Submitted by Greg KH, Linux kernel developer delivered a keynote in the Linux plumbing conference about the health of the ecosystem. His message was essentially that distributions that don’t contribute to the ecosystem have to rely on the whims of others which is unhealthy for them. Here is an introduction the development model and some interesting statistics about the Linux kernel code. Update by TH: Rebuttals are appearing all over the web, like this one by Canonical’s Matt Zimmerman (“He’s refuting a claim which has, quite simply, never been made. […] When this sort of thing happens on mailing lists, it’s called trolling.”), or this one by another Canonical employee, Dustin Kirkland.
Is none other than Broadcomm.
For a company that on the surface seems to think that the word Linux does not exist and encounters the wrath on many a Laptop owner who can’t use the WiFi networking because it has a Broadcomm chip (without lots of effort that)it is surprising that they are openly contributing to Kernel Development.
That is probably for embedded Linux, and not for getting their drivers running on it.
I thought it was quite nice and reassuring to see how many independent developers are contributing to various software projects.
Broadcom also buys a lot of companies, so a lot of that could be coming out of established operations of places like ServerWorks. I don’t know how much any of their acquisitions actually supported Linux development, though.
All that tells me is to never use a piece of GPL’d code, otherwise I might get self-righteous people telling me what to do with the freedom I’ve been given.
Targetting Canonical with this blind zealotry is like cutting one’s nose off to spite your face.
I’ve got retorts, I’ve got articles I can link to that debunks this whole talk – but I can’t be bothered. It’s like talking to a brick wall.
These people can go off, and keep chasing that kernel rainbow that’s so important to them; that’s their freedom. Their freedom. Not yours.
That’s nice, but they have every right to do this, and IMHO are right on track with it. Sure you’re not “obligated” to give back completely (I’m not talking about “making changes available” but “helping get them to base”), but you DAMN WELL SHOULD.
There are people who in their good will and free time put loads of work into the thing you’re using for free AND making money off, it’d be nice if when you DO fix or enhance something you go to the extra little bit of trouble and try to get your changes backported.
Canonical SHOULD be ashamed of itself in this regard. I love Ubuntu, in fact I’m using it now, and I’m not fanatical about this so I won’t stop using it for this reason but I’m disappointed in them for being so lazy in this.
I mean they are all “marketing bang” happy about “we’re going to invest in the ecosystem and GUI” lately but giving back their own kernel changes and enhancements seems like an impossible chore for some reason.. doublespeak perhaps?
Shuttleworth recently invested millions into FOSS – again. What exactly is it that Canonical should be ashamed of? That they’ve invested more money into FOSS than you ever will?
Well what millions is that then, where, into what FOSS?
The fact of the matter is ordinary developers of their free time contribute more than them. Ubuntu is a great distro but they need to put more effort into contributing more back to projects code,it’s really that simple considering how popular it is.
I’d like to point out about in the update that person puts up employees, which is completely wrong. If only 130 people contribute to Linux at IBM then it’s about the same, it’s about how many contribute in the company not how many are employed.
Whats how many years it’s been going got to do with it, it only takes a few people to contribute patches at a fast rate.
Edited 2008-09-21 10:43 UTC
Did you not follow the news? http://osnews.com/story/20282/Canonical_to_Fund_Upstream_Linux_Usab…
It was even on OSNews.
Or is it that you can never be pleased no matter what they do?
Edited 2008-09-21 10:41 UTC
Which has only just been announced, we’re not talking about employing professionals we’re talking about what they contribute code wise.
Still more free time people like me contribute artwork than them, I dont get a page on OSnews about it.
Edited 2008-09-21 11:01 UTC
There’s more to software than code.
Zealots have a one-track mind. That plays over and over.
And why does it matter whether they code themselves or whether they hire others to code?
Ummm…artwork? We’re talking code, not changing the color of the background and thinking it makes you more special than a company that does, in fact, contribute code and money.
You should quit with the trolling.
> Did you not follow the news?
> Or is it that you can never be pleased no matter what they do?
Fine. That tells us that they invest into news and PR as usual. Or do you assume we will see this time some kind of result upstream?
“Ubuntu is a great distro but they need to put more effort into contributing more back to projects code,it’s really that simple considering how popular it is.”
Projects need more than “code”. Publicity ain’t free, exposure ain’t free.
Also, DOES ANYONE HERE BELIEVE THAT UBUNTU/CANONICAL IS MAKING MONEY? Not me, I think that they’re bleeding money (I believe even Shuttleworth announced it)…
They are bleeding money and Mark has had to contribute money to the project three times as far as I can count. That’s 30 million dollars out of his own pocket to a project which has still to make a red cent. Suse and RedHat have the luxury of having rich parent companies that had infused them with cash. They both sell their distro commercially while ubuntu has stuck to their guns that their distro would always be free. They could have capitilized on the popularity of the brand a long time ago and sold the distro. In-fact only recently have they started packaging Ubuntu in retail locations.
Canonical has one guy giving away his own money to a project he believes in, Suse has Novell, RedHat had TimeWarner at some point. He decided to focus his distro on what they can do best which is packaging and they seem to be doing it rather well. They don’t have to contribute all that much to the kernel because they use almost the exact same kernel that Debian does, they contribute bug reports and fixes to the debian and they in turn contribute patches back to the kernel in necessary. I don’t see the issue here.
This is a stupid argument, and just shows sour grapes. So what if Ubuntu is more popular, instead of complaining about how about doing something about it. How about seeing what exactly is it about Ubuntu that makes people new and old want to use it.
Mark Shuttleworth has a business plan. It may or may not be making money now, but you can bet your bottom dollar that he is expecting it to some day. He is not infusing money into the project as an act of charity, but as an investment.
As for the benefit of the Ubuntu marketing, if there was no Ubuntu, some other distro would take its place as a darling of friendliness.
So? He’s investing in something he believes in. There are a lot of easier, faster, and less risk-prone ways to make a decent profit from one’s capital. While I’m sure he’d like to see a return some day, if he were “in it for the money” he certainly wouldn’t be heading up a Desktop Linux company.
It used to be Mandrake. But they developed a long-running reputation for poor quality, made some incredibly bad business decisions, went into the French equivalent of Chapter 11 bankruptcy, and lost the throne while they tried to sort it all out internally. Ubuntu is doing a lot better on all fronts than any distro which has attempted to blaze this trail before. That generates a lot of jelousy among users of distros who wanted their distro to achieve the popularity that Ubuntu enjoys today.
See my previous thread on how people hate it when that happens, and how they react. It’s all so very transparent and predictable.
Edited 2008-09-21 20:56 UTC
Was just arguing against the opinion that Ubuntu is a philanthropic project.
There is nothing wrong with being for profit, but the poster I was replying to was suggesting that as an excuse to give Ubuntu extra leeway.
Disclaimer: I am Vista User who also dabbles in Fedora. I have used Ubuntu, but do not plan to again, nor would I recommend it to others as I am of the opinion that in the medium to long term it is bad for the floss ecosystem.
Exactly. I think you hit the nail on its head there.
Many people attack Ubuntu a lot all the time – just because they would like to see something else enjoying its popularity: SUSE/Novell, Fedora, Mandrake, BSDs etc.etc.
Heck, also I may have been guilty of that a few times when Ubuntu was still young and I hoped that they would just concentrate on improving vanilla Debian instead of developing their own distro. However, later I had to admit that the big Debian project seemed all all too slow and cumbersome a ship to turn so Ubuntu’s choices started to make more sense to me gradually.
Red Hat and Novell are really popular in the enterprise — unlike Ubuntu.
When you read “The Linux Ecosystem, what it is and where do you fit in it?” you’ll notice that this keynote was an answer to questions by Ubuntu fans. Yet, even tough it was kinda “provoked” by Ubuntu users, Cannonical employees cried foul and started to attack Novell because they didn’t like the answer.
Sane and sober people who have been paying attention notice no such thing. They notice an unprovoked attack by Greg K-H and the resulting food fight which he successfully provoked.
I suggest that everyone ask themselves whether this “story” and the controversy surrounding it is worth our wasting any more time and effort on.
You are kidding, right? You are so “sane and sober” and didn’t even bother to read the text?
Edited 2008-09-22 09:52 UTC
Markus,
I clearly said “sane and sober people who have been paying attention“. Paying attention implies being familiar with the whole sordid story, from Greg’s first gratuitous (and erroneous) attack, which is where the story really began, through the present.
In paying attention to this thread, I can’t help but notice that you are doing everything you can to stir things up. I cannot help but wonder about your motive.
But I suggest, yet again, that this non-news that Rahul felt was so newsworthy just isn’t worth wasting any more time on. Unfortunately, judging from history, it seems likely that Greg is going to continue to stir the pot every chance he gets. And I’m sure that people will take the bait every time.
Red Hat has a parent? Red Hat got successful on its own.
SuSE also was successful before that was bought by Novell.
Huh? Red Hat gives away Fedora, and Novell gives away openSUSE. Both Red Hat and SuSE were given away on magazine CDs since the 1990s.
Red Hat and Novell moth sell enterprise support contracts, but that’s no different than Cannonical.
I can’t remember ever having seen one but maybe I only missed seeing those magazines?
Anyway, SuSE was long described as partly a non-free distribution by sites like Distrowatch. There were no free SuSE installation ISOs to be downloaded anywhere, but you could install SuSE by using the more difficult and time consuming ftp installation method. OpenSUSE (and its commercial variant SLED) came only after Novell had bought SuSE.
Edited 2008-09-21 23:56 UTC
Obviously. I moved just a few months ago and threw away lots of old CDs. Magazine CDs with S.u.S.E. Linux were among them.
YaST was open source, but not GPLed in the past.
That are things of the past anyway. Today’s business model of Cannonical and most other distributors like Red Hat are similar: Sell support contracts to enterprises.
