Submitted by As we mentioned in a previous article, Red Hat advocate Greg DeKoenigsberg claimed that due to the much larger amount of code it’s contributed, Red Hat is a better open source citizen than Canonical, adding, “Canonical is a marketing organization masquerading as an engineering organization.” A Computerworld blog retorts that that’s no insult; and that marketing Linux could be just as important to the cause as contributing code. UpdatedThere’s a long-running feud between engineering-types and sales and marketing-types, with plenty of mutual disrespect to go around. Among coders, I find that there’s a sentiment that great code will speak for itself, or, at the very least, that the salespeople and the marketers are somehow missing the point, focusing on the wrong things, or just plain making things difficult for the engineers by making promised that the development team has to deliver on.
Now, this is a dynamic that plays out in individual organizations, and isn’t the case in the anti-Canonical world. Canonical’s “marketers” aren’t making promises that Linux developers must deliver on, so I think the anti-marketing bias is residual. Or maybe it’s resentment that the Linux distro that gets all the glory isn’t an engineering powerhouse, but just a re-packager that’s riding everyone else’s coat tails.
Since one contributor to one of the many independent open source projects that makes up the greater Linux ecosystem is actually contributing to the project as a whole, there isn’t that much reason to get all “tribal” about who’s doing more than whom, but I think that this particular feud actually has a lot to do about the professionalization of certain open source projects. A large proportion of the serious contributors to Linux are paid employees of companies who are involved in Linux professionally. Their prestige and success is tied up in how well they can portray themselves as leaders in the field. When a company that doesn’t lead the pack in code contributions is nevertheless a mindshare leader, it rankles.
This is only an issue if you’re looking at Linux advancement as a zero-sum game, though. If there’s only so much prestige to go around, then you have to fight over it like hungry dogs. But if Canonical is raising the profile of Linux, then that raises the prestige of everyone who’s involved, especially the biggest code-contributors. So, in other words, can’t we all just get along?
Update: As mentioned in our more recent post, Greg DeKoenigsberg apologized for his inflammatory remarks.
We also changed the first sentence of this posting to clarify that the advocate or “partisan” was Greg DeKoenigsberg, not David Neary. We didn’t mean to imply that the original GNOME census was a partisan endeavor.
Marketing is good but it depends on who is the target. If the target is Linux users that use other distros, then there is no gain. It’s just altering the Linux ecosystem (and I tend to believe that this altering is for worse). If the target is Windows users then Ubuntu contributes with all this marketing.
Since Linux market share on the desktop for the past 5 years didn’t grow much, if at all, but Ubuntu’s market share (within Linux market share) grew tremendously, then it is safe to conclude that Ubuntu didn’t help Linux gain more users, Ubuntu mostly moved the Linux users from another distro to itself.
It is definitely a good thing to have the best distro represent Linux, but is Ubuntu the best? I may be wrong but I don’t think so.
And imagine for a second that Shuttleworth decides to stop pumping money because Ubuntu is not profitable. The void left in Linux ecosystem would be huge.
The marketshare of Linux might not have grown that much (it did a bit though), but the market as a whole did grow as well.
So Linux might have been at a little below 1% of the desktop market, it’s now a little over 1% of the market.
Which means in absolute numbers Linux did grow quiet a bit. Not as much as Apple obviously.
Edited 2010-08-03 17:48 UTC
No it didn’t grow based on statcounter:
http://gs.statcounter.com/#os-ww-monthly-200907-201008
I think Ubuntu has actually been bad for Linux due to it having zero corporate appeal. A weird brown er.. aubergine OS that sounds like a curry dish is not going to go far with the typical CTO. Name and presentation mean a lot and Ubuntu doesn’t have either.
Linux Mint has a much better name and comes with a complimentary theme that results in a cohesive offering. People have heard of Linux, Mint is associated with value, and green has wider appeal than brown or pretty purple. Shuttleworth should fire his design team and hire the Linux Mint guy.
Ubuntu has tons of corporate appeal. The engineering, the name, the color of the icons doesn’t matter; only the ability to buy a support contract does. Linux Mint has zero marketability in the corporate world due to it being a community project.
Ubuntu has been good for corporate Linux. It’s really pushed the idea that Linux has a place outside of the server room. There’s not another company that does that. Red Hat sells desktops licenses, but they would really rather sell just server licenses.
