Canonical zu der Community gibt es genug. Danke an Immundus/CMD_Keen für den Link, ich denke er sagt sehr viel aus:

    Seraphyn Blog

    September 18, 2008 at 14:24

  24. […] the company that brought Mono and patent threats to GNU/Linux also comes Greg Kroah-Hartman, who neglects to mention his affiliation when he disses Ubuntu. Greg is, of course, a well respected contributor to the Linux kernel, […]

  25. Launchpad is a cool application … but it is not FOSS. If it was FOSS I could submit patches to it as well as run my own if I wanted to … last I checked I can do neither.


    September 18, 2008 at 14:28

  26. @bronson: Next year Launchpad will be open source ( and one of the only code hosting / issue tracking / etc. platforms out there that is open source .. SF and Google Code and a lot of others are not. )

    I read the talk on Gregs homepage. And it is so negative and just screams bad character. Really really lame key note for such a important conference.

    I cannot read the LWN article, though ( No way to pay electronically ).
    But I can read the comments ( http://lwn.net/Comments/ ). And people are really silly. How can a small company like Canonical be compared to IBM and Google?
    At least give some relative number contribution/employees for example. And do it for all projects that make up a that specific distro.

    Google for example does a lot of work because of Andrew Mortons work. But they have lots and lots of kernel developers working on their own kernel and don’t contribute that work at all.

    So according to GregKHs logic all the well paid Google engeneers should leave Google?

    Yeah baby, you tell’em!

    I can only repeat myself. Utterly pointless and it goes to show: Do not trust any statistics you did not fake yourself.


    September 18, 2008 at 14:58

  27. @TomM: Canonical promises that Ubuntu will always be available for free, and goes quite a bit further than that by sponsoring its distribution in various forms. Naturally, though, we have no control over whether another company chooses to charge for it. The fact that Ubuntu is available for purchase at Best Buy with a physical CD, printed documentation and 60 days of installation support is perfectly consistent with our values.

    http://www.fsf.org/licensing/essays/selling.html is a good resource regarding the sale of free software.


    September 18, 2008 at 14:59

  28. “However, no one, certainly not Canonical, has ever claimed that Canonical does as much Linux development as Red Hat or Novell. He’s refuting a claim which has, quite simply, never been made.”


    “I also saw the Google Video. There he did not attack Canonical from the start but only when people guessed that Canonical was the one of the biggest contributors… maybe that made him snap or something .. who knows ..”

    I saw the video too, and so that I say … the claim HAS been made … certainly not by Canonical, but there is a lot of indirect(*) support for such claims from people connected to Canonical somehow …

    (*) so that no one can point a finger to any obvious lie

    oh and btw, if you are soo much concerned with Greg being Novell employee, and that Novell is a competitor to Canonical, how do you explain that he praises Red Hat for example, isn’t that a competitor too?

    and if you say “Of course, none of these errors impact his fundamental conclusion,” why do you then write “his data is inaccurate” in bold when we all already know that “his data … was incorrect by two orders of magnitude”?

    – now THAT is what I call trolling


    September 18, 2008 at 15:00

  29. It seems NOVL actively encourages its employees to try to start these useless barfights in the open source community. It seems to become a tradition in their company: Miguel de Icaza with Mono, Michael Meeks with their wannabe OOo-Fork, GregKH on the kernel.

    It seems like the company doesnt cherish developer egos enough so those ego get distracted and troll and harass others.

    Or its not leadership incompetence, but they just do what they where payed for by MSFT …

    Well, probably not.


    September 18, 2008 at 15:02

  30. @T0m: LWN is a great resource and well worth the subscription if you can work out a means of payment. If you’d like to read this specific article, email me and I’ll send you a free link.


    September 18, 2008 at 15:06

  31. @Mike, oversights are always possible. If you have references to things which ought to be submitted upstream and haven’t been, please email me about them.


    September 18, 2008 at 15:08

  32. […] Matt Zimmerman discussed, Greg’s “Linux ecosystem” seems a bit unfairly limited to the kernel, gcc, and […]

  33. @mdz, thanks I will contact you if I come across this again.

    In the meantime, can the Ubuntu kernel developers do better than:

    “Fixed in Hardy”

    How about a reference to the commit if a patch was created or used?

    Thanks! :)


    September 18, 2008 at 15:16

  34. @apokryphos, because Ubuntu is popular and has been the first to expose users to certain technologies, it’s understandable that sometimes they credit us for it. We’ve acknowledged this as an issue, and will attempt to offset it by explicitly crediting these contributions in our pre-release communications. There is a delicate balance to strike, of course, because Ubuntu is (like every other distribution) highly dependent on a great many projects, and crediting everyone fairly is an impossible task, but we’re trying to do the right thing.

    “Bullet-proof X” refers to a system we assembled to help Ubuntu users recover automatically or semi-automatically if their X server failed to start. It relies on open source components contributed by others, like displayconfig-gtk (and of course X.org), but is, in itself, a feature added to Ubuntu by Canonical. The specification is here: https://wiki.ubuntu.com/BulletProofX


    September 18, 2008 at 15:17

  35. @k3ninho, Mark does his best to keep up with the happenings of the project, but it is a very busy place, and he is a very busy person. I don’t expect him to know exactly what is going on at every level. He has occasionally misspoken in public about something we are or are not doing, and generally acknowledges and corrects his errors.

    If you can point to something specific which you feel is inaccurate, I’m happy to respond.


    September 18, 2008 at 15:21

  36. @neo, I’m certainly in no position to retract something which was said by someone else, least of all my boss. ;-)

    I can say that, in my own opinion, good security is very difficult to quantify, and studies like the one Mark references cannot tell the whole story. There are objective measurements to support most any ranking of the major distributions.

    Canonical sponsors free security updates for several releases of Ubuntu at any given time, and this is made possible by the excellent folks on our security team, which is likely a fraction of the size of Red Hat’s or Novell’s. They work very hard to respond to security issues by quickly prioritizing, analyzing and patching them, and on top of that, they find time to work on proactive security measures in cooperation with Debian (http://wiki.debian.org/Hardening).

    I’m proud of the work they do, and am not particularly interested in statistical comparisons. We measure our own work in this area, and I think we serve our users well.


    September 18, 2008 at 15:33

  37. I think Launchpad is a distraction from the issues here; it is in many ways an entirely separate project from Ubuntu, and many other projects make use of it which are unrelated to Ubuntu. Certainly, no one will claim that Launchpad is a direct contribution to the Linux kernel or any other project. I think it does do a good job of collecting relevant data about bugs in one place, and Mark has said that he will make Launchpad itself a contribution to open source by releasing the code.

    The launchpad-users mailing list would be a better place to discuss it.


    September 18, 2008 at 15:42

  38. @kavol, put simply, I think Novell is probably more worried about Ubuntu than Red Hat at the moment.