Red Hat knows that to fulfil those contracts, it has to employ many developers. Every Red Hat Enterprise Linux is supported for seven years, after all. Somebody needs to do all that bugfixing, backporting, etc.
Cannonical OTOH wants to sell support contracts as well. But as Ubuntu supporters pointed out: Ubuntu’s contribution is to report bugs and not to fix them. See http://osnews.com/thread?330932
So who does the bugfixing? Well, most of the time that are employees of Red Hat, Novell, Nokia, IBM,… who are part of the upstream project.
RedHat and Suse did get successful on their own, initially. Just like Ubuntu, they were popular because they were community based projects that made it easy (relatively) for the average user to install linux without having to compile everything from scratch. However both distros, though they contributed, did exactly the same thing that Ubuntu does now, they were integrators, they didn’t start really pushing contributions until after they got parent companies with huge amounts of money to back their projects. RedHat being the first to get this financial backing and already having dealings with IBM has contributed the longest. To top it off Suse, RedHat, and Mandrake (now Mandriva, stupid name) though they gave away their distro for free still sold their products through retail. The versions they gave away were usually “community” version which had things missing from it.
One of the main reason’s Ubuntu is so popular (at least in my eyes) is that they give you a basic 1 cd install (while the others used to give you a DVD or 5/6 disc) and EVERYTHING else you need is available via the repos, which are already setup for you and ready to go (which wasn’t always the case, I’m glad they changed that). I’m glad to see other distros are following suit. I downloaded Fedora the other day and was able to download one live CD instead of a DVD’s worth of crap I won’t use. The idea isn’t new but Ubuntu took the concept polished it up and ran with it.
Which is what Canonical really does. They are integrators, distributors. They are basically a more popular version of that guy who puts Linux Mint together or any other small scale distro out there that focuses on what they want their distro to be like and devote their resources to it. Tell me how many kernel patches do those guys contribute back? I can tell you, not many. Don’t let Ubuntu’s popularity and enigmatic leader fool you into thinking that Ubuntu is anything more than just some scratch your own itch distro that just happened to scratch a whole bunch of other people’s itches.
I dont see RedHat giving away RHEL or SuSE giving away their equivalent.
Every openSUSE X.1 release corresponds to a SUSE Linux Enterprise X.0 release. Eg. SUSE Linux Enterprise 11 will be based on openSUSE 11.1. The number of supported packages is smaller, eg. SUSE Linux Enterprise 11 won’t include KDE 3.5, and a default desktop is selected. Other than that openSUSE 11.1 and SUSE Linux Enterprise 11 will be pretty much identical.
That being said, Cannonical’s so called “LTS” releases only offer three years of support for desktop apps. Every openSUSE release offers two years.
Red Hat goes a different path with Fedora. A Fedora release is more like an early beta release of Red Hat Enterprise. Red Hat releases all source code of Red Hat Enterprise, even of BSDLed apps that don’t require a source release. While not official, there’s CentOS. RHEL/CentOS is fully supported for 7 years, not just lousy 3 years for desktop apps and 5 years for sever apps.
130 work on Linux….do you mean Linux the K]kernel or do you mean 130 total work on porting Symphony and Lotus and all their other software to Linux? I really doubt it’s the latter. They’ve probably got a lot more people than that working on the userspace software. And that’s most of what Canonical does too. They just can’t do it on the scale that IBM does.
You obviously don’t get it…
Money is of no concern here. We’re talking about developers doing the right thing for the right reasons which in the end benefit everyone.
I also work on one FOSS project, and there are basically two kinds of people:
1. People who when they fix or enhance something, rush the changes upstream (patches or direct commits if they got the right)
2. People who don’t even say they got something fixed or enhanced and rather use their own copies etc.
#2 usually makes more money (mostly because they don’t care ABOUT the tool, they just want to get the job done), but is more destructive than constructive. We need more #1.
Now don’t get me wrong. I like Canonical, can’t really say if I like Shuttleworth, don’t know the guy. Anyway I’m not bashing here, but it seems to me that they don’t try to finish the job when it comes to changes and enhancements. I’m not an inside man so it’s a bit questionable how much this article can be trusted, given Canonical’s small size etc.
Edited 2008-09-21 14:14 UTC
Ubuntu does help upstream _A LOT_ with regards to bug reporting:
“Ubuntu users report bugs, the ubuntu developers wait for upstream to solve them, and then backport the fixes to the ubuntu packages.
In this senario, bugs are reported and fixed, but few patches goes from ubuntu to upstream.”
(Quote Jonno. LWN comment)
> Ubuntu users report bugs
In launchpad rather then upstream.
> ubuntu developers wait for upstream to solve them
You nailed it.
Well, isn’t exactly how it should work? Ubuntu developers get the bugs, make sure they are valid (e.g. they are not caused by their own customizations) and send the relevant ones to the upstream.
Nobody is more familiar with a project than the original developers.
IMO, it’s better than making your own patches without distributing them, which can lead to even more bugs because the fixers are not familiar with the upstream project. Think about Redhat and PERL or ask upstream developers about bug reports from Gentoo users.
Maybe that’s just me, but I see developers for a distribution as integrators, not core developers. Perhaps Novell are doing more work on projects, but I have never been impressed by the quality of their distribution. Of course, your mileage may vary.
No. Distro’s developers should fix bugs and add features upstream and that’s exactly what Gentoo, Debian, Novell, Red Hat, etc. do.
As I know KDE best, I give you examples from there: Stephan Kulow is in the top 3 of KDE bug fixers. See http://commit-digest.org/issues/2008-09-14/ (under Bug Killers). He’s a Novell employee: http://en.opensuse.org/User:Coolo and http://news.opensuse.org/2007/08/09/people-of-opensuse-stephan-kulo…
Many SUSE-specific features that ship with the openSUSE 11.0 version of KDE 4.0.4 were integrated into KDE 4.1. They couldn’t be integrated into KDE 4.0 because openSUSE’s and KDE’s release cycles are different. Contributing distributors like Novell/SUSE and Red Hat are not stupid. They know: Not committing their patches upstream just increases their work load. They know it because in the 1990s both companies tried to get in front of the competition by doing so.
Employing Free Software developers is a key element of the business of Novell, Red Hat, IBM, Apple, Sun, and so on. It decreases their cost: 5 companies employing each 2 developers and collaborating on projects is cheaper for one company than employing 10 and doing software on their own. Google doesn’t contribute kernel patches because it’s a nice company. Google uses Linux on its servers and needs to tweak it for performance or if a bug occurs.
Then there’s Cannonical… a company that in theory wants to make money by selling support contracts for its Linux distro. Yet, Cannonical commits less Kernel patches than Google: https://www.linuxfoundation.org/publications/images/table4-companies…
You may not approve the keynote but the keynote is a symptom from a larger issue. Ubuntu is quite successful among the mainstream, but Cannonical’s contributions to Free Software don’t match their success (yet). Ubuntu’s/Cannonical’s not-so-active support for Free Software development starts to backfire, because the hard core aka the Free Software developers noticed it.
Many KDE developers switched to openSUSE, no matter if they are employed by Novell or not. Novell-sponsored services like the openSUSE Build Service make their developer life easier. Nowadays you often see the SUSE mascot Geeko in the lower left of KDE screenshots. Organizations and individuals who are not affiliated with Novell like ArsTechnica and Aaron Seigo advise openSUSE as platform with the best KDE experience.
Now openSUSE is on #2 in the DistroWatch charts.
You see: Actively supporting Free Software development not only helps the software the distributor wants to distribute but also gives it publicity.
Why? Ubuntu has gotten more people to try Desktop Linux than any other distro before. It got me back into it, after years of ignoring because the distros of 1998 (the last time I tried Linux) weren’t up to the task of being a desktop. In the interim, I tried and really liked FreeBSD (if you can imagine that), as well as used Windows.
Ubuntu got me to try Linux again, and from there I moved to Debian proper. I think this is the case for a lot of people, they try Ubuntu and see what the fuss is about with this Linux thing, and then eventually, they move to another, more suitable desktop for their needs.
Yer, yer, yer, we get it. Everything negative that is said about Ubuntu is zealotry, because of course, Ubuntu and Canonical are the shining lights who will save the open source world – an open source world they don’t contribute much to apart from soundbites.
Bottom line is, if I’m going to believe the hype from Canonical about Ubuntu and if I’m going to shell out for a support contract I would like to know that they have people on their payroll who know about some of the more core software they are putting in there, are maintaining said software with others and can answer my questions from the horse’s mouth if and when I ask. At the moment, I am more likely to find documentation and explanations from Gentoo’s Bugzilla and forums than I am from Canonical.
It’s typical of what surrounds Ubuntu. Rather than acknowledge the problem, let’s write rebuttals and debunkings on our blogs. It’s what we’re good at. The biggest rebuttal in one of those blogs is “He works for Novell!” Not a good sign.
Additionally, he tries to shift things around by just implying that Canonical is maintaining parts of the desktop infrastructure. Well, I didn’t see Canonical coming forwards to help maintain KDE 3.5.x for a LTS release when the KDE developers had said they would support it for as long as people wanted it. I also don’t see all that many Canonical addresses on Gnome’s mailing lists, nor do I see them having a hand in driving forwards GTK and Gnome 3 and having some vision for it apart from “Let’s make it like Mac OS”. So, that line of reasoning doesn’t work either.
Canonical has it’s faults, sure. Looking at patch numbers isn’t how to determine that. Running the CD in the drive, is.
How does booting a live CD give an overview about how much Cannonical contributes to open source?
Holy unsubstantiated and sensational claim, batman!
For the Novell stuff, if you really can’t see why it’s disingenuous to not disclose how much you have to gain by smearing another company well maybe you need to revise your sense of business ethics, or even ethics in general.