On the other hand, Ubuntu hasn’t done enough. They really need to start looking at the entire ecosystem corporations need. They need to make deploying and maintaining a Linux only network easy, and they really need to create a first class development package.
I do agree with you about Mint being better then Ubuntu.
So if Novell made a distro called ClownTimeOS that used a circus theme that wouldn’t matter as long as they offered support contracts?
Linux Mint isn’t being marketed to the corporate world but the name and color scheme would be an easier sell.
I’m not seeing major corporate rollouts of Ubuntu. I think it has been a waste of time for Linux overall.
It’s funny as to how many people prefer Mint when Shuttleworth has a team of designers and Mint is mostly the work of 1 person.
Mint has as much negative connotation as Ubuntu. “Would you like a mint, dearie?”
Well you might have not been looking in the right places. And most corporate rollouts are not publicized. Though the problems with corporate Linux desktop adoption are more related to Microsoft’s solutions being more integrated and complete, and have little or even nothing to do with names or color schemes. That is why Apple’s good looking OS is a total failure in the corporate environments as well.
Remember the Luna theme on WinXP, or how about the Aero themes on Vista and Win7?
I’ve seen apps designed to create as much pain as possible made the standard for large corporations, so the UI is not the deciding factor when decisions are made.
I’ve seen one job posting in my area that called for experience with Ubuntu on the desktop, and looking around there are a couple more around. Not a lot, but some.
http://www.indeed.com/jobs?q=ubuntu+desktop&l=
As JAlexoid mentioned and I hinted at, MS has a lot of support technologies that add value to Windows desktops. Ubuntu really needs to create an answer for them (easy implementation by unskilled admins) to really step up as the corporate desktop Linux.
Money can’t buy taste.
Huh? So hotshot designers work for free these days?
Novell does make a circus OS…
Hah.
Meanwhile, in the real corporate world, what matters is management tools that make deploying, updating and maintaining dozens, hundreds and thousands of servers and workstations as easy (read: manpower-free) as possible. Windows and AD and group policies, however crappy they are, do a decent job of providing these things to the average J-Random-DumbAdmin.
… and a big part of appeal of MacOS X and Linux for random user is that those operating systems are out of reach of those management tools. No enforced virus checks.
Almost all the “corporate” stuff is in the intranet, as HTML applications anyway.
Again, hah.
Meanwhile, in the real world, IT people and management people don’t buy systems they can’t manage. Most users do not need or have any access to their own systems, not in a corporate environment. Everything is managed and locked down. Most internal apps are with web based or .net-based these days, it’s true, but control of user behavior (e.g. installing apps, access control) needs to be manageable.
Now that is an outright false statement.
No it’s an opinion.
The corporate world has completely ignored Ubuntu in case you hadn’t noticed.
Pretty much no other company has put a serious effort on pushing the Linux desktop, so while Ubuntu hasn’t succeeded on the corporate world it has helped Linux grow in mind share.
I don’t think that’s something that hurts Linux.
Which is fantastic, if you only care about buzz and winning contests in Google Trends. Hype will only take you so far though and it hasn’t exactly meant success in the consumer space either.
They’re still relatively newcomers but it will be hard to justify the Ubuntu hype a few years from now. It should be pretty obvious by now that something more than lots of buzz is needed, and in my opinion whoever thinks that things are gonna change significantly and users are gonna flood to Ubuntu (or any other Linux for that matter) is only kidding himself.
Thinks are going to change significantly. It might take 10 years, but that’s no problem because financial problems or failure to gain mindshare in short term are not real problems for Linux as a whole.
Linux is pretty much unstoppable. And it’s only going to win more when people shift more and more to different form factors (phones and tablets).
Don’t see how things are gonna change. We’ve all been hearing that for the last 15 years, and all is just a matter of time.
Another 20 Ubuntu releases are not gonna make the slightest difference. You can’t expect different results if you keep using the same process that has been always used.
Edited 2010-08-05 20:51 UTC
You’re right about mobiles, tablets, and other similar form factors. There is no denying Android is a success, and that is technically a Linux distro.
Linux on the desktop will always be for the nerds though. Which sort of makes it even funnier that Ubuntu is trying to market to normal users. On corporate workstations it has its place also, but for the normal home user? Laughable.