    As to pointing out the errors, a misstatement of fact deserves to be pointed out regardless of the conclusion it’s being used to support. I think that Greg’s methodology is not very sound, and this should detract from the credibility of some of his claims.


    September 18, 2008 at 15:46

  39. You can flame me all day long, but IMHO GKH is right!

    1. Ubuntu adds loads of drivers to the kernel to increase hardware support but apparently has absolutely no intention to clean them up & get them in the main kernel.

    2. Often stuff is fixed in Ubuntu but not pushed upstream to get it fixed there as well.

    3. Quite some times Ubuntu doesn’t give credit to the original source – e.g. Compiz.

    So, while Canonical for sure is smaller and has less money than RedHat or Novell but the widely spread opinion (with the exception of the buntu zealots) that Ubuntu takes without giving back doesn’t come from nowhere.

    Therefore I would suggest you change your attitude from “Upstream can get the source from our version control if they want” to actively trying to get your fixes & additions accepted upstream every single applicable time.

    If you do this the public view will change but for the moment GKH is right.


    September 18, 2008 at 17:32

  40. I agree with you Matt, it seems that his agenda was purely to slam Canonical for his own benefit. He is clearly not interested in providing constructive criticism or fixing a grievance that he has with you. He purely used that talk as an opportunity to slam Canonical. While the importance of Canonical’s contribution to the kernel is certainly debatable, I hope that no one takes him seriously.


    September 18, 2008 at 18:10

  41. Greg KH is indeed a zealot that likes to rant about things. He may do some good work, but his presentation has never been professional to me.

    His talk at LPC was an inappropriate–especially with the obvious Novel/Ubuntu conflict of interest.

    He could have made the point once and moved on–that would have been a productive jab. Instead he thought it was worth our expensive time to rant on and on. IMHO.

    Next time, I’ll sleep in.


    September 18, 2008 at 20:15

  42. I see in the blog entry that lots of statements are refuted but not corrected. What are the real numbers for the kernel? What are the real numbers for gnome?

    Mike McGrath

    September 18, 2008 at 21:06

  43. Novell are still a proprietary software company, and it seems some there just can’t stand anyone using their free code – in the way it was originally intended I might add. What Canonical do is far and away above what many other 1-dollar distro’s out there do, at least they do more than just re-brand and repackage an existing one. So why are they a target? Because they’re popular?

    Integration and marketing are high value contributions to any movement, and looking at pure numbers and counting LOC is the sort of bullshit you get from people more interested in property than freedom. With that attitude, one wonders why he’s interested in GNU or Linux at all.

    Perhaps he should be looking at commitment to freedom rather than lines of code? Anecdotally, Redhat wins here, Canonical has a few questions, and Novell come a big fat proprietary plus patent-deal-with-microsoft-infected last.


    September 19, 2008 at 00:55

  44. Canonical contribution to kernel could be small, but Launchpad , Ubuntu , Ubuntu Server is really great goods to free knowledge and free culture.


    September 19, 2008 at 03:15

  45. I would disagree with the “credit given to Compiz” part.

    Every Ubuntu user who’s more or less advanced (and actually cares) knows that it’s Compiz. In all videos, it’s “Compiz also”.

    Just because it’s given a human and an understandable name in one menu item, it doesn’t mean it’s being credited.

    Oh, that said, someone should check out the numbers: http://dustinkirkland.wordpress.com/2008/09/18/whats-behind-gregkhs-latest-rant/

    … and continue to rant on how efficient canonical is given their size, age, and revenue.


    September 19, 2008 at 03:59

  46. While I did find GKH’s speech to be strange, I too would be interested in correct data. The claim has now been made that his figures regarding Ubuntu’s contributions are incorrect. Could we then have the correct numbers, so we could then do true comparison to what others are contributing?

    That said, I don’t think that this was about Novell vs. Canonical. I think Greg was speaking as a kernel-developer first, Novell-employee second. Besides, his talk did not have pro-Novell spin to it, in fact his figures clearly showed that Red Hat, not Novell, is the biggest single contributor.


    September 19, 2008 at 06:09

  47. I think he may be right about his “Linux ecosphere” – but I live in a “free software community” which includes GNOME and KDE and Open Solaris.

    I may not use Open Solaris, but as free software it is part of what I see as my community, and GKH tries to divide that community into “Linux” and “not Linux”.

    And that worries me.

    Linux is part of the free software community, and Canonical greatly helps the free software community by providing Ubuntu as a desktop user distribution (though I would prefer to have a stronger Kubuntu, Ubuntu is a great contribution!).

    So if GKH says “you don’t contribute enough” because you don’T do as much as others for Linux directly, I think he misses the bigger picture.

    I could install Kubuntu on my wifes computer, and she can work very well with it.
    And that’s something you gave me and many others.

    Arne Babenhauserheide

    September 19, 2008 at 08:08

  48. @TomM:
    Did you seriously do any thinking before you posted? I have this nice package I want to sell you for only $20.00 contains OpenOffice + Firefox they are available for free via download but I want to make them not free by selling them on CD. Please note I have nothing to do with the Open Office project or Mozilla.


    September 19, 2008 at 08:22

  49. @mdz: Two data points:
    (*) The 2007 Linux Collaboration Summit keynote ( http://gh-linux.blogspot.com/2007/06/mark-shuttleworth-keynote-at-linux.html )
    (*) Mark’s own blog: May 12 2008 ( http://www.markshuttleworth.com/archives/146 ).

    These state that greater collaboration among downstream and upstream is a big thing for Mr Shuttleworth. If I’m feeling kind, that’s a call to Ubunteros to integrate better. If I’m feeling mean, that’s a big gap between SABDFL’s talk and the Canonical/Ubunto walk. And if I’m being cynical, both are adverts for Launchpad.

    Ignore the mean and cycnical comments because Free software confirms my suspicions that life is better when we work together. I’d like to know if you think there’s a real disconnect between the things SABDFL has been saying and what Canonical/Ubuntu/Ubuntu’s volunteers do with respect to Upstream. And then, whether you think GKH was fair to make the points he did.

    (BTW: Launchpad is a good tool and the recommendation that people make use of its bug-catching, translation, release-scheduling, ppa-hosting, all-singing/dancing features is fair. But the mainline Kernel is still smarting from BitKeeper and wary of proprietary solutions, and I’d go as far as to say that diversity — the bazaar — is good. So perhaps Launchpad.net would make a good centralising *website location* but never a centralised *hub* for diverse projects like Linux Kernel, GCC, X.org and so on. Is there any scope for a web transport interface so that Lanuchpad can allow you to submit kermeloopses or track bugs in b.kernel.o and/or b.mozilla.o? I hear you already: show me the code!)