Canonical has no interest in spending what funds they have on maintaining a dead-end project like KDE 3.5.x, or 3.x for that matter. The KDE project is trying to move forward, Canonical and other distributions are respecting that decision and agree with it. Your “investigations” really don’t amount to anything when apparently you do your “investigating” by perusing the GNOME mailing list. You take an opportunity to cheap-shot Ubuntu at wanting to be Mac-like (Which it isn’t, that’s as false as saying KDE wants to be Windows and GNOME wants to be OS X) after deriding the contributions Ubuntu has made to the GTK/GNOME development process.
If you would take the time to analyze the information presented in the articles responding to the presentation (Information that really should not have been omitted from said presentation) you would notice that it’s very unrealistic to complain of Canonical not doing more development on core pieces of the system including the kernel. The company is small, the main investor and contributor of funds is Shuttleworth, and you expect it to pump out the same numbers the old-name boys like Red Hat, Novell and IBM can boast? Canonical has never refused to give credit where credit is due, and is trying to do the right thing by getting Linux to more people and making it accessible. This criticism is not constructive at all, it’s obvious and in poor taste to point it out as if it’s some sort of failing, as if Canonical is being a bad guy. You want them to spend more money on full-time developers doing nothing but working on core parts of the system? I’m sure they would love to do just that if you were to tell them where they could come by this mystical yet-untapped treasure trove of funds.
P.S. — I really am interested how you are so quick to talk trash about the support services Canonical offers. Did you get one for your own personal desktop? Do you have some huge personal datacenter or server farm that you operate, and you’ve had experiences with all the major vendors? Somehow I doubt these things. In other words: Where’s the beef?
Edited 2008-09-21 23:44 UTC
In what way? Currently Gentoo contributes more to stuff like the Linux kernel with probably less resources than Canonical does, and if Shuttleworth wants people to fork out and get Ubuntu supported in a data centre or on a bunch of desktops (yep, hardware support is needed there as well) then Canonical aren’t doing anything to convince me.
The so-called rebuttals didn’t do anything to rebut that either. They don’t even contribute much upstream to desktop projects either, despite fairly feeble claims that that is the market that they’re in. Calling it unsubstantiated and sensational is plain wrong because it is substantiated. If Canonical want to change that, they know what they can do.
Hmmmm, no. I’m not exactly Novell’s biggest fan around here, but I’m just interested in whether there is something there or not. Unfortunately, Canonical’s contributions and ability to support people don’t match up to their marketing hype, and yer, it’s slightly below the belt, but that’s life.
If you think it unethical and shocking for a company in one market to point out the shortcomings of a competitor then I think you and Canonical need to take a trip into something called the real world. Canonical knows what it can do to shut Novell up here.
When you get yourself a clue give us a call, right sweetheart? KDE 3.5 is by no means a dead-end project, and will exist for as long as there is demand and contributors, as said by KDE’s developers themselves. It will probably be maintained for a few years yet, and yet, Canonical couldn’t be bothered to contribute something to that.
All you’re telling me here is that Canonical will merely support what other people are supporting and will contribute little themselves, even when it makes little logical sense to do so and hurts their userbase. Repeating Canonical’s part line as fact isn’t going to do anything to make it less wrong.
Apparently Canonical doesn’t respect the decision of KDE’s developers to continue to maintain KDE 3.5 as well, and help them to do so when it makes a lot of logical sense within a LTS release.
It was a silly, knee-jerk decision which wasn’t based on feedback or dialogue with KDE’s developers, and has led ultimately to the detriment of Canonical’s userbase.
Well, yes they do because code, dialogue with others and contributions that match up to the hype and the soundbites count in my book, and in the open source world.
At the risk of repeating myself, I still don’t see Canonical addresses on Gnome’s mailing lists and I don’t see Canonical taking an active role in the future of Gnome, and particularly Gnome and GTK 3. If they want to be on a par with Mac OS in two years then you would have thought this was important. But, whatever.
There is ample evidence for the Mac worship, and denying it is a bit sad:
http://itmanagement.earthweb.com/osrc/article.php/12068_3757246_2
….our goal, very simply, is to make sure the Free software ecosystem can deliver a Mac OS-like experience, or an experience that will compete with the Mac OS………..We see Apple as the gold standard of the user experience.
I don’t see what contributions they have made, apart from a rather dubious six-month fixed release cycle that delivers nothing but tangible features bumped to later versions.
I think you should take the time to read around here and elsewhere, because you clearly haven’t. Gentoo and companies like Mandriva have less resources than Canonical, and they contribute more to software components like the kernel than Canonical does.
No, but I expect it to at least be on a par with companies like Mandriva and non-profits like Gentoo.
When all else fails, tell everyone that Ubuntu is user friendly and accessible. Unfortunately, they’re not doing anything other distros aren’t already doing.
Errrrm, they could actually do what they’re promising rather than dishing out soundbites about it, and get people to want to pay for something they’re doing?
Because when I look for one I want to know that said company knows enough about the software they package.
Ha. I take it you’ll be repeating that as the Emperor runs naked through the town square?
The only thing the GPL makes you do is make your code available; that’s it. The only reason it seems more complicated than that is because people keep trying to skirt around it in various creative ways which leads to all the licensing lawyer crap, but that’s true of all software licensing no matter what the terms are.
How does freedom imply immunity from criticism? You can do whatever you wish, but that doesn’t mean that I or whomever else won’t try to persuade –and persuade is the operative word here– you to do something else.
It’s all the more silly to claim that your freedom is being encroached upon when in regards to a written article that you choose to read. Unless you actually have people in your yard with pitch forks –which would be a form of intimidation not mere persuasion–, I’m hard pressed to see how your hand is being forced.
If one’s definition of freedom includes being free of social pressure of any kind, society probably isn’t the place for you.
It’s pretty tough giving tens of millions to a Linux project and raising the platform’s public profile dramatcally, only to be told that you are some kind of freeloader. What does Shuttleworth have to do next, offer a pint of his own blood every time he writes a cheque? FWIW, Debian isn’t on that list either. People contribute what they can and in most cases that means contributing what they are best at, which isn’t necessarily working on the kernel.
I sometimes wish Linux folks generally would show a little more gratitude. An instructive graph not in this article would show the number of Linux users compared to the number of Linux contributors. I suspect that the number of Linux users who get an amazing operating system completely for free would be in the overwhelming majority.
Edited 2008-09-21 09:20 UTC
First, sorry but Debian IS in the list, and od contribute more code to the core kernel than Canonical. Secondly, you and Kroc are completely wrong. They are not blaming Canonical for “morally infringing” the GPL or similiar nonsenses… Greg is just pointing out that if you want to collaborate in the kernel ecosystem, if you want changes to happen (like Shuttleworth was asking) you have to be involved in mainstream and you have to send patches and be active in development. This is completely correct and I hope that Canonical will respond to this “show me the code” in the best way.
Just a few notes:
– Novell employees are eager to start open source bar fights ( Meeks, de Icaza come to mind )
– Novell has a lot of problems with “open source morality” of their own (I won’t point fingers
– GregKHs view of the “ecosystem” is twisted. “Plumbing” would be the right word
– Negativity helps no one
– Canonical is tiny, if you took relative numbers you could say the same for Google or Xandros, Archos .. the list goes on for ages. They just don’t scare Novell.
– Ubuntus strength is not their kernel work
– Being at the whim of others might not be such a bad thing after all and might have worked for Canonical
And finally it is no wonder that a company like Novell has a attitude like “don’t take our code and don’t contribute” .. they just don’t understand open source.
Red hat is on a very different level in that regard.
What are Ubuntu’s strengths?
Novell does not understand open source? Why is Novell the second largest Kernel contributor then? Why is openSUSE the distribution of choice for KDE developers? (As for example seen here: http://movingparts.net/2008/09/15/opensuse-103-kubuntu-hardy-fedora… )
Well, obviously their GUI and usability work, especially that related to GNOME. You cannot concentrate on everything and be good at it all, you have to concentrate. Although – according to many – Ubuntu is already spreading its focus in all too many directions.
But I suppose there are always people who have nothing better to do than whine, what ever you choose to do…
Also, like many have already pointed out, and Shuttleworth is always ready to admit, Ubuntu is only building on the well-tested Debian base. Lots of the basic work, like the work related to kernel etc. is already done for them by Debian. Debian base is very good and there’s no need to reinvent the wheel in that sense. But what Debian lacked, traditionally, was related to its desktop usability, and that is the field that Ubuntu has concentrated on. Why should they reinvent the wheel and start competing with Debian by concentrating on the exact same tasks that Debian already does for them?
If we are talking about open source and free software, why should we support camp mentality, this camp against that camp etc? Co-operation is and should be the key word in the open source world. Contributors to free software are supposed to cooperate with each other instead of competing against each other. Somebody is good at something and concentrates on that and it will benefit the whole community.
Unlike Novell, Canonical is not a big old corporation, and lacks resources compared to Novell. In fact, Canonical’s resources are still a bit too much the same thing as Mark Shuttleworth’s money that he has generously given to the Ubuntu development. The man is sure rich but he doesn’t have endless amounts of money and has his right to decide what kind of projects he sees more important to give money to.
Edited 2008-09-21 18:50 UTC
What Usability work?
The usability work where Mark Shuttleworth fought tooth and nails to keep the “About Gnome” menu item?
Apart from the press release(s) have they actually done anything?
Huh? Yes, of course they have. Otherwise I would still be using Debian on my desktop right now too, instead of Ubuntu. I love Debian but I love ease of use even more and therefore decided to swicth my main desktop to Ubuntu.
The huge amount of other distributions based on Ubuntu is another testimony to that.
However, many things sponsored by Canonical/Ubuntu are now part of GNOME and Debian too so they are not Ubuntu-only things. But that’s is how open source projects are supposed to work. You are not trying to reinvent the wheel but cooperate with others.