Ubuntu Unity is doomed to fail, it is competing with both Android and Intel/Nokia backed MeeGo. Corporations don’t take Ubuntu seriously, and despite what you might think, home users still have no idea what Ubuntu or Linux is. That is proof of Canonical’s true marketing prowess, it is only people that hang around online forums that have even heard of Ubuntu.
The thing is, partially because of Canonical’s marketing, people at least know about Linux, which puts it closer to any kind of adoption that the previous “Linux? WTF is that?”.
Linux marketshare won’t skyrocket (or even grow significantly, or at all, or whatever) just because of Ubuntu, but you wouldn’t ever get there anyway if people doesn’t even know Linux exists.
Rather than complaining about Canonical being code leeches, other companies should capitalize on Ubuntu’s marketing work same as Canonical takes advantage of the development efforts of RedHat and the likes.
RedHat isn’t even trying to push Linux on the desktop, so why are they angry about Ubuntu taking all the (so to speak) glory? If they want the mind share then start placing ads, sending CDs and making deals with OEMs.
Don’t really know how much of that can really be attributed to Canonical’s marketing, for the sake of argument, let’s say that is the case. Again, that is a good thing, but the problem is that you don’t only have to get people to know about it and try it, you also have to get them to stay.
Not that I can prove it, of course, but I have the feeling that way more users know about Ubuntu and Linux that those actually using it. There’s this claim among many in the Linux community that the system is far superior to Windows and all it needs to succeed is to become more known, and I disagree strongly with that. I think it has a long way to go before it can actually compete with Windows (or OS X), and worst of all, is not even going the right way.
Canonical might be doing a lot to hype the system, but they’re doing very little to make the system live up to the hype. They don’t even have the resources or the business plan to fix bug#1, which they themselves set as the ultimate goal. After all, there’s a big difference between making a grand statement and being able to fulfill it, and Canonical can only make the first part happen.
Edited 2010-08-06 09:00 UTC
That’s still more than others have been able to do.
Declaration of a grand goal can be a big motivating factor.
Maybe others were prudent or honest enough to avoid creating the hype without the goods to back it up.
Edited 2010-08-06 12:50 UTC
It is simply a pointless fight.
The desktop form factor will be obsolete long before Canonical ever gain any sort of ground there.
That is the mistake Canonical has made: They’re fighting a giant on their ground. That isn’t have you win any battle, the only way to beat Microsoft is to make them fight in a new arena.
Red Hat figured this out a long time ago, that is why they concentrate on corporate needs. Canonical is trying to create a cloud story, but again, its a losing battle because Red Hat and others have a better story. Ubuntu can’t offer the middleware or management functionalities like Red Hat can, or the security story for those customers.
Google figured this out, that is why they’re getting a strangle hold on the smartphone sector. Canonical thinks they can beat Android/ChromeOS in this segment, but they have generated absolutely zero OEM support for their offering. They can’t even begin to compete with the Intel/Nokia offering, MeeGo.
These are all very similar to the failures behind Launchpad. Despite the number of projects hosted there, no reputable project actually uses it. Linux Foundation members have all flocked around the openSUSE Build Service and git(orius) combination. Instead of helping out the rest of the community though, they insist on being stubborn.
Canonical seem to be doing quite well with legacy desktop substitutes with Ubuntu. But wouldn’t be Canonical be more successful if they based their non-desktop offerings on MeeGo or KDE Plasma, which are designed to be customized for different form factors. Instead they customize Gnome, where the problem seems to be if you are not upstream contributing to the desktop UI, then you are a not a first class player.
Depends what you define as “doing well”. They still count there users in the 50 million region at the high end, and 20 million at the low end. That might sound good, but think about how many Apple and Microsoft have within this form factor and it becomes measly.
They have next to no OEM momentum at all, which means there simply isn’t enough demand. Most of their users aren’t buying machines with Ubuntu pre-installed, they’re downloading Ubuntu for free or ordering CD’s at a cost to Canonical. No matter how you slice it, Ubuntu isn’t generating funds currently, so its hard to say it is “doing well”.
Canonical continuously attempts to enter markets with a less than convincing story for justification to customers. Ubuntu Netbook Edition is yet another example of this.
Does IBM count as part of the corporate world? Does Google?
< tribalism >
In Finland, it’s at 2.67%.
In USA, it’s at 0.72%.
< /tribalism >
Where did you find those numbers? I’d love to see what they are for us here in Sweden.