    Cheers again,


    September 19, 2008 at 10:57

  50. Just wanted to say Ubuntu rocks. Ubuntu is the only flavor to have figured out that the UI is the most important aspect to end-users. Non-developers form about 99% of computer users in the world. If users really cared about kernel speed and kernel crashes, everyone would of moved away from Windoze alongtimeback.


    September 19, 2008 at 12:36

  51. As far as putting credit where credit is due, you might consider putting back the statement on http://www.ubuntu.com/products/whatisubuntu that Ubuntu is based on Debian which provides the majority of the packages.

    Integrating Rosetta with upstream localization processes (as a mere “contributor” submitting translation patches with self-initiative, not as the centralized location for POs) would also go a long way in reducing bad blood with your upstream.

    I think the issues only manifest where Ubuntu views itself as upstream instead of consumer. Get rid of the notion that Ubuntu (as a distribution) should host and develop actual applications, and give each of those a nice proper “independent” home – there’s no reason why Apport or Jockey shouldn’t be Ubuntu-agnostic.

    And first of all, don’t attempt to solve the problems for Ubuntu. Solve them for the whole Linux ecosystem, like with Upstart, and work hard on cooperating with other distros on one solution that solves the problem for everyone. Don’t take the easy way out by hacking together a quick Python GUI for some Ubuntu-specific settings, but get both the settings control capabilities and the GUI into upstream, even if that involves not doing it in Python.

    Compiz didn’t aspire to be a good composite manager for SUSE Linux, but aspired to be a good composite manager for X servers. Which is why it’s now used everywhere around. NetworkManager, PackageKit, PolicyKit, ConsoleKit and PulseAudio didn’t just aim for getting something up and running on Fedora/RH systems, they solve problems at their core, for a large group of people across all distros.

    If you want to be respected as a good contributor, then get rid of the idea that you need to adapt the original system in order to make it usable. Instead, work on the original system until it doesn’t need Ubuntu-specific adaptions anymore. A lack of that mindset is why people are critisizing you all around.

    Jakob Petsovits

    September 19, 2008 at 12:47

  52. […] https://mdzlog.wordpress.com/2008&#8230; “We all have bias, and the best that we can do is to disclose it so that others can take it into account when hearing our ideas.  Unlike the presentations given by other Novell employees at this and other conferences, Greg’s slides omitted the Novell logo.” […]

  53. mdz,

    Don’t be a weasel. You can certainly disagree with your boss claims and you should if you think it is incorrect which it certainly is. Why don’t you just admit that?

    “I’m proud of the work they do, and am not particularly interested in statistical comparisons. We measure our own work in this area, and I think we serve our users well.”

    I am sorry. You cannot claim you are better than your competition and then don’t publish stats. Red Hat publishes stats which is not comparative and you can certainly make do this.


    This is about honesty and integrity. Not competition. In fact, if you don’t measure how will you know whether you are doing better or worse by your own standards from last year?

    “I think Launchpad is a distraction from the issues here; it is in many ways an entirely separate project from Ubuntu”

    The issue is of Canonical decision not to work cooperatively with various upstream projects and keep their own code closed like launchpad. You say your kernel is close to upstream yet you have patched many many things. Why didn’t you spend your effort upstreaming them?


    September 19, 2008 at 19:06

  54. […] desktop packaging and installation, and its code contribtion is much more significant. So goes the reasoning of Canonical CTO Mark Zimmerman, who also complains that Kroah-Hartman was not prominently […]

  55. Lets look at linux contributions another way. If UBUNTU was not on the scene, linux would still be there in server land and only for server land. There would be no inroads in the less affluent countries.

    The linux kernel is ubiquitous, in that it serves the database and web servers, the desktop, the mobile phone and probably a whole group of process control equipment. I see Novel’s role as addressing the building of bridges to Microsoft, for Microsoft, with most of their development in that area.

    Fedora is developing new functionality, new drivers, attacks on poor performance, and the support for new hardware, new algorithms, etc.

    Mandriva, PCLinuxOS, Debian, Gentoo, and so many other distributions and developers are working in concert to bring linux to the masses. The English speaking world does not have exclusivity on intelligence. Development of linux in other then English is widespread.

    But I want to stress the following analogy. In a company, there is engineering, sales, purchasing, finance distribution, transportation management, inventory management, telecommunications and even more. No one person is an expert in all. By that example, UBUNTU has been superb in Sales and Marketing, and if you ask the man on the street to name a linux distribution, UBUNTU will be the first one to be uttered. And to make linux more user friendly and easier for desktop support, Canonical has done a tremendous job.

    My favourite distribution has not done as much as Canonical to get linux into DELL, ASUS and to the man on the street. My favourite distribution has to recognize that it needs to compete for the desktop. My favourite not Novel’s SUSE.

    Leslie Satenstein
    Montreal Quebec.

    Leslie Satenstein

    September 20, 2008 at 00:54

  56. I saw the keynote live. Of course, Greg KH is just trolling… It seems like it is working. :-)

    The easiest and best way for Canonical to respond is to just hire a couple of kernel devs, and let them loose. Just remember to follow the “upstream first” principle. If you want to really be taken seriously in the enterprise space, you need to do that anyhow.

    Jim Pick

    September 20, 2008 at 00:57

  57. Different companies – different approaches

    The open eco-system is very diverse and different companies plays different roles – that’s why it’s so powerful.

    Some contribute to the engine parts, some to the presentation layer, some wraps it all up and put it in nice boxes and distribute it, some gather the community.

    It is called specialization and is an integral part in an economy.

    Canonical/Ubuntus contribution to the parts Novell-Greg mentions would be to act as a “bug-proxy” and leverage their community to get input, consolidate that bug-information and keep track of the progress.

    As a “bug-squasher” Canonical/Ubuntu is doing a great job that all distros benefit from.

    Of course, I look forward to the day when Launchpad.net becomes open source – not because I’m interested in the code, but to many contributors out there it’s like Mozillas EULA – they don’t like it, and they prefer to choose something worse to click an “I-agree-button”.

    Open Launchpad up, stop bitch-slapping each other around and get on with business – lets grow this marketshare and aim for Ubuntu bug #1


    Niklas Andersson, TechWorld Open Source

  58. I’ve always said that Canonical took the work largely done by Debian and repackaged it for their own repos while adding on only slight improvements to vartous Gnome packages and adding some stuff like the upstart init system.

    I’ve also said that Canonical benefited in a large way from contributions by Novell (and earlier Sun) that attempted to address and improve a large number of the Linux Desktop’s usability issues.

    And people heard me and they misunderstood that to say that I was accusing Ubuntu of ‘leaching’ other people’s work.


    They are being damn SMART.