These are a direct quotes from Wikipedia:
Also many new things in GNOME desktop environment were first introduced by Ubuntu. And so on.
Such as?
I know of Upstart which seems to be pretty successful and also used in other distributions.
There is also bulletproof-X which is less successful and reportedly can cause problems.
Apart from that I am not really aware of much that Ubuntu has contributed. It may just be that I am ill informed, but my impression is that there is not much else. (and from my erading of some mailing lists, a further problem is that when they DO fix something, they do not always communicate well upstream to let them know a fix is available…)
To me Ubuntu seems to wait for others to do the work, then the marketing spins up and it claims “Great Feature X is now available to Linux users Via Ubuntu!”
(I assume all this can be debunked by a link to a contributions page on the ubuntu website?)
You have to compare Ubuntu to its mother distribution, Debian. Basically, Ubuntu is only aimed to offer a more user friendly “Debian experience”.
Just put an average person in front of a PC and ask them to install, configure and use Debian and Ubuntu and then compare the user experience. Do you think they will find Debian easier to use? Hardly.
Although Debian has seen many great improvements in its usability lately, Ubuntu is simply much easier to install, configure and use to an average person, and also implements some new advanced technologies faster than Debian (stable) does. A stable Ubuntu release is usually also more stable than Debian Testing, not to mention Debian Unstable release, is.
You simply get a relatively stable and cutting edge Debian-like distribution, readily configured and streamlined for typical desktop use by installing and using Ubuntu instead of Debian.
Well, many things, maybe small or usually rather invisible, like improving automatic hardware configuration, improving GNOME menus and their structure, having sane software defaults instead of offering dozens of applications choices for the same tasks, replacing some default GNOME applications with newer and better alternatives – like when replacing the GNOME browser with the (then) better choice Firefox, etc.
So what if some other people may have usually developed the software used in Ubuntu? Ubuntu does not need to reinvent the wheel every time, just use Debian as base and then pick up and offer customers a good and streamlined selection of open source software in an easy to use form. If others have failed to do the same as successfully, it is not Ubuntu’s fault.
Edited 2008-09-21 22:17 UTC
No, but the notion that Ubuntu is the only distribution providing a half-decent installer is complete baloney. When anyone is pushed as to what Ubuntu actually does the only thing you usually get back is “Oh, it’s user friendly” painting over the fact that it’s really no better than what OpenSuse, Fedora or one of the other smaller distros like PCLinux are putting together.
Yes, you may be mostly right about that. However, one secret to Ubuntu’s success is, of course, Ubuntu’s Debian-like package management and huge software repositories. Other distros (not based on Debian) have been improving in that front lately too, but it has taken surprisingly long time and they may still be behind Debian and Ubuntu in that respect.
Other Debian-based distributions like PCLInux may have had other problems, like there maybe being doubts about their future etc.
As to package management, I remember years ago trying to convince a local Redhat-based distribution of the benefits of apt-get (apt-rpm) and how it could improve the usability and their popularity a lot. (Their distro had no repository system at all and all packages had to be downloaded and installed one by one, including all the dependencies. The so-called RPM hell.). They were not convinced, and have also been out of business for years now. But, like we know, also Redhat and many other RPM-based distros have eventually seen the light and use apt-like tools for their software management now too. But many still see that Debian-based distributions like Ubuntu are not only pioneers in this field but still do software management better than others in the Linux world.
Edited 2008-09-21 23:33 UTC
I’m sorry, but I fail to find the innovations. Don’t compare Ubuntu just to Debian. Debian was never been meant to be easily installable be Joe User. Debian is made for servers and system administrators. Other distros made moves similar to Ubuntu before.
Cannonical wrote a new installer. Wow, basically every distro does that. I remember Caldera Open Desktop (or whatever it was called) from around 1999. Its installer was great. Yeah, the hardware had to be mostly manually configured by that time but it was as easy as it could get; and later it offered the user to play Tetris. That was awesome.
It took other distros years to catch up with its usability.
The problem with Ubuntu’s installer is that it’s too simple. It doesn’t offer a real “advanced” installation routine if the user wants to. The user has to get the Alternate CD that comes with a text mode installation.
Compare that to YaST: Inexperienced users can just click “Next”.
I’ve first seen a live CD installer in BeOS. The same applies for a Windows based installer that installed the OS into a virtual partition (BeOS 5 Personal Edition did that).
OK, that’s not Linux and that stuff was not open source, but IIRC Knoppix came before Ubuntu and did that whole live CD installer thing.
Ubuntu replaced a few default applications, eg. ship Firefox by default and not a gazillion apps that all do the same. Ubuntu was hardly the first distro to do that. My example is again Caldera Open Desktop. I got that one from a magazine CD. Unlike “old SuSE” Caldera didn’t ship on 5 CDs, but 1. I can’t remember which defaults Caldera used (hey, it’s been 10 years) but I was a total Linux noob then and I was not confused.
Most of GNOME’s polish comes from Sun Microsystems, not Ubuntu. And KDE… well… not really polished at all: http://www.flickr.com/photos/19616885@N00/2535182834/
Great new OS innovations or not – but that’s neither what this story is really about. It is more about basic level bug fixing and other such code contributions, in what Greg Kroah-Hartman calls the Linux ecosystem, and where he – in between the lines – loudly praises his own employer, Novell as the misunderstood hero of the story, when compared to Canonical.
Edited 2008-09-22 00:04 UTC
Cannonical doesn’t even do simple bugfixing — at least not upstream or, to be more precise, only to a very small degree. That’s OK. Free Software permits that.
Instead of getting angry, Cannonical employees and Ubuntu fans could just admit that.
Instead they deny the article’s numbers.
You’re going to need an awful lot more than that I’m afraid, and it’s the sort of wishy-washy comment about what Ubuntu does I see an awful lot of.
I don’t see them I’m afraid. I don’t see Canonical addresses on Gnome’s mailing lists, see very little in the way of downstream patches pushed to upstream, and for a company that is trying to outdo the Mac I see no discussion at all on GTK and Gnome 3 when you would have thought that this would have been vitally important to them.
In short, it’s all hot air.
Right. Then of course one might wonder why they chose Debian Unstable, which by definition is not well-tested.
They chose Debian Unstable as a compromise between stability and bleeding edge software.
If they had used Debian Stable, or even Testing, they would be two releases behind on all fast developing software like Firefox, Gnome, music players and what have you.
For me, the main reason I find Ubuntu palatable as a distro is that it offers vast debian-style repositories with more or less bleeding edge software.
Who told you that Debian Testing is two releases behind?
Debian Testing is bleeding edge, quite often with packages newer than Ubuntu.
And how can you say Debian Unstable is a compromise?
What would be in your opinion really bleeding edge? Debian Experimental?
At least get informed before posting.
“They chose Debian Unstable as a compromise between stability and bleeding edge software.
If they had used Debian Stable, or even Testing, they would be two releases behind on all fast developing software like Firefox, Gnome, music players and what have you.”
Packages generally migrate from Unstable to Testing after a 10 day period without serious bugs. This lets packages be far more up to date than a 5 month old Ubuntu release. It’s also the type of policy that would have prevented something like Ubuntu breaking Xorg in an update to their stable distro ( http://ubuntuforums.org/showthread.php?t=240957 , http://www.oreillynet.com/onlamp/blog/2006/08/ubuntu_xorgcore_updat… ).
Yeah, basing on Stable can leave one behind a bit. Testing? Not so much. Testing seems a much more sensible target to me.
But how often does Ubuntu synch with Debian Anything these days? I don’t actually know
That’s call deflection. I see a lot of contributions from Novell compared to Ubuntu.
Basically you are saying: “ignore our problems,,, look at the pariah Novell. They do more work than us but we want you to look down on them.”
This is why I never liked Ubuntu. It all seemed like a college students marketing class lab.
http://mdzlog.wordpress.com/2008/09/17/greg-kh-linux-ecosystem/
http://dustinkirkland.wordpress.com/2008/09/18/whats-behind-gregkhs…
Updated the item with links to the rebuttals for completeness’ sake. Normally I would turn it into a new item, but there’s no need for that now seeing how new the current one is.
Just so you all know.
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 and, the kernel, and instead of acknowledging the problem the best they can do is to try 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? Canonical contributes less than Gentoo, a free community distro without a single payed developer…
Edited 2008-09-21 11:10 UTC
Your enemy’s enemy is your friend.
Do you really think that Novell, Red Hat, MontaVista, Debian, Mandiva, Gentoo, and every other Linux distributor conspire against Cannonical?
Of course they all compete with one another!
People, these Linux companies need to make a living. Despite their fluffy bunny and green meadows image, these companies need to make money, and they are one another’s competitors.
This is the prime reason why Greg attacked Canonical the way that he did. And it’s the prime reason why the Canocical folks rebutted. If they weren’t competing, there wouldn’t be a need to act like a bunch of children in public, they would’ve worked it out together.
-1 for both Novell and Canonical.
I asked if he thinks that all Linux distributors form a conspiracy against Cannonical. That’s different from competing with each other.
Style of competition varies from company to company, and from distro to distro. Novell, unlike most Linux companies, is still very much an old-school company at heart. They’re the middle aged guy, with the bald spot on top, in his new red sports car trying to act hip. But they are not really hip. They’re just play-acting. Novell’s old-school tactics should be familiar to anyone who has ever watched Sun, IBM, and HP go at each other over and over in too many convention keynotes. It is not unexpected behavior for a “mixed-source” company like NOVL.
I’ve not seen these kinds of tactics employed nearly as much by the companies that started out as Linux pure-plays.
-10 to Greg K-H, though. He knows better but did it anyway.
My opinion of his character just dropped a few more notches.
Edited 2008-09-21 15:31 UTC
“It is not unexpected behavior for a “mixed-source” company like NOVL.”