Play around with
http://gs.statcounter.com/
Sweden has 1.2%. I’m a little bit disappointed with our western neighbour that are otherwise pretty tech-savvy ;-).
EDIT: here are the stats I checked:
Finland 2.67
Germany 1.49
France 1.93
Spain 1.76
Sweden 1.2
Brazil 0.92
USA 0.72
Mexico 0.63
Edited 2010-08-03 20:54 UTC
Dang, heh. Well, maybe it’s that old Finland<->Sweden rivalry, Linus is finnish after all
Well, he’s a Swedish-speaking Finn. Oh, that’ll make a nice battleground :p
Uruguay beats them all with an incredible 5.5%
Yea but the creator of Linux left your tribe for the placid burbs of Beaverton, Oregon. He even blogs about shopping at costco. Assimilation complete
I think it’s a bit bogus to emphasize the marketing aspect of Ubuntu.
What Ubuntu does is – it’s making a distro that friggin works and looks nice. Try any other distro, and you’ll quickly see how polished Ubuntu is in comparison. For many distros, it almost feels like they didn’t even try it before shipping.
Yes, I include Debian in those broken distros. Ubuntu doesn’t attempt to say “you know, this distro is mostly for experienced users” as an excuse. If Ubuntu is broken, it’s considered broken by powers that be, instead of saying “yeah, you just need to do some fixup to config files here and there”.
Ubuntu is really a breath of fresh air in the Linux world. It has really raised the standards and what you can expect to work without manual fiddling.
Edited 2010-08-03 17:24 UTC
Boy, were those standards low.
Yeah, they were. Fedora made an exception occasionally (I still have fond memories of FC1), but Ubuntu raises the bar constantly.
I think everyone should shut up about Ubuntu and get to work. 😉
Ubuntu should really talk to the upstreams early in any new development and coordinate better. When people agree on development it will be much easier for Ubuntu to get their changes accepted upstream as well.
When I see the top contributors of kernel, X.Org, GNOME, etc. I think it’s usually Red Hat who raises the bar most. Well of course the bar raises on all distributions thanks to Red Hat’s way to contribute to projects.
Distributions which support proprietary software gets usually praises in reviews.
Depends on what the contributions are.
Microsoft also contributed to Linux by adding support for their hypervisor. But did it benefit anyone else?
There are different types of contributions and RedHat is mostly interested in the server-side of things.
Common, lots of people were using S.u.S.E. (old spelling intended) happily since the end of the nineties. The Ubuntu installer and administration tools looked pale compared to what Libranet 3 provided. And, let’s not forget about Mandrake/Mandriva, who have always focused aggressively on an easy to use desktop.
SUSE (in every incarnation) contributed *a lot* to the Linux kernel and userland. Ubuntu stole the limelight to some extend from SUSE and Mandrake, but only contributes upstream marginally. Red Hat is excluded intentionally, since it committed its own desktop suicide (due to low margins). Fedora is nice, but not for the average user.
I have always found Ubuntu average (hardware support is good, but there are often regressions between releases). And it is worrisome, that there are so few popular distributions left that contribute significantly to the ecosystem.
I tried the last version and found it boring not pleasant. Looks like Windows 98 in steroids.
I agree. This is probably why Ubuntu is so popular. I recently installed Arch Linux because I wanted a rolling release distro and I wanted good KDE packaging. I was surprised at how hard it was. I knew Arch Linux was supposed to be for power users, but it almost felt like it was out to get me. Thankfully it’s all done now, and it’s just as friendly as Ubuntu now.
I find Mandriva just as polished and new-user-y but with better hardware support. Canonical just doesn’t include the same amount of required hardware support do to the focus on open source with closed source firmware and such being opt-in.
When it does come up in conversation, the usual reason for someone reverting back to Windows is “it won’t support my wireless” or similar; something that could easily be fixed by including the closed source hardware support by default. I think *buntu gets far more attention than it deserves for it’s place in the meritocracy. It’s more popularity than substance. There are more user friendly distributions which have been at it longer but without the marketing push to prop them up. In my own new user days, it was Mandriva’s GUI draketools that finally got me over the hump long before Canonical was a twinkle in Mr Shuttleworth’s eyes.
I also take issue with default setting choices and claiming a production distribution based on testing/unstable parent branches.
At the same time, I’m very happy that Canonical is showing that marketing is important. They also contribute a crapload of patches back to Debian because they do focus on the testing and unstable branches. Those are fantastic benefits which they deserve credit for.