    WHY the hell does Debian go through social turmoil over licensing pettiness? Why the hell does the developers of the Linux kernel do through the hell of trying to make sure everybody that writers drivers for Linux knows about the GPL license and tries to get as many drivers into the kernel as possible?!

    Because they want to OTHER PEOPLE TO USE IT!. Debian talks all the time about end users and people modifiying and redistributing their work.

    Ubuntu is one of those people. It’s a perfect situation. Win-win for everybody.

    Not to mention making a good UI is FUCKING HARD. If programmers spent 90% of their time working on developing good user interfaces AND documentation for their software then that software is vastly superior to the same software if the developer spent 100% of the time working on new features.

    Ubuntu is able to make a usable Linux system with relatively good success and marketability because the massive amount of work that they’ve put into improving the user experience. This isn’t going to show up as code, because code isn’t as important as making software usable.

    THAT is what makes Ubuntu so wonderful and THAT is something that nobody else has been able to do so far with a 100% no-cost Linux operating system.

    Nobody else has been able to pull that off and everybody, including Greg HK, should give a large amount of respect for Canonical for being able to make something that is largely usable.


    September 20, 2008 at 20:06

  59. Actually, it makes perfect sense. Greg is a kernel guy. The only part that can be correctly called “Linux” is, technically, the kernel. Everything else is what’s been bolted on to make a more rounded operating system – also known as “bloat” to people in the embedded world. As techies and geeks, we are supposed to be pedantic about such definitions, and I for one refuse to succumb to the marketing love of fuzziness and misleading definitions. Linux is the kernel. One of several possible kernels. GNU forms the core of the O/S outside the kernel. Everything else pretty much optional and up to user preference.

    For a techy to argue that Gnome/KDE/etc contributions have anything to do with Linux is to fall into the definition trap. These are just GUIs on top of that sometimes help and sometimes hinder the user in getting things done. If the GUI is being tailored to particular kernel, then the GUI designers aren’t doing their job.

    As for the various distributions, they are just that. Distributions of other peoples work, sometimes adding something, mostly re-packaging then providing services around that package. All the distributions are basically providing a very useful consolidation of what many (most?) users expect to find on a modern computer. Each distro has it’s own ideas of how to do that, unfortunately. This means they keep on constantly reinventing the wheel and adding more and more branches to an already overcrowded ecosystem.

    I can’t comment about the rest of his talk, his patch numbers, or anything else because quite frankly it reminds me of the old Microsoft vs “Linux” marketing and political lies that keep flying around. Everyone trying to define and redefine things to to suit themselves and their agenda. I’m sick of it. Whether Greg is just turning on the Novell/Microsoft FUD or actually has a real beef – I don’t know. So long as he continues his excellent work on one of the many kernels available, and one of the most popular ones, then I don’t know if I care much. I do know that anyone who currently works for Novell/Microsoft, or any other company that cooperates closely with Microsoft, is not someone I’ll listen to without some serious filtering off possible intent.

    Now we devolve into a the disjointed rantings of an old and tired geek. Forgive me.

    When someone comes up with a distribution that allows you to EASILY (as in, no editing of multitude of obscure configuration files), select and install different kernels (to use Linux, a microkernel, BSD kernel, etc), select, install and use different GUIs (or none), and provides glue-ware to make it all work then I’ll be damn impressed. But right now I can only view the proliferation of different distributions to be more of an ego massaging wankfest than a solution to the problem of making non-Microsoft operating systems more mainstream.

    Yes, Ubuntu seems to be popular right now. But it still doesn’t install correctly on 1/2 the systems I’ve tried it on without me needing to edit config files and hack at it. Nor do Debian, RedHat, Slackware, Suse, FreeBSD, OpenBSD, or any other distribution of GNU/Linux or GNU/BSD or other open source systems I’ve tried so far. From not recognizing common, established hardware to weird DHCP issues, all installs have required editing of various config files and intimate knowledge of hardware and/or protocols. Nothing so far has just worked out of the box on 100% of my test systems. Systems that use standard, everyday, well-known components and are a mix of self-build and well-known vendor machines. Until THAT particular issue is resolved – work on what’s out there without user intervention – then we’ll continue to lose the battle. Unfortunately Microsoft software DOES install on those systems.

    I say unfortunately because for me the enemy is still Microsoft, and so long as all of us keep bickering about the little shit then the robber baron behemoth will continue to rule the ecosystem most of us depend on for our livelihoods. We need to get together and push for standard “best practice” ways of doing common tasks so that our energies can be spent actually innovating, not just reinventing the wheels that others have already invented. Sure, it’s ego-stroking cool to write something that others have already done and have people prefer your version over others. But does it help overall? Probably not. It just creates more polarization as people invest their ego rather than their skills into solving the problems.

    Yeah, I guess I’m still an idealist. Someone needs to be. Good night.


    September 21, 2008 at 09:33

  60. What a surprise. Someone points to a serious problem of Canonical/Ubuntu, which has been noticed by a a lot of developers from many projects, including X.org, gcc, the kernel, and instead of acknowledging the problem the best you can do is trying to stir up the zealots and the nutjobs (like schestowitz of boycottnovel.com fame) with this “he works for novell!!! he is an enemy of the people!!!” crap.

    If he attacked Canonical because he works for Novell, how comes that he repeatedly pointed up that RED HAT contributes many thing to everything? For fuck’s sake, Canonical contributes less than Gentoo, a free community distro without a single payed developer…


    September 21, 2008 at 12:08

  61. Novell, is just upset because the real Linux users don’t like their Microsoft position and avoid them like the plague (one wonder’s what dreams of Linux domination that Novell has that somehow Ubuntu”s existence threatens)?

    1- Red Hat Enterprise (the Cent OS version) I can’t even find dependable, or installable, Gwenview with Kipi Plugins AT all…. that an educational group wants to use in a poor school with the CentOS version of K12LTSP.org – Linux Terminal Server Project for Schools based distro. Otherwise the Red Hat stuff works great (they do a good job with the server side, but Red Hat is terrible with tending to Desktop Needs. Red Hat does SELinux and that is a great thing, nothing compares to it. RH does servers well.

    2- Ubuntu does a great thing with Desktop needs (with the Edubuntu layer it is pretty cool). That is until you load the LTSP install and it does not see 80% of the devices (including an old Jammin 125 device that was sold by the LTSP founder’s Disklessworkstation.com – does not work at all with Ubuntu’s LTPS and there are Youtube videos of other devices that don’t work with Ubuntu (that work fine with Red Hat K12LTSP without any modifications at all). So, Ubuntu as a server or a LTSP server is NOT GOOD. Ubuntu does not do servers well, and they really are missing the point in the commercial server space when they DO NOT embrace SELinux (so if they ever plan to make money, in the server support area, they better wake up and start loving SELinux).
    Ubuntu’s repositories to install Gwenview (add remove menu) installs the KDE 4.x version that is very broken and does not support Kipi Plugins, so Ubuntu does make big mistakes as well. Debian – if some app has Kipi Plugins as part of it, then the Kipi Plugins should be part of the dependency list for that application (why force newbies out to find Kipi… that does not make any sense)?