Unnecessarily using a stock ticker instead of a company’s name in an attempt to look cool really only makes you look like a complete idiot, you know.
Off-topic, irrelevant, and incorrect.
Were you truly unable to come up with anything more appropriate, Adam?
Edited 2008-09-21 18:05 UTC
Off-topic? Sure. Irrelevant? Guilty as charged. Incorrect? That I take exception to.
I don’t know where this trend for talking about companies as if their stock ticker was their name started (probably on some trendy-yet-entirely-clueless ‘business’ site), but it’s stupid. The company is called Novell. Not NOVL. Knock it off already.
In this case, it was actually quite appropriate, as it was intended to call attention to the fact that they *are* publicly traded, and thus are driven by their responsibility to make a profit for their investors. Different companies handle that in their own ways. But Red Hat’s way of doing it is decidedly not the norm. Novell, being an old-school company from the 80s, handles it in a more typical way.
I could have spelled all that out explicitly. And as it turns out, I’ve apparently had to do so anyway. But using ‘NOVL’, in that case, was a succinct way to draw attention to that aspect. Take the context into consideration before you post knee-jerk criticisms, please.
That said, you won’t catch me doing things that really *do* qualify as silly and detrimental to credibility, like the writing Microsoft with a ‘$’ thing.
And yes, business oriented sites do indeed commonly refer to companies by their trading symbols, in both articles and in comments.
No, they conspire against Microsoft.
Don’t think so.. that’s more ECIS territory.
The second linked article is a survey by the Linux Foundation. It also includes a graphic about where the kernel contributors come from: https://www.linuxfoundation.org/publications/images/table4-companies…
There are even some companies on the list, I’ve never heard from before, like Linutronix. Cannonical/Ubuntu is not on the list.
I don’t know if the Linux Foundation is nowadays a competitor to Cannonical, but I don’t think they are. 😉
OT:
And I love this comment
Mark Shuttleworth, with his playboy image and willingness to piss people off, is on his way to being Linux’s Steve Jobs. This is a good thing, as Linux has no pretty faces in the press (apologies to RMS, LBT, ESR and BP).
“Mark Shuttleworth, with his playboy image and willingness to piss people off, is on his way to being Linux�¢�€™s Steve Jobs. This is a good thing, as Linux has no pretty faces in the press (apologies to RMS, LBT, ESR and BP).”
Wow, dunno where that came from, but am I to understand that whoever wrote it figures Steve Jobs is a pretty face?!
I’d rather wake up next to billg than that bear. *shudders*
I find both reactions quite funny.
First: I don’t think the keynote was held to bash Cannonical. Even if they do only small contributions, every contribution is welcome. I think the keynote targeted Ubuntu fanboys who see Ubuntu and Cannonical as the best things ever happened to open source. Cannonical’s main contribution to FOSS is probably marketing. The term “Ubuntu” reaches out to people who have never heard of Free Software before. It’s cool to reach the mainstream. However keep in mind that FOSS is nothing without its developers. Just don’t forget where the developers come from.
Secondly: Some of the counter arguments are a shot in the own foot. Quote from the first one:
Yeah, he shouldn’t make that statement. I’m a KDE user and thus my main focus of interest is obviously there and only to a lesser extend GNOME. I don’t count patches or anything but I’m closely following blog postings etc. So Cannonical has about 3 employees who work on KDE. That’s Harald Sitter aka apachelogger, Jonathan Riddell, and maybe Celeste Paul aka seele. There’s a page in Kubuntu’s wiki that’s not really conclusive: https://wiki.kubuntu.org/Kubuntu/Teams It lists one person, Eugene Tretyak, who’s directly involved with KDE.
Compare that to openSUSE/Novell: http://en.opensuse.org/KDE_Team
Not counting the members from the community, Novell employs 18 people who work on KDE. That’s more than a tenth of Cannonical’s entire work force.
The second reaction compares Cannonical’s company size with Ubuntu’s success. Again, IMHO that’s a shot on the own foot. It clearly states that Cannonical’s work force didn’t grow enough to keep up with its success. Cannonical does not have a work force of adequate size. One “LTS” release every two years that supports desktop components for three years. By comparison: Every openSUSE release (roughly every 6-8 months) is supported for two years. SUSE Enterprise for seven years. Red Hat’s enterprise offerings are in the same league of support length.
I don’t know how many people Red Hat employs to work on FOSS, but according to Wikipedia Novell employs over 500 free software developers. Cannonical’s overall work force is just 130.
So IMHO when Cannonical want to be respected, their contributions to FOSS should keep up with their success. It’s even for their own good. Why should a big company deploy Ubuntu on their workstations when they know that Cannonical can’t offer them the support they need? They turn to Red Hat or Novell because they know: They don’t need to upgrade their desktop PCs after three years, because they have seven years of support. How is Cannonical expected to be financially successful when the paying customers turn to somebody else?
It’s not like Cannonical is a little startup with little money. Yeah, Cannonical is a new company, but founded by a billionaire. Cannonical/Shuttleworth has the financial resources to increase Cannonical’s work force. It’s time to do so or Cannonical’s as well as Ubuntu’s reputation will shrink.
Edited 2008-09-21 12:07 UTC
> Cannonical has about 3 employees who work on KDE
afaik they have only one employee, Jonathan. Others contribute independent of Canonical and are not related to that company. So, I would assume Greg’s statistics are valid for KDE too.
Boah, what a lot of Canonical bashing!
Having done some work for some of the bigger FOSS projects, I can assure contributing isn’t as easy as fixing something, running diff and uploading the result to some Bugzilla or -patches mailing list. Participants in these projects are highly targeted experts, have more knowledge about their pet than you can learn in months and it’s next to impossible for an “outsider” to get it right for them or their project.
Much better is to exactly pinpoint failures or possible points of improvement and let these experts do the remaining. Once a problem is well documented, it’s most times a snap for an project expert to actually fix it.
If distributions and project specialists work together this way, it’s most effective for both. Let’s do everybody what they’re good at.
The best way to make a regression is to edit code that doesn’t belong to you (there’s often the risk of edge cases that are not obvious just from the location of the fix).
Don’t forget this one: http://linuxhaters.blogspot.com/2008/09/free-as-long-as-you-give-ba…
Personally I think LinuxHater have about as much credibility as all the equally lame and retarded Windows hating sites out there. If you hate an OS so much that you have to create an entire blog about how much it sucks and how much you hate it you need to either get out and get a life or seek professional help.
Not really. There are many shortcomings that LH is entitled to point out, and when you actually read some of the things that LH has gone through it has to have been made by someone who wants Linux to get better, but is a bit frustrated with the status quo today.
I can read about those exact shortcomings more coherently elsewhere and without the retarded, foulmouthed language.
Greg KH shows some statistics of Canonical’s actual contributions. The two rebuttals from Canonical employees suggest that Ubuntu developers don’t like what those statistics tell. If those numbers are incorrect, then Canonical clearly needs to provide more accurate statistics. But do the rebuttals really refute the numbers presented in the keynote?
Matt Zimmerman writes that Greg KH’s figures are wrong. But what Zimmerman actually says is that the figures in Greg KH’s earlier “Google tech talk in June 2008” were wrong. However, Zimmerman agrees that the figures in the current keynote are actually more or less correct. So, the only factual inaccuracy that Zimmerman is able to point out in the keynote concerns Greg KH’s claim that Canonical hasn’t been contributing to binutils at all. Zimmerman refutes this by telling that a Canonical employee has actually contributed one patch to binutils.
Dustin Kirkland’s rebuttal doesn’t try to challenge Greg KH’s statistics. Instead, he points out that Canonical is a small company with little resources. Then he suggests that the motivation behind Greg KH’s keynote was jealousy of Ubuntu’s popularity. Kirkland also suggests that many first time Linux users would have never heard of Linux if Ubuntu didn’t exist. This kind of claims are, of course, impossible to prove.
Kirkland also writes that “Canonical and Ubuntu actively contribute to GNOME and KDE, as well as dozens of other open source projects”. However, he doesn’t back up this claim with numbers, like Greg KH does in his keynote.
Less than two years ago a core Ubuntu developer Scott James Remnant wrote in his blog: “we have a policy of not doing our own software development, but only packaging what others have developed”.[1] I’d expect that Ubuntu’s development policy might have changed since then, but this policy (that appears to have still existed two years ago) might help to explain why Greg KH’s statistics show such a small number of contributions from Canonical.
If Greg KH’s statistics are correct and if Canonical thinks these statistics make Ubuntu look bad, then perhaps Canonical should figure out ways to contribute more, instead of writing rebuttals and living in denial?
[1]
http://www.netsplit.com/2006/11/27/slippery-slopes/
Edited 2008-09-21 12:32 UTC
Of course he doesn’t, because he knows that Ubuntu/Cannonical is obviously not the only distributor that contributes to KDE and GNOME. I compared in another posting the number of contributors to KDE from Cannonical and Novell. Cannonical has 3, maybe 4, and Novell has 18. Nokia/Trolltech also sponsors several KDE developers, not to mention all those people who work on Qt.
I don’t know how many KDE developers are employed by Mandriva.
I’m not following the GNOME development closely, but it seems to me that the majority of contributions come from Red Hat. GTK is currently mostly in the hands of Imendio. See http://www.gtk.org/development.html
As I said before: I’m not bashing Cannonical. I think every contribution is great no matter how small or big it is.
Having Cannonical among the FOSS sponsors is great. Cannonical employees just shouldn’t pissed when somebody compares their team to another team.
“I don’t know how many KDE developers are employed by Mandriva.”
For the record, I believe it’s three. But then we have a far smaller budget than Canonical.
Small or not, I’d say it’s extremely well used. With one of those developers being Laurent Montel. There are not many weeks the last two or three years he is not among the top five developers on the KDE commit statistics.