True though, I wouldn’t claim Debian is a new user distro; especially with the way I build it out from a minimal install and custom package selection.
Edited 2010-08-03 19:21 UTC
I’m not convinced of this. I randomly test distros and I didn’t find Ubuntu to be better than openSuse at detecting hardware and Ubuntu has a poor track record when it comes to updates.
I don’t think any distro is leagues ahead when it comes to being user friendly. My opinion is that they all need to bake for 3-4 more years. Linux is fine for servers and cell phones but the desktop is a work in progress.
Distros do a ton of stuff apart from hardware detection. My experiments with opensuse 11.3 indicate that it’s quality is just not as good as Ubuntu 10.04.
Case in point: I was unable to get browser-originated sound (youtube) from my usb headset. I don’t want to debug it, because in Ubuntu I don’t have to. I guess I’ll delete the opensuse partition and try Fedora. THEY should be able to configure PA ;-).
(My work machine is still happily running Ubuntu 10.04)
Well that is your experience and mine is that neither distro is clearly ahead of the other.
Distros certainly do a ton of stuff like upgrades and Ubuntu required 8.04 LTS users to perform a major upgrade just to install OpenOffice 3.0. This is pretty pathetic when 8.04 LTS came out in 2008.
They have a lousy history when it comes to updates and that goes beyond anecdotal evidence. Purchasers of 8.04 Dell machines were told to keep their OS frozen so nothing breaks. Ubuntu also broke some 9.04 netbooks. Call me crazy but you would think that Dell machines pre-loaded with Ubuntu might be tested first before sending out updates.
So we can exchange anecdotal experiences all day but as it stands Ubuntu has very poor history when it comes to updates and on that alone I do not consider it to be “linux for humans” or other such non-sense. I sure as hell don’t consider it to be “linux for admins” either.
Really? I have just installed Mandriva 2010.1 Powerpack and I found it a lot more polished (and complete) than Ubuntu. Hardware detection and configuration were simply amazing (I put it on a notebook).
Same goes for PCLinuxOS. And openSUSE isn’t bad either.
Ask yourself why you think Ubuntu works, then look at who has contributed to the parts that you believe made the experience better.
Canonical is a packager of Linux, they develop some things around maintenance that are nice – like janitor and software-center. They also develop a decent development platform in Launchpad, but very little of the software you think is improving things is written by Red Hat.
I think what makes the difference is that Ubuntu is just tested better. No matter who does all the development, if a component is broken when it’s time to push out the release, your distro doesn’t benefit from it.
Actually, Ubuntu offers some of the smallest numbers of testers around. Less than 10% of their users are running the development releases, so its really difficult to make that argument.
Where the real testing comes from is that Ubuntu are generally at least 6 months behind Fedora, and Fedora users are actually giving worthwhile feedback to fix things.
Ok, let me rephrase – Ubuntu releases are better tested than other distros.
Again, I can’t agree.
Of the *free* distributions, perhaps, but how does that serve the ecosystem well?
No one in the Linux sphere tests as completely as RHEL, no one.
Ubuntu does integration. Maybe they tested it less, but if labels match up in different menus, if file type associates are set correctly, if processed that are supposed to be automatic really are automatic, then it is “better tested.”
Simple experiment: On a modern distribution install and open GIMP and Kolourpaint. Go to File->Save As. What directory do you see by default? You SHOULD see the XDG picture directory; if you don’t then this is an integration failure. This specific example not withstanding my experience is that Ubuntu wins at these kinds of tests more often than most other distributions.
Test 2 – plug in your phone through your usb. Can you configure cellular networking “trivially”? Ubuntu just asks you for your operator and you are set.
Test 3 – plug in an usb headset. How easy it is to make all the sound come from that headset?
Test 4 – configure encrypted home directory
Test 1 – This is due to xdg-user-dirs – which isn’t something that is maintained by Ubuntu guys.
Test 2 – A NetworkManager feature developed by RedHat.
Test 3 – Depends on the the headset, I have never gotten my headset to funtion on Ubuntu or any other distro.
Test 4 – Ubuntu was last to the party here, almost every distro I have used in the last 3 years allows you to trivially set up an encrypted home dir
Polished or blindly copied on other OS ?
I have to disagree. Just look at Fedora 13… It is as well polished as Ubuntu.