    3- Novell, they just want to beat everyone up and go to bed with Microsoft and poison Linux with Microsoft proprietary code, that they will force the Linux folks to then use ONLY Novell to get a system to work. They are as bad as Microsoft with their attempts to twist people into “lock-in” MS and Mono technology. I used to buy every Novell box version that came out… that stopped with they went to bed with Microsoft…

    4 – I have consulted or installed many Linux boxes… Servers get Red Hat (unless they can’t afford it, then they get CenOS, and I tell them that if they want to switch to Red Hat at some time that by using CentOS that it is pretty easy to do). Don’t use Fedora as it is way to unstable for ANYTHING these days…. Desktops on laptops and private machines, all get Ubuntu. When KDE 4.x grows up and can be used in about a year, then Kubuntu will offer some very interesting “mobility benefiting desktop/cloud Plazmoid eye opening flexibility” that no other OS desktop will be able to match.

    Where’s the beef (I think that Novell would like some of the Anti-Novell hate heat put onto someone else… when they fail to acknowledge that NOVELL deserves every inch of the disrespect that they are getting. If they stop sleeping with Microsoft, then they are welcome back in to the list of suggested OS versions here (otherwise, they don’t exist at all and so it is worthless to even mention them to any customers as you can not tell what radical rabid move that they will make next – like send their employees out into the field to FUD Ubuntu.

    Wake up Novell… doing this will just get you less respect than you have now. Stop sleeping with Microsoft.

    Red Hat – get your stuff working on the desktop. RH is not even in the game (I can’t understand why OLPC selected RH to do their desktop, as to RH the desktop is foreign to them to begin with).

    Ubuntu – get with the SELinux bandwagon or be forgotten in the server space. And make the LTSP stuff work with all terminal devices (vs just about 5% of them).

    Opps - Where's the beef?

    September 21, 2008 at 13:13

  62. You wrote: “Greg is failing to disclose his bias”

    The first slide says “gregkh@suse.de”. IMHO suse.de is pretty obvious.


    September 21, 2008 at 15:57

  63. My opinion is that both Greg and this comment on his presentation are wrong. Greg may have been negative, but I definitely understand that this commenter’s post is laden with its own negativity, no matter how carefully disguised behind diplomatic wording. Except for at the very end of the comments, where there is a clear appeal to people’s emotions by using terms like “trolling”.
    I hate it when people do that because it is always a weak position that shows an attempt to hide behind an “Us versus Them” position, sort of like saying “you are either with us or you are with the terrorists”.

    Ooh, that is so bad! Yeah, it is always bad to give criticism that does not agree with your own views, is what I read in it. So let’s put down anybody that dare to have the balls to find flaws so that they can be addressed and fixed. I mean, you must see the absurdity in sugar coating your feelings to try to coerce people? People are going to see through it, no matter how much you try to qualify your responses.

    Unfortunately, the only real thing that these comments do not qualify is the actual contributions that Canonical makes. Where are the numbers if there is a disagreement about their veracity? If it is upsetting to have people comment on the lack of them, why are they not quantified by Canonical? I am sure it is not that much problem to write a single line of text in a file somewhere each time there is contribution to keep an accurate and, hopefully, open record. Then Greg will have the right numbers to begin with if he decides to scrutinize Canonical again.

    Other than that, I think making negative criticism to negative criticism is a sign of poor character. People make criticisms all the time. If you don’t like them then make sure to engage, but don’t resort to silly diplomatic tactics that smell very bad to people who are straight-forward and not easily coerced (of which I think many Linux developers are the same type).

    I have to agree with another poster here in that you are just trying to stir up the Ubuntu zealots and the nutjobs . Which is not hard to put past a Canonical employee. So it is not a very credible rebuttal to the presentation at all. If were coming from an independent knowledgeable third party, then it might carry some weight.

    Raymond Martin

    September 21, 2008 at 16:28

  64. Hmmm…

    linux, the kernel is certainly important… but the problem seems created by the monolith nature of linux (kernel) in the creation of Linux (GNU-systems)… linux is more kernel+drivers, and besides advantages it also posses it’s own problems.

    But since the size of Ubuntu is so small, any demand would be relative. So the problem seems to be the success in the choice of users. Well… if so… call users to patch the kernel. Not the best idea and against the GNU philosophy. Is linux (kernel) changing it’s goals?

    I would like very much to see BSD-Ubuntu using a driver compatibility layer with, huh, linux. BSD has a new pretty good kernel… but it would not be practical to drop a lot of work… but never say never! It would be an interesting exercise, and would help to keep the dream unaffected by saboteurs of the dream GNU started.

    The dream is more valid than a patch. So be aware against taking-overs.
    What is valued always demanded eternal vigilance.

    Dutra de Lacerda

    September 21, 2008 at 23:10

  65. It’s things like this that make me proud to be a die hard slackware user. I can’t believe people still write crap like this. Who cares?


    September 22, 2008 at 01:38

  66. Greg KH quantifies exactly what many people in real distro’s think and know to be true. Ubuntu is all there for the purpose of making Mark S the next few hundred millions. The value of a clone of a distro (as ubuntu is a clone of debian) is very low. Clones of Clones of Clones (ubuntu derivates) even lower. Ubuntu will fade in time with people realising that all the hype that you all generate is just that: Hype. RedHat and SUSE do things properly as opposed to just talk a lot about nothing special to generate hype. Ubuntu is damaging the Linux business deployments, since it is an inferior quality distribution driven my amateurs, who do not contribute anything to the community. Greg KH is correct and shows the numbers, no point whining you all … it is just a fact that giving free CD’s away ain’t a business model or ain’t helping coders. Ubuntu’s demise is just a matter of time.

    Andy Linux

    September 22, 2008 at 04:29

  67. Some practical observations.

    At my workshop I install Linux distros on a variety of older refurbished hardware, which is then donated for a small hardware refurbishing fee to the digital divide. On anything with at least a P3 1GHZ 256MB, Ubuntu is the distro of choice. Anything less gets Vector Light. In the last 3 years this has amounted to over 1200 Ubuntu and 400 Vector systems systems on hardware from all major vendors and Asian clone makers.

    I also do installs for modern hardware, the latest being a Dell XPS M1330, set up with multiple partitions to test the latest distros from all and sundry, and an AMD 64 clone to test various 32bit distros for usability with multimedia applications and peripherals.

    I can say without reservation that out of the box Ubuntu performs flawlessly on all older and modern hardware that it gets installed on in my workshop.