Not that one can expect that most are able to contribute to upstream like that, but it clearly show that budget size are not all that important when it comes to working with the comunity.
More than Canonical then, and you have the stalwart Laurent Montel who really is a KDE person through-and-through. Not that that reflects on Jonathan Riddell, but the guy clearly could use a lot more help.
Also for the record; its one from Canonical.
No, but I don’t think this has ever been the goal.
All of this is a classic flamewar pattern. We used to get them via NNTP but since blogs are the new usenet we get them via HTTP instead.
Greg could have just corrected the numbers of his Google Tech Talk and could have still pointed out where this puts Canonical relative to others.
However stomping on Canonical at every single data point is IMHO clearly just meant as flamebait.
On the other side the “rebutals” could have just pointed out which areas Canonical and Ubuntu can probably easily found among the top ten contributors instead of personally attacking Greg (pretty certainly everybody reading the slides knows that he is working for Novell and definitely all who attended the conference) or hiding behind excuses (if amount of contribution would be directly correlated to number of employees or profit. Debian and Gentoo wouldn’t be at better positions).
We can of course still hope for a real rebutal, e.g. someone ignoring the flamebait and just listing which project’s Canonical in particular or Ubuntu in general are highly involved in.
Probably not at the same level of the software stack that Greg used but maybe higher up, since any level of the stack is important for the whole anyway.
You know, I could write something up describing why I sigh and shake my head every time this topic comes up. Why it is that despite all the carefully prepared charts and tables of statistics… all the cherry-picked anecdotal evidence… and all the solemn affirmations of objectivity that inevitably show up… the real motivation is always so painfully obvious. I could write something up about all that. But I really could not say it any better than Morrissey:
http://tinyurl.com/4p5bkg
Edited 2008-09-21 14:00 UTC
What? Obvious? What’s so obvious about an employee of a financially battered company with a tattered reputation trying to bad mouth one of it’s biggest competitors? I mean, really, there cant possibly be any ulterior motive at play here. Right?
While KH’s talk may not have been the most positive thing, Canonical’s kinda blown the response. Crying loudly that “Novell put him up to it, they’re jealous, it’s a corporate attack!” is just as bad, and no less idiotic. The second rebuttal lists some numbers, but no sources for those, I’d like to see them. KH’s data may be off, but I seriously doubt it’s a malicious attack by Novell, like Canonical wants to frame it.
Ubuntu is more of a “glue-and-marketing” distro than a “tools” distro, so they probably don’t have as many contributions to the toolchain as others. Big deal. Not fixing something because you don’t need to isn’t a bad thing. What IS bad is creating patches that nobody tries to send upstream. As long as Ubuntu isn’t doing that, I say, “who’s counting?” And I use Gentoo, so I’m not biased in either party’s favor.
Edited 2008-09-21 14:33 UTC
Ubuntu shines more than other distros because they take all the disorganized crap that is the Free Software, proccess it and offer it to the user in a digested and organized way. And that to me has more value than the most brilliant kernel hack.
Edited 2008-09-21 14:35 UTC
Free Software is crap? Wow, never thought of that. Thank you for your insightful comment.
I’d also never thought that every distribution other than Ubuntu sucks. Tell that to the osnews.com editors to make them never post “lies” like those again: http://www.osnews.com/story/19895
Other “lies” are posted there: http://www.linux.com/feature/119783
I can’t close my posting quoting the biggest “lie” spreader out there — ArsTechnica:
http://arstechnica.com/reviews/os/hardy-heron-review.ars/5
Imagine this: Every single Linux oriented company just spent all their time and money hacking the kernel. Yes the kernel may be a lot more awesome than it is now, but hardly anyone would be able to use it, because installing it would be more complex than a Gentoo Stage 1 install. That would suck.
It would also suck if all the companies just focussed on fronted/desktoppy stuff, because it might be easy to install, but none of your hardware would work.
We need companies like Canonical that focus on making an easy to use distribution, but we also need companies that sponsor kernel hackers so our webcams and printers work.
I havent listened to the talk or read any transcripts, but it seems to me like GregKH just went out to beat down on Canonical and not do anything particularly productive.
Canonical creates a distribution from an already existing infrastructure, namely Debian GNU/Linux. Most of the core work is done for them. They do more in userland than to the core. It is nice to see my distribution, Gentoo, having more contributions to the kernel though. It just shows that community distributions can contribute more in some areas than larger ones like Ubuntu.
The problem is NOT Ubuntu not contributing to upstream. It is following this up by “claiming leadership” over the projects, asking other distributions to work off their kernel “as it is more stable than what others use” etc etc.
What is worse than someone not pulling their weight is if they don’t pull it and then try to boss those that do round.
Shouldn’t open source and free software be about cooperation more than about small camps (or big corporations) fighting each other?
Canonical concentrates especially on the desktop level development instead of kernel level work. Big deal? Isn’t concentration good? Somebody is good at something and concentrates on that and it will benefit the whole open source community – because the source code is open and free to to use.
People, like the Ubuntu developers (most are or were Debian developers too), were happy with Debian but there was need to improve its usability, especially on the desktop level. That – desktop usability – was and still is the main focus of Ubuntu and Canonical, not kernel development.
So maybe Canonical could try to contribute a bit more to kernel development too, by sending patches etc. Let’s hope they will find more resources even to that too. Although many people already accuse Ubuntu and Canonical of focusing on all too many things simultaneously…
But Canonical is not a big rich corporation. Mark Shuttleworth – as rich as he may be compared to most people – doesn’t have endless amounts of money to throw around. Most developers hired by Canonical have been desktop developers.
Simply – their main focus is and probably will stay on the desktops.
I fail to see the point in GKH’s talk. I do not think it is in dispute that Canonical contributes less than many companies to the core (Linux Kernel, GCC, X). For that matter they do not contribute as much as many companies to the user visible components.
It is funny in a way. I do not hear or see Canonical arguing that their contributions are in volume or nature comparable to Novell’s or Red Hat’s. You get a random voice from the community upon occasion but nothing compared to the numerous others decrying Ubuntu’s contributions and existence for that matter.
Canonical is not trying to control the wave, merely surf upon it. The fact of the matter is that Canonical is a tiny company. They do not have the resources to effect great change, nor has it been my impression that they are trying to. They package and present Linux in a way that is appealing to many users. Whether clever marketing or actual accomplishment does not matter, many people have bought into the vision.
Canonical have, or listen to, ideas and attempt to implement them to make a better desktop. Their focus is there, and even then not on core design but upon assembly and fine tuning.
Do not confuse popularity with size, or merit for that matter. Canonical is popular and hope to be profitable one day, but right now they are small and trying to effect change to the users benefit in a way their size can accomplish. Hiring a team of kernel hackers will not help accomplish Ubuntu’s goals on the desktop as going a different route.
Mark Shuttleworth in particular is known for stepping on the toes of establishment and the merits of his ideas are much debated. Indeed in many ways he steps beyond his sphere of influence, but at least he is not afraid of speaking his mind. Nor is GKH. The same talking points could have been discussed without targeting Canonical, he has an axe to grind and given the lack of Novell logo I would say it is personal.
GKH’s slides not having the Novell logo on them most likely means Novell did not bless the message. There is nothing sinister in the act, it simply means GKH was talking for himself, not Novell.
But you do get Mr Shuttleworth saying that other other distributions should base their kernel of Ubuntu’s as it is more stable and better tested… and you get him projecting a leadership image where he is leading the Gnome community down a path and asking others to move to his tools instead of Canonical to upstream’s.
What this presentation shows that while Canonical talks the talk, it is not yet walking the walk. It is valid and debunks the leadership myth that Mark Shuttleworth is trying to project.
It sounds like Shuttleworth but I have not seen that particular accusation before. Can you provide a link.
This I can see as an extreme interpretation of Shuttleworth’s talks but I do not really agree with the interpretation. Can you give examples of this?
And you are arguing against Mark Shuttleworth’s projection of leadership, not Canonical’s contributions. If GKH has problems with Mark S. then he should have based the talk on such and backed this up with accurate facts.
http://lwn.net/Articles/290156/
As for the second, yes it is my cynical reading of the proposal from http://www.markshuttleworth.com/archives/162 (specifically the reply to the first comment.) and other blogs on planet gnome where developers were asking for the code.
To the cynic in me, even the talk of synchronisation talked of flailing – Ubuntu suddenly was using a kernel not being used by another distro, an xserver with the same issue and a couple of other important bits too. (since that time, Debian 5 and OpenSuse 11 have also made the same choices, so Ubuntu is no longer alone).
And for Ubuntu 8.10, there is the “late” switch to kernel 2.6.27 – something Fedora had already been planning to go with. TO me this smacked of synchronisation to cut the workload. Not a bad thing in itself, but how large is the iceberg?
I have also noticed that in some places, there is massive resistance to change in Ubuntu, as was seen in the discussion over whether to go with Empathy by default over pidgin for Ubuntu 8.10. This caution is good in some places, but bad in others. Me personally, I like a little bit of Chaos.
uh…. read the original interview, not a response from a Red Hat developer?
He is responding to the point that with release synchronization Ubuntu gains the stability of the Red Hat/Novell kernels without contributing anything back.
Shuttleworth’s argument is that Ubuntu provides better security updates (which I disagree with but that is another matter) and hardware comapatibility. Thus all parties gain from the common release, Ubuntu from Red Hat’s stability patches and Red Hat from Ubuntu’s hardware compatibility. Nowhere does he say others should use Ubuntu’s Kernel, but instead that all parties would benefit from a unified release schedule and patch set.
Read into it as you like but he does not say “Everyone should use Ubuntu’s Kernel.”
OK, obviously after reading the blog entry and responses I came away with a different impression but so be it. Ubuntu is still working on getting into the game and they seem to want to play by different rules. Time will tell if this form of contribution can be effective.