Debian is lighter than ubuntu… And simpler – for a intermediate linux user – to use.
Really ? Is that a good news ? I don’t think so. Don’t you need a driving license to drive a car ?
Ubuntu is really a breath of fresh air in the Linux world. It has really raised the standards and what you can expect to work without manual fiddling.
Really ? Is that a good news ? I don’t think so. Don’t you need a driving license to drive a car ?
Sorry to butt in, but yeah, that is a good thing. No OS should ship with broken configuration in place. An OS should work as advertised. It shouldn’t be a minefield of bugs and configuration niggles which an end user is compelled to solve before being able to use his machine.
Regardless of what Canonical contributes upstream, the Ubuntu project delivers an OS that has sane configuration by default. I won’t claim that it is bug free or doesn’t have it’s occasional niggles, but overall, Ubuntu just lets you use your machine without making you jump through needlessly complicated hoops. Best of all, they do it without dumbing down Linux. All the raw power of the system is still available at the tips of your fingers, if you choose to dive in.
On top of that, it makes adding the naughty, proprietary bits and bobs (which everybody but RMS adds) relatively painless, it has the advantage of the depth and breadth of the available Debian packages and it has enough critical mass to get software projects to release their newest releases as .deb files and Ubuntu fosters a vast, easily accessible and friendly user community which is invaluable as a resource of information.
While all the other distro’s have one element or another of what Ubuntu has, none of them have the picture as complete as Ubuntu.
Side note: I think elitists should only be allowed to use computers when they have produced their own silicon wafers, wrote their own ISA, etched their own circuitery and programmed their own OS from scratch.
Wouldn’t it be nicer if we had cars that one could drive with no prior knowledge and no need for a license? If we had such cars which could drive themselves, would you suggest that it was a negative when compared to the ones we have today. Keep in mind that at one time you used to have to turn a crank on your engine to get your car to run, make sure the choke was pulled out before starting, push it in afterwards, and there was no such thing as an automatic transmission. Are we worse off now because driving has become easier?
Ubuntu is about HALF as polished as Mint. Of course Mint is an Ubuntu remix, but still. Cannonical has made many, many puzzling and dumb choices with Ubuntu over the years.
What Ubuntu doesn’t do is – it’s not sending many of its changes to upstream projects. Instead, Ubuntu contributes its changes to Canonical’s hosting website, Launchpad. Then, all the upstream projects are free to create accounts at Canonical’s Launchpad and use Canonical’s version control system, so they could benefit from all the “cool Ubuntu innovations” that are hosted there. And then, if a large number of FOSS projects become dependent on Canonical’s services and tools, Shuttleworth might become a very powerful man in the FOSS world.
It’s a clever strategy, but it doesn’t work as long as most of the upstream projects refuse to become Canonical’s clients. In stead, Shuttleworth is seen as a bad guy because he keeps Ubuntu’s contributions stored at Canonical’s hosting services instead of sending them upstream. And this negative public image makes Shuttleworth less influential in the FOSS world. So it’s really no wonder that Shuttleworth gets very upset when people keep pointing out that Ubuntu contributes too little code to its upstream projects.
So Shuttleworth must just hope that his evil scheme of world domination pans out and that FOSS projects are stupid enough to allow Canonical to become the new centralized upstream for Ubuntu’s upstream projects – not by producing lots and lots of code but, instead, by controlling the tools of production.
When Ubuntu is broken and users are complaining, your overlord at Canonical says: “Ubuntu is not a democracy. We won’t fix it.”
http://www.osnews.com/story/23039/Kicking_in_Open_Doors_Open_Source…
Who gives a damn? Let contributers contribute in the way they can. Is Linux better off because of the contributions of Canonical or Red Hat obviously.
Sometimes it would be better if people kept their own sectarian thoughts to themselves. I can’t believe that Ubuntu server is any threat to Red Hat, so why the upset? both have their own niche in open-source ecology.
People give a damn because if Canonical managed to somehow dislodge RedHat from its leadership position and their roles were reversed, unless Canonical changed its practices, there would be 16 times less people contributing to upstream projects.
Everybody would lose.
This is not an issue as long as Canonical remains a small bit player, but it has grand ambitions and in some areas it pretends to be more than a bit player. This can lead to trouble.
Thing is, if people spend their money on Canonical instead of Red Hat, where does that leave Linux as a whole?