    On the recent hardware described. Fedora and Mandriva also performed well, albeit with a bit of network setup tweaking for wireless, but OpenSUSE had a lot of trouble using the inbuilt Intel wireless rather that the inbuilt Ethernet., on the Dell laptop, and both Fedora and openSUSE couldn’t use older webcams without compiling a driver, an impossible situation since the supplied kernel headers didn’t seem to match the installed kernel. This is not a problem with compiling drivers for Ubuntu kernels.

    Notice that this is all Kernel and boot processing stuff, so apart from the end user choice of Ubuntu for its ease of use package management, it wins hands down for trouble free installs on disparate hardware. This must be due to Canonical making sure every possible hardware driver is included and that they all work.

    Since the same cannot be said of all the other major distros tested in my workshop for hardware compatibility (I do try to test the very latest), I surmise that the drivers have been patched by Canonical for their distribution and sent upstream to the kernel org,, but these haven’t yet got into the kernels being supplied by the competition.

    So even if Canonical are not doing so many low level patches, they must be doing more than the others with driver patches to make sure their distro works out of the box on a wider range of old and modern desktop and laptop hardware.

    This is exactly what Mark Shuttleworth said Canonical would do, and my results affirm that they have done it. Until the other major distro’s catch up on this very important kernel usability factor, my users will continue to love using Ubuntu on their hardware.


    September 22, 2008 at 07:05

  68. @Andy Linux:
    That is a sophisticated form of conspiracy theorizing. I’ll give you that ..


    September 22, 2008 at 09:01

  69. “””So even if Canonical are not doing so many low level patches, they must be doing more than the others with driver patches to make sure their distro works out of the box on a wider range of old and modern desktop and laptop hardware. “””

    That’s why people complained that Canonical don’t contribute back to the linux community.

    They have patch for their own system. They don’t want any other Linux distro to install as flawless as their own ubuntu. So every if they spend the time and resource to fix driver bugs, they don’t feed the patches back the main kernel. This sucks.


    September 22, 2008 at 11:39

  70. Have to agree with Matt on this one. Canonical’s contribution is of a different focus, and they admit they didn’t enter the Linux game for kernel development. Indeed, why start there! It’s obvious Greg KH had a beef to pick with Canonical, and isn’t over it yet.

    Whether you use Ubuntu or not, you should peruse its forums to gain knowledge! See http://www.thegsblog.com/?p=251


    September 22, 2008 at 13:13

  71. Ubuntu is not contributing as much as others? Where’s the problem?
    Ubuntu says it’s contributing more than RH to the linux kernel? That would be bad!

    I didn’t count a lot of patches from GKH to KDE, so what he does is certainly bad….

    And all of you haven’t put a single line of code in myProgramIsTheBest.py, so don’t call yourself hackers!

    It can be quite difficult and time consuming to get patches incorporated in the kernel. And you have to count on the ego of the maintainers as well. (Like the guy who offered a new scheduler, got blasted and saw the maintainer “re-invent it” and write the same thing, just a bit buggier ….)


    September 22, 2008 at 14:09

  72. Greg is absolutely right.
    Here is article of different author.
    Canonical does not give much to the Linux community


    September 22, 2008 at 14:23

  73. Why are most people picking on Greg here? He was only responding when Ubuntiets picked on him for not including Ubuntu’s contribution. Well, he only showed how much Ubuntu’s contribution was to the community.

    Ubuntu is free to use GPL code and newbies are welcome to use it as their first distro. But please don’t expect the community to fall at Mark’s feet for “expanding Linux market share”.

    Heck, Ubuntu must be smoking pot if it wants to be considered equal to Red-hat and Novell / Suse. All Ubuntu does is use upstream code and apply dirty patches to make Linux easier for home/Soho user.

    As a Linux user who was introduced to FLOSS via Mandrake half a decade ago, I know how much of a help distros like Ubuntu are to newbies. But please dont equate Ubuntu with Red-hat and Suse, they belong to a totally different plane. And unlike Ubuntu, they work with the developer community to ensure their work is also useful to others. And they don’t claim to be what they are not.

    Ubuntu has a ugly face. Greg, like a few Debian developers earlier, had the guts to point it out. Responses from Ubuntu developers so far have been rants against Greg rather than meaningful responses.


    September 22, 2008 at 14:28

  74. Oh, gosh, Ubuntu should just switch to BSD maybe then. Linux’s desktop share will plummet by a third then right off the bat ;)


    September 22, 2008 at 14:53

  75. Novel can’t take me (a nobrainer + ex-Windows lover) to the Linux World or ‘Ecosystem’ of the Linux community, but Ubuntu can.

    Learning Linux is a big pain for any nobrainer like myself, but Ubuntu is not that hard especially when the great Ubuntu community is standing with me, I’ve learnt a lot, have been helped a lot by others, now, I can even pass on my Linux experience to the newbies.

    Ubuntu may be a shit distro. to you geeks, but it’s really means so important to me.


    September 22, 2008 at 15:38

  76. How about you guys get your s*** together with Dell and release that freaking Mini 9 with Ubuntu?
    its nowhere to be found (except US where its in pre-order limbo) on the planet while the XP version is.

    On top of that your friends at Dell say sweet things like you can ‘upgrade’ to XP or call Ubuntu a mini-OS both indicating that Ubuntu is a toy that you can get rid of and get a real OS. Who needs enemies with partners like this?

    What some jealous dev says at an uber-geek fest wont make it past our ears.
    The phenomenal failure that is the Dell/Ubuntu Mini 9 collaboration will.

    Many reviews now that I have seen (just read an ozzie one this morning) dont even mention that there is an other option that XP.

    Worry less about jealous enemies and more about your partners.

    Most netbook sites are filled with people who are going to the Acer One instead because they are tired of waiting (Acer will sell 1 million netbooks in September) and Ubuntu is getting its fare share of the blame for this since the hardware is out which they associate with Dell and the software (Canonical) is not.
    Sure, many will be buying the Mini 9 and putting the Ubuntu Remix on it but the reasons why its better not to do it are clear.

    To make it worse, there has been no announcement, nothing but speculation about to why and when (or if) the Ubuntu version will be released.

    Ubuntu is missing and no one knows why.

    Maybe you should worry about that a bit more. Or who knows, maybe let the people know WTF is going on.


    Guy lafleur

    September 22, 2008 at 15:46

  77. Without Ubuntu’s user-friendly features, Helps from the Community, and Canonical’s successful marketing, I wouldn’t step into the world of Linux, wouldn’t have correct attiude to learn Linux, wouldn’t be able to learn any essential Linux skills.

    Now I can easily learn to use Debian, Arch, and many other distros. with no pains, I can freely swtich to other popular distros. with ease.

    But now I probably won’t try any Novel’s product again!!