Launchpad can be used as an effective tracking tool with the ability to push bugs upstream to bugzilla among other tracking systems. This does give some benefit of a single point of entry for entering bugs. Personally till they open source it completely the idea of launchpad is “meh, so what” to me. Have no real need for it personally, though if it works with upstream as advertised it does make life simpler by providing a single point of entry for problems in Ubuntu.
On the other hand I am quite fond of bzr and use it for my projects. I use the svn connector to push change sets up to the company repository and can definitely see the benefit in using zr for tracking internal changes prior to pushing upstream.
You see it as a late switch. I see it as a reasoned decision based on stability and hardware support. It would have taken more work to backport and test than to update to .27 and test. Shrug, different views I guess. Personally I would have been critical of Canonical had they not done the update to .27.
I do not think there is much debate that it will happen, just if the 8.10 release was the right time for it. The counterpoint to the resistance to change are some of the unwise changes that Ubuntu did too early, like PulseAudio.
Blehh…, what exactly is your point, VistaUser? Many other leaders in the opensource community tend to be much more aggressive and dominant than Mark Shuttleworth has been: Linus, RMS, ESR, the participants in the Mono/.net debate etc.
Shuttleworth is rather considered a relatively diplomatic person who tries to be in good terms with others (which, by the way, is probably one reason why Ubuntu and Mr. Shuttleworth are so popular). But he is interviewed and asked questions about this and that all the time, and of course he supports his own projects and their goals. But I don’t see him trying to force his opinions on anybody, not to mention trying to control the direction of other projects like GNOME or other distributions – anymore than anybody else (like Novell, a big rich company) is.
@irbis – they also provide the code to go with their leadership or aspirations thereof.
This is turning out more confrontational than I intended, but I stick by my point.
You do? The whole point is outlined right at the beginning of that article. It’s an answer to questions asked by Ubuntu fans at a talk given by him at Google.
Who submitted the link for this story? More importantly, let’s look at what company they work for. Oh it’s Red Hat. I’m not even a Ubuntu fan, I’m more a Fedora man but Red Hat employees posting links to unsubstantiated tripe for osnews doesn’t seem quite right to me. Lost some respect for you there Rahul.
You or I may not like how most of the people involved made asses of themselves but it’s still newsworthy.
I think the news is worthy enough and my summary is balanced regardless of the pointless trolling in comments. I have been a osnews contributor for a long time and it is not related to where I work for.
To avoid precisely such accusations such as yours, I have usually limited myself to just some specific news but the osnews editor mailed me recently asking me to be “deputy” for Linux related news. I guess your post is a good example of why I should have refrained. Feel free to mail the editors and tell them. Your loss.
… when people start using words like ‘ecosystem’ about a technological product. Next thing you know we’ll see words like proletariat.
But of course Ubuntu’s popularity is what makes it so unpopular with the die-hard counter-culture free as in freedom zealots who need to ‘rage against the machine’.
After all, that’s what counter-culture is. Accessability to the masses and popularity is the last thing they want.
See the overemphasis on X and the Kernel, when frankly X on it’s own without a decent library to turn it’s vague calls into something useful (which is why we have several dozen libraries like QT, GTK, etc) or a decent WM (Gnome, KDE, XFCE, FluxBox, etc) is a JOKE, and most of what makes ‘GNU Linux’ actually useful isn’t linux itself. (The GNU folks want to make a stink about making the distinction, fine, let’s call it as it is)
Edited 2008-09-21 21:47 UTC
I still don’t get, why Canonical gets blamed for not contributing to Linux. They are distributors – not so big company that glues different pieces of software to their Ubuntu distro. When something does not fit or work the way they want, they fix it. So the software they are putting into Ubuntu works without major fixing. Kernel, gcc, xorg devs makes the stuff, that Canonical thinks is good to distribute. IMHO even if some software has major hole to be fixed, Canonical will better wait while original authors and developers do that.
According to Greg Kroah-Hartman any contribution to GNOME or KDE also benefits OpenSolaris so he cannot consider Linux desktop environments like GNOME to be part of his Linux ecosystem – unlike the Linux kernel or X.org. However, also X.org (or many other things in his “Linux ecosystem”) is used by Solaris and BSD too so I cannot see how it would be any more Linux-specific thing than X.org is…
He says that he had to draw the line somewhere – but could the choice of things included in his “Linux ecosystem” have something to do with the amount of those contributions by his company vs. those of Canonical/Ubuntu, and with the fact that Ubuntu is the most popular distribution and that his employer, Novell, has been recently attacked by free software supporters (because of Novel-MS deals etc.), and that Novell therefore needs more credibility as a genuine Linux / free software supporter?
Canonical is still a small company that is not even profitable yet whereas Novell is a big old IT corporation. Of course big companies like IBM, Novell and Redhat have more resources to put into Linux kernel development. Good for them. But Canonical is doing its small part in the Linux ecosystem too, especially in the desktop environment space (that Greg Kroah-Hartman decided to drop out of his “Linux ecosystem”). However, of course all the parties could try improving their commitment, contributions and cooperation.
…As to X.org – and if I remember correctly, Ubuntu was the first Debian-based distribution to use modular X.org. That work later benefited Debian too. So their work concentrated on on implementing X.org technology rather than developing it. Probably the same still.
As a small new company that is yet to make a single dime of real profit, Canonical probably just doesn’t have the resources to throw to all directions like to basic level X.org development. If you want to develop X.org, and even want to be paid for your work, Canonical is not the best option to look for as an employer. That is why Canonical doesn’t have many X.org developers hired. The same in Linux kernel development.
Ubuntu’s (and thus Canonical’s) main focus is only in distributing and building a user-friendly Debian- and Gnome-based desktop distribution, mostly build from tools developed by other people than themselves.
The obvious line would of course be at the Kernel since that’s the only thing that’s Linux specific. Why he didn’t we dont know but it does lend itself to unflattering speculation.
Ubuntu get lots of credit for usability. But they are getting credit for the work of gnome hackers generally in the employ of Novell or Redhat.
Trouble is most Ubuntu users think that Ubuntu wrote the stuff. Ubuntu made Launchpad, that’s nice, the live/install single CD is a good idea. But mostly they get credit for menu layout and Brown theme.
There are plenty of smaller distros that contribute very little. Trouble is Ubuntu count for a huge percentage of the user base of linux, but they do not contribute to the community as a whole.
Look at it this way, if Ubuntu went away tomorrow, Linux would initially lose a lot of users. But it’s development wouldn’t even skip a beat. That seems a bit skewed to me.
“There are plenty of smaller distros that contribute very little. Trouble is Ubuntu count for a huge percentage of the user base of linux, but they do not contribute to the community as a whole.”
This statement makes little sense to me. The community consists of both users and developers, so if they have a huge percentage of the user base, then they contribute hugely to the community. If those users were lost, the community would lose all those people. Contributing to the community does not mean just to send in patches and code. It means bringing in new users, and helping it to grow. Lots of OSS websites proclaim that contributing means using the software and reporting bugs, not necessarily contributing actual code. One such piece of OSS software is the Linux Kernel. With that, I would say Canonical and Ubuntu contribute a great deal to the community, maybe not to the people that seem to require everything to be open, but definitely to the rest of the community.
Agreed, to an extent. If those developers are developing for Ubuntu, those changes are not going upstream, so they will not be missed. If they are developing for other upstream projects they will simply switch distros, they may not be as happy in another distro, but they will have to switch. So the community still has those developers.
As for bugfixes, if the users are reporting upstream, then they are a huge help. Most will be reporting to Ubuntu, and here is the problem. If the Canonical team don’t fix the problem, at least they may submit the bug upstream. But what we see, is that if they do, historically the patch doesn’t make it upstream. So the bug still exists for the rest of us anyway.
Bug reports are broken for Linux if we have to report to the appropriate package. What if, as is often the case, a program only fails when it is used with another program, window manager, hardware driver, etc?
For a project, tracking a few major distros bug tracking system for problems related to their own software isn’t very hard. For users or the distros to specially report the bugs to individual projects, it is.
This rant is basically Enterprise Linux vs Desktop Linux. Red-Hat and friends contribute a lot of code, yes. What you all forget to mention is that it is code THEY and their corporate friends need. It adds 0 value to Linux as a Desktop OS. Ubuntu could run just as well on kFreeBSD(if the Debian port was still alive) or Minix. Shuttleworth chose Linux because hardware manufacturers are more likely to contribute drivers to it because of the license and the pressure from the Enterprise Linux world.
Edited 2008-09-22 04:33 UTC
No it’s not. It’s more than that. I want Ubuntu to be a better citizen because it represents the user. With no code contributions, not just to the system, but also to places like gnome, it is the enterprise calling the shots.
Other than the occasional rant where Shuttleworth tells the community where it is wrong, nothing is going upstream and Ubuntu have very little control on where packages are heading. If Ubuntu really want to make linux the user os I hope it will be, then invest in some programmers who work on the upstream packages they want to use.
“This rant is basically Enterprise Linux vs Desktop Linux. Red-Hat and friends contribute a lot of code, yes. What you all forget to mention is that it is code THEY and their corporate friends need. It adds 0 value to Linux as a Desktop OS”
Red Hat leads the development of GTK, HAL, DBus, NetworkManager, PulseAudio, PolicyKit and dozens of other desktop components without which desktop linux of today would not exist
https://fedoraproject.org/wiki/RedHatContributions
I would seriously recommend you look up which projects Red-Hat has contributed to, since you’re statement is very far from the truth.
It is not fair saying that Ubuntu does not contribute nothing. An operating system is not only made up of software, and Ubuntu contributes in many important ways. The number of users is very important, since they spread the operating system and create opportunities for the developers.
Maybe Ubuntu is not the best of the existing Linux distributions, but it now is the most popular one in a market controlled by Microsoft products. It would be better to join efforts to achieve a change to Linux operating system than involve in this internal fighting that does not benefit at all the Linux world.