Canonical even has a stated goal of not reinventing anything themselves – that means they would never create things like udev or dbus, which have both made Linux a whole lot better. Granted they’ve changed this, but those changes are being done behind closed doors for the most part. Things like Unity are just bad, the whole community is working on gnome-shell, but Canonical has to do their own thing?
If Canonical wins, Linux loses. You can’t have multiple enterprise leaders in the Linux sphere, Novell proved that. Red Hat has to win in that arena for Linux to keep improving rapidly since they’re the only ones actually driving Linux as a whole forward.
Both Vistauser and Lunitik make the point that if Ubuntu “wins†and Redhat looses this is bad for Linux, however, this seems to me to be based on a wrong headed view, Linux is not a single hierarchical structure with all the contributers attempting to race to the top. Its not a case of Redhat or Canonical assuming “the†leadership role but “a†leadership role. I think this wrong headed view may explain why Lunitik “hates†Ubuntu which is a shame.
Now turning to some other points, It seems to me that through Ubuntu, Conical has tried to create a usable desktop, and make it known, and available to ordinary users. To this it has had some success. I disagree with Lunitik that Shipit was a waste of time, for me it was the only way to get Linux I was on a 33K modem (with only a lease line alternative) at the time and the nearest Linux magazine a 200 mile round trip (although I did get a Suse boxed set once). Certainly Shipit was significant for me, less so now that broad band is now beginning to penetrate even into darkest Africa. I have tended to stick with Ubuntu and not because it is crap, having used Xandros and Mandrake in the past I would say that Ubuntu is certainly better than Xandros and I prefer its less proprietary philosophy, more to my taste than Mandrake now Mandriva and as a desktop more responsive and easier to setup than Opensuse. Is Ubuntu bug free – no but is Fedora or even a non Linux OSes such as Vista? No for me on the desktop no bug has been a showstopper and usually easy to fix or work round.
I don’t see, how Canonicals effort here has damaged Redhat and would argue that Ubuntu has improved the culture of Linux for desktop users from the bad old RFM power of the commandline days. Is Canonical steeling business from Redhat? on this I think I would take an old fashioned competition is a good thing line, but are they even in competition? If I was setting up a Enterprise IT system I wouldn’t be thinking about Canonical (am I being unfair – possibly) but here Redhat is dominant. One might consider using Ubuntu server in a small to medium sized business where Canonical aims its Ubuntu server. I would have thought that CentOS was more damaging to Redhats business than Canonical.
Would I like to see Canonical make more contributions upstream, yes, but I understand Canonical has had some difficulty getting some of its modifications excepted upstream. I have no doubt that Canonical has no wish to kill the goose that lays the golden egg and little doubt that canonical will attempt to play its part as a good citizen
Edited 2010-08-04 08:16 UTC
Red Hat basically hires every important developer in the Linux ecosystem. They are leaders in every facet of Linux, and they are the single most innovative company in the open source arena. Canonical is a bit player, they do not innovate as a rule. I wish people would stop crediting Canonical for getting the word out about Ubuntu too, that was their users by way of word of mouth. Canonical has never had a hand in that.
All you can actually say about Canonical’s contributions to the desktop is they have created nice themes – and even then, most do not like their color schemes. People read reviews, and they dislike distros like Fedora because of the default theme, not realizing that it is trivial to change such cosmetic things. Beyond that, they have created tiny semi-useful tools like computer-janitor and done things in a lazy way such as apport and upstart. They have also ridden the coat-tails of Debian and Synaptic and built on top of those tools some nice interfaces for package management. They do next to nothing within important upstreams that span the entire community, and don’t deserve any type of leadership role because of this.
You don’t see how taking money away from Red Hat – away from the guys actually developing software – is damaging? Can you imagine where Linux would be if Canonical put Red Hat out of business? Canonical has no development expertise, they are a packaging company that can throw together lame python apps that do simple tasks. Until they start contributing in a significant way, I will always consider them leeches.
You know what other companies do when their contributions don’t get accepted? They clean them up until they are acceptable. It really speaks volumes in my opinion to the Canonical developers skills if they can’t get things accepted in places like GNOME, it means either they can’t write portable code, or their coding standards are laughable.
I think this http://www.jonobacon.org/2010/07/30/red-hat-canonical-and-gnome-con… is interesting and shows Canonicals perspective.
Edited 2010-08-07 08:38 UTC
In spite of Red