    September 22, 2008 at 15:58

  78. Ubuntu is a great distro for people to start and get used to Linux. I started off with Mandrake sometime in 2001-02, dual booting XP, and frankly it blew me off. Many new Linux users feel the same when they start with Ubuntu today. Mandrake was in 2001-02 what Ubuntu is in 2007-08.

    But today, configuring a GNU/Linux system is so easy even with a puree distro like Slackware, unlike back in 2002 when Mandrake had to do a lot of patching and scripting work on top of vanilla packages to make them usable to a newbie. They used to a fantastic job, and frankly, i’m still surprised their “Control-Center” approach was not widely adopted by the FOSS community. Its sad Mandrake could never capitalize their brand well for all the work they had done.

    Still, even in its heydays, Mandrake folks had the modesty to say they were only building on top of what is already available, and never claimed to be saviours of Linux. Back then, if Mandrake had moved on, Linux would have lost 1/3rd its user base too! But they didn’t claim to be doing a great service to FOSS community , unlike Canonical’s Boss and some Ubuntu developers.

    Modesty seems to be missing in Ubuntu. Its sad to see Greg being chastised for his presentation, when in reality, he only presented facts and figures when some Ubuntu “supporters” derided him for initially ignoring Ubuntu.

    So much has been made of Greg working for Novell, one would assume Greg only works on Novell projects, or that his contributions to FOSS start and end with Novell. Heck, Greg continues to maintain Gentoo packages even today! Does Ubuntu not use Greg’s code in Ubuntu??????????????????????????????? His list of contribution is so long, that it looks like his individual contribution may surpass that of all Ubuntu combined.

    Is such a man not entitled to his personal opinion without it being attached to the company he works for? Is he not entitled to get a little angry when someone picks on him at a technical presentation for not including some minor distro’s contribution?

    Matt even mentions that Novell maybe afraid of Ubuntu more than it is of Redhat! Matt, please wake up and smell the air. Novell freed up Greg of all other responsibilities to let him focus full time on Linuxdriverproject, which will be used by every Linux distro in the world. And you are claiming that Novell is somehow trying to sabotage Ubuntu’ reputation through Greg’s presentation!.

    Ubuntu team is looking more like a cult, that has a bloated estimation of its self worth.

    PS: My earlier post was under the name “ashyanbhog”, updated my wordpress profile after that.


    September 22, 2008 at 17:52

  79. Interesting commentary on the whole thing…

    I didn’t go to LPC, or watch the video, and am not able to read the LWN article either (no subscription). I did, however, read this article and presentation and commentary on Greg’s blog (http://www.kroah.com/log/linux/lpc_2008_keynote.html); and it seems to me that…

    1) The remark on ‘5 or 6 patches’ was likely an off-the-top-of-his-head guess, and he was simply correcting it here with the best data he could find for what he looked at.


    2) He is making the case for anyone attending the LPC that they should not let their projects (or employers let their projects) become solely dependent on an upstream provider; whether that employer is Canonical or someone else; he just used Canonical as an example since he already had some data…

    So then we come to this comment in this article: “our kernel consists almost entirely of code we receive from upstream.”

    This is exactly what Greg arguing against. He was basically saying “don’t rely on upstream; be your own distro/project, and provide patches back up”.

    So okay, Canonical/Ubuntu is okay with being solely dependent on upstream and thus not sending much in way of patches up. But the normal developer (in Greg’s estimation) shouldn’t be and should be pushing to be able to (when he/she can) push patches up-stream to lighten their long-term support load – it would benefit everyone.

    And with that I can agree. Now perhaps he should have abstracted it away from any single contributor sponsor to help get the point across (likely a more professional thing to do), but that seems to be the jist of what he was (at least) trying to say, and that is a message that should have been appropriate for an LPC keynote – communicating to everyone how to be a player in the entire ecosystem, helping lighten their work load, and everyone else’s too.

    Now, as far as Canonical/Ubuntu and similarly situated developers/companies being a downstream and relying on (mostly) Debian (or whoever) – perhaps the solution is for Ubuntu to deliver more patches through Debian (or whoever) instead of directly to the project maintainer (or to show they are and that their upstream source is saying ‘no’). This would have the single advantage of removing (a) duplicate patches submitted by the upstream provider as well, and (b) creating a patch which depends on an upstream provider’s patch that may or may not be submitted. Either way, it would help with (and work towards) what jist (so far as I can tell) of what Greg was trying to say (IMHO).


    September 22, 2008 at 18:36

  80. Matt,

    don’t worry about all the haters.

    First they ignore you,
    then they laugh at you,
    then they fight you, ( <- you are here )
    then you win.



    September 22, 2008 at 19:32

  81. Here is typical reply of Ubuntu user (see many similar above)

    >Without Ubuntu’s user-friendly features, Helps from >the Community, and Canonical’s successful marketing, >I wouldn’t step into the world of Linux, wouldn’t >have correct attiude to learn Linux, wouldn’t be able >to learn any essential Linux skills.
    >Now I can easily learn to use Debian, Arch, and many >other distros. with no pains, I can freely swtich to >other popular distros. with ease.
    >But now I probably won’t try any Novel’s product >again!!

    Look at this numbers
    Novell and RedHat make the Linux. You simply can’t use any linux distro (including of course Ununtu) without using Novell’s work. So please first find out what you are talking about and only then say that you won’t use something.

    The question is not in how many users Ubuntu attracted. The question is that Canonical use work of other companies and does not give back anything. What if RedHat and Novell will do the same? Linux will stop to evolve. That’s the main point. If you use work of other people for free you must respect their work and give something back.

    Another point is about “Ubuntu’s user-friendly features”. If you try any other Gnome based distribution you’ll find that all of these features come from Gnome (the upstream) but not from Canonical. That’s the point.

    All Canonical is doing is just marketing. And that’s why free software world does not respect them.


    September 23, 2008 at 07:22

  82. […] vs. Community? Jump to Comments On Gregs remarks about Ubuntu not giving back to the community and the reactions to it: Funnily I just did organize this SFD and just wrote that […]

  83. I liked this rebuttle… kind of a “don’t be evil”-sh language.

    In any case, what Canonical primarily does for Linux is make it popular. This helps feed a nice cycle of increased demand for the OS, increased demand for developers (across the board – RH, Can, Deb, etc…), furthering the development of the kernel and surrounding systems, resulting in a better system, and more demand.

    Why must each major player in this dynamic look the same and play the same role? The whole of the Linux environment (OS, users, dev community, etc…) is just too large for this kind of view.

    Brian van Doren

    September 23, 2008 at 18:20

  84. It’s been interesting reading the comments here. I got through half of them and it seems the same arguments are getting repeated, and to be honest they scare me. So, I thought I would just post a comment and call it a day.