Today I see just Ubuntu at the International Day of free software in Costa Rica.
http://www.nacion.com/ln_ee/2008/septiembre/21/aldea1709989.html
Google translated article here:
http://translate.google.com/translate?u=http%3A%2F%2Fww…
..
Greg is a decent guy and I personally think Ubuntu is not bad a distro. I don’t believe there’s anything sinister about Greg’s presentation.
My attitude is, don’t get too serious. Has Ubuntu contributed to the Linux community, yes. Is it a good thing for Ubuntu to contribute more to the upstream? Yes. So there you go.
Sorry if this has been said before, i think one of the best things with regard to ubuntu is that it has raised the profile of linux no end. Many new people coming into linux usually start with ubuntu and then sometimes move onto other distro’s when confident with linux.
I think it’s good to look inward, to question certain practices to ensure that linux doesn’t go off track.
However with the free ubuntu disk send out, ubuntu ecosystem (users, comminutity, manufacturers(dell) etc..) i think that ubuntu has put a lot in.
It puzzles me to see Greg Kroah-Hartman attack Canonical the way he does and makes me ask what his motivations may really be.
Canonical Ltd. is only a very small company that is not even profitable yet, unlike a big old corporation like Novell, Greg Kroah-Hartman’s own employer.
Canonical’s main focus is sponsoring Ubuntu and also offering some customer support for it. Other things done at Canonical are just extra. A small new company has to have a clear focus like that. I bet many other small companies sponsoring Linux distribution development (not to mention if they are not even profitable yet) do not yet send in that many kernel patches either.
On the other hand, Ubuntu as a community is rather big. Quite a big portion of Ubuntu developers do not work at Canonical. Ubuntu is only partly a commercial distribution. By keeping Ubuntu open Canonical is able to leverage the talents of outside developers willing to contribute rather than having to do all development within the company itself.
It is not right to say that Ubuntu would be the same as Canonical Ltd. This may be the cause for some of the misunderstandings regarding Canonical’s and Ubuntu’s contributions to the Linux ecosystem.
According to a study approximately 30% of desktop Linux installations run Ubuntu. It is safe to say that it is about the same on developer desktops too.
According to another study, the one referred by Greg Kroah-Hartman, the biggest group of kernel and gcc contributors are what he calls “amateurs”, that is, those who are not directly working at some commercial company. Now, how many of those people may be running Ubuntu Linux, and are thus, in a way, Ubuntu developers too? 30% maybe? Something that Greg Kroah-Hartman again forgets (like also the role of Debian project as the mother project of Ubuntu, or the role of Linux desktop environments in the Linux ecosystem, and where Ubuntu’s contributions are strongest) as it doesn’t seem to suit his scheme of downplaying the role of Ubuntu and Canonical in the Linux ecosystem and praising (at least between the lines) the role of Novell, his own employer in it?
Well, that figure may be too high(?), but that’s not the point. Whether the actual percentage is a bit lower or not, the figure certainly gives you an idea why many commercial Linux distributors may have gotten jealous and afraid of Ubuntu maybe eating their bread…
Also, because of its huge popularity, Ubuntu must run on quite many developer desktops too – although it may be impossible to give accurate numbers. One has to draw the conclusion that quite many kernel, X.org or GCC developers must be running Ubuntu on their developer desktops too. In a way you could say that they are Ubuntu developers too, as Ubuntu is also a community project. That community aspect is something that you may forget if you look at the Linux ecosystem from a corporate and business point of view only.
That is a very valid and good point. let me point out an example. When Compiz was initially released, who do you think was the first community to extend (despite DR’s objections) and package the project. Within a month the project was being packaged, hacked and tested on Ubuntu under Quinnstorm and friends. They were all Ubuntu users. Ofcourse later on they botched the whole thing when they forked into Beryl, but now that project is back in the fold and is now known as compiz-fusion. I can give you plenty of examples just like this one. So while Canonical as a company may not have a lot of contributions, the community as vast as it is has contributed to many projects.
Maybe the reason why Canonical isn’t contributing more to the kernel is because it’s already being kept in excellent shape by Novell, IBM, Red Hat, the Linux Foundation, and others.
You’d also find that most Ubuntu users are using proprietary drivers, and therefore their bug reports are virtually disregarded by kernel devs unless they disable the drivers and reproduce the bug. Most Ubuntu users I probably wouldn’t be able to reproduce the bug at will; I know I haven’t been able to in the past.
What Ubuntu with all its end-users *does* know is the desktop, and some of its own desktop-related projects have been adopted by other distributions.
You might find, for instance, that Gentoo doesn’t contribute much to the kernel. But Gentoo users contribute a massive amount to the ecosystem through their excellent HOWTOs for doing virtually anything you ever wanted to do with Linux.
Have to agree with Matt (the linked blogger) 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, any beginner should peruse its forums to gain knowledge! See
http://www.thegsblog.com/?p=251
Basically, see my post in response to mdz:
http://mdzlog.wordpress.com/2008/09/17/greg-kh-linux-ecosystem/#com…
To summarize:
and:
And directly per Canonical:
Edit: Fixed per formatting issues with copy/paste.
Edit: Fixed per HTML/UBB tags
Edited 2008-09-22 17:49 UTC
And when Ubuntu patches its kernel everybody goes: “It’s different from vanilla! Ubuntu sucks!”
Also, a vanilla kernel makes it easier to send bug reports upstream instead of having them rejected.
Hmmmm, no. You would like to have as much upstream as you can, but you will never be quite in step with upstream though. Everyone has to patch at some point. What you want, however, are people upstream who know what they’re doing on your behalf, which is why Red Hat’s developers end up contributing a lot to the kernel.
Not the point. It would be better for Canonical to have developers contributing stuff upstream that Ubuntu needs to make what they’re doing better, which will possibly be desktop related, and making sure they have the experience in-house to support what they, their users and their customers need and actually have some vision for what they’re doing.
I was not going to post anything further under this essentially valueless story topic. (Not referring to your thread, but to the story topic itself.) But then I ran into this little gem posted in the comments on your site. It is so very apt. And funny because it is likely so true. Regarding Ubuntu’s position in the community and the detractors seemingly coming out of the woodwork, lately, to disparage it, he says:
“””
First they ignore you,
then they laugh at you,
then they fight you, ( <- you are here )
then you win.
“””
Edited 2008-09-22 19:12 UTC
Funny that you’ve quoted that, and even funnier, and rather sad, that it doesn’t actually apply. Using that statement assumes that Canonical are actually doing something to ‘win’ in the first place. They’re not:
1. Canonical and Ubuntu can’t win and can’t surpass Mac OS with the contributions they’re making to open source software today.
2. Point 1 means that any apparent progress is all down to marketing, soundbites and hype, and even Mark Shuttleworth will run out of money to finance that at some point as so many companies, most of them VC funded, have done before them.
3. People aren’t fighting Ubuntu at all I’m afraid. All they’re doing is pointing out points 1 and 2.
I’ll re-phrase your quote as it actually is, and it really is rather sad:
– First you think you’re being ignored so you create lots of hype.
– When people point out said hype and look for something more concrete (i.e. lines of code) you think people are laughing at you.
– The questions continue and you believe that people are fighting against you because they won’t go away.
– You fail to contribute the lines of code necessary to actually create something that surpasses Mac OS (or even to identify what’s needed), you fail to generate money off the back of that because hype and soundbites aren’t enough, your benefactor can’t put money in forever, you run out of money and you lose.
How can you employ developers when you are not making money?
I think this is the core issue. I saw a figure of 130 people at Canonical but last I heard (from sebastian; he who packages gnome) there are only about 10-15 developers working at the distro.
How do you expect them to contribute the same as Red Hat or Novell?
Shuttleworth has failed at providing a business model for Ubuntu: for four years Ubuntu has not made a profit.
That’s why there’s an increase in the enterprise aspects of Ubuntu. Let’s see how that goes.
Debian, Gentoo, Mandriva and others contribute developers and patches to many projects, and they all manage.
If that’s really true then Canonical has no future whatsoever as any organisation, let alone a commercially viable one. What on Earth are the other 115 hangers-on doing in an organisation that Shuttleworth set up to try and move open source software, and the desktop more specifically, forwards?
No one does. People are just questioning why they contribute less than Gentoo, Mandriva and other organisations versus their marketing exposure, the funding they have and the number of people they employ.
With 115 to 120 people not directly contributing to code that will make the enterprise market sit up, take notice and think “Wow, we have to have that” then I can’t see them being successful ever. They’re going to sign off with a whimper in a couple of years as no one can continue to bankroll that. I blame Shuttleworth if that’s true. All they’re doing is packaging Gnome and all the same software as other enterprise vendors are doing, with nothing that differentiates them.
If Shuttleworth wants to make money he needs to stop jacking off over Mac OS and start creating a competitive alternative to Windows Server out of open source software with none of the administrative and licensing overhead. If Canonical does that then people would crawl across broken glass with their flies unzipped (too much Red Dwarf) to get a copy of Ubuntu and give them their bank details when they got there.
From there, the desktop is a whole lot clearer and there’s money to be made on the way there. Any Linux vendor who does that really will have Red Hat and Novell worried.
Edited 2008-09-22 21:58 UTC
Oh please, all you Ubuntu lovers go home. Ubuntu is a desktop distribution. By all accounts, they give bugger all back to the kernel team. The talk/presentation was about giving back to the kernel. Greg was simply referring to this, and nothing else. He has publically acknowledged on his blog this.
Get a life you Ubuntu lovers…
Personally, I think Ubuntu is vastly overrated. You’re all a bunch of lemmings (or should that be lemons?).
Dave
we know canonical contributions is very low…
somebody can post their contribution to other project like gnome?
maybe it’s worse?