    Since when are FOSS sponsors and contributors so touchy about “proper credits” and “contributions”? Really there is only one way to solve this argument; close it all up…

    This is the only way you would ever control people to use “proper credits” when using software you were involved with developing, and certainly the only way to keep other distributions or companies from “using but not contributing”. It’s Open! This is going to happen. If it bugs you that someone may be making millions off of your hard work and you are not then don’t do work in the FOSS world.

    I guarantee that if we start making an issue out of this nonsense it will only hurt everyone involved with FOSS. Remember your core values, swallow a bit of ego, be happy with your own production and contributions, and forget what someone else is or is not doing \\unless of course they violate a license, and there’s no claim of that here\\. Attitude is everything, and if we start bickering with each other it’ll only take longer for FOSS to rule the world.

    I’d rather be reading the same old brainwashed argument about how Linux sucks compared to a generic Spanish view than read about petty squables between communities that in the end have the same goals but a different road map to get there. We’re a family, let’s forgive each other and forget about it. Besides, my laptop has issues with suspending still, so don’t you people have something better you could be doing? :-)


    September 24, 2008 at 22:49

  85. […] the scope were limited to the Linux kernel; Ubuntu CTO Matt Zimmerman is quite candid on the point, saying: Canonical is primarily a consumer of the Linux kernel. It is one of the building blocks we need in […]

  86. +100

    You do not have to worry about corporate Oracle shills (what essentially RH and Novell are).

    Keep on doing your thing, guys.


    September 26, 2008 at 16:33

  87. Well said Matt and Mark. You are doing a great job. Please continue to make Ubuntu better.

    Irresponsibility shown from Greg. If I was his bossat Novell I would tell him so. He is not helping Linux with talk like this.


    September 28, 2008 at 11:23

  88. I’ve been an Ubuntu user since day 1.
    Instead of repeating what so many people have already
    said, I’d like to say only this:

    1. Send your patches upstream instead of hoping for
    upstream developers to pull from you.

    2. Hire a few kernel developers.
    Hire Greg Kroah-Hartman if you have to. :-)


    Yesudeep Mangalapilly

    September 28, 2008 at 12:08

  89. I have a theory about this…

    You know how GKH mentioned that there are companies (Novell) that are looking to hire developers, also implying that Canonical is “bad” and people should quit it?

    (putting on tinfoil hat)
    Novell’s deal with the devil has made them unpopular among quite a lot of people, especially those inclined towards working on Free Software.
    Maybe they are now stuck with a large wad of cash, looking to hire people in their ongoing attempt to be “bigger and better” than Red Hat, but they’re having problems finding applicants who are experienced in free software development — since those are the exact
    same people who dislike them!
    Hence the attempts at vilifying Canonical, trying to divert attention to them, all the while showing that “we are second best” in all those charts…
    Why else would he raise fact that Novell are looking for people, while supposedly complaining about Canonical’s inadequacies?
    (taking off tinfoil hat)

    It would also be interesting to see his patch-count charts normalized per the profit each company makes off of Linux… (I certainly don’t see the point of normalizing per total # of employees of the company
    — the size of the kernel team would make more sense.)

    (not an Ubuntu user myself, BTW.)


    September 29, 2008 at 12:36

  90. Ubuntu’s focus is a well integrated desktop; Redhat’s and Debians are not. Off the top of my head, they’ve contributed bazaar and upstart, which reflect that focus. Its popularity ON THE DESKTOP is a reflection of that desktop focus…

    AFAICT Canonical is also tiny compared to, say, Redhat and Suse.

    The unpaid workers of Debian and Ubuntu are kind of irrelevant to the whole discussion. They are free to contribute to whatever they want to. Unpaid Debian developers who happen to alao contribute to GCC could just as easily change over to being unpaid ubuntu developers who also contribute to GCC. I don’t believe there is a corporate mandate from Canonical NOT to contribute upstream, and AFAIK there’s not a specific mandate for Debionites TO contribute upstream.

    Honestly, if your focus is the desktop, then the kernel and gcc are basically good enough. Linux desktops are still a bunch of half finished crap.

    Darryl R

    September 29, 2008 at 13:25

  91. […] den Benutzer mittelfristig nichts Gutes bringen wird. Natürlich gab es kurz darauf entsprechende Gegenreaktionen von Ubuntu-Mitarbeitern. Die Diskussion geht mittlerweile auf LWN weiter, es gibt auch Artikel, […]

  92. Thank u


    October 5, 2008 at 21:03

  93. […] http://www.kroah.com/log/linux/lpc_2008_keynote.html – Matt Zimmerman, CTO for Canonical, responds: https://mdzlog.wordpress.com/2008/09/17/greg-kh-linux-ecosystem/ – Points out Greg only considers kernel, X, and compiler contributions, not Gnome, KDE, or anything […]

  94. Linux


    October 12, 2008 at 20:33

  95. Does anyone still remember how was SUSE before being acquired by Novell? I still remember.

    SUSE was bought but didn’t seemed to go nowhere for some time as if Novell didn’t knew very well what to do with it, and only then we saw a really OPEN SUSE (Yast) and a really OPEN Ximian Evolution. Fedora was almost dropped by Red Hat if someone still remembers the discussions at the time.

    RedHat dropped the common user for the more profitable corporative ones and was so “closed” that CentOS appeared.

    Before Ubuntu appears in the scene the easiest of the big ones (those that give professional support) for the end user was Mandrake/Mandriva and in no way the effort of having releases often were a motto for anyone around the big distros.

    Sometimes I hear this discussions and it seems to me that everybody eats too much cheese.

    Mário da Silva

    October 22, 2008 at 02:39

  96. […] Linux) ao tamanho da contribuição da Canonical ao ecossistema Linux, sobre as quais já há uma série de outras respostas […]

  97. […] sobre la participación de Greg en su plenaria. y aquí les dejo muchos de sus comentarios de su post original Quizás el nombre de la charla esta errado y veamos por que: Greg Considera que el […]

  98. […] contestaron posteriormente a la crítica de forma razonada: En la primera respuesta, se justifican las motivaciones de Canonical y se ponen en duda los números de la charla. En la segunda respuesta, se cuestiona la presunta […]

  99. […] Quizás no venga al caso, pero Greg Kroah-Hartman es empleado de Novell; ojo al piojo. No voy a tirar al piso ahora lo que he recomendado como lectura, pero éste y otros datos (como por ejemplo, la relatividad de lo que se considera un ecosistema Linux) se explican un poco mejor en éste artículo de Efrain Valles y en una de sus fuentes, Matt Zimmerman (empleado de Ubuntu) y su artículo en respuesta a la charla de Greg. […]

  100. […] time, Matt Zimmerman replies, noting that while Ubuntu has lots of users, it simply does not have a lot of developers working on […]

Comments are closed.

%d bloggers like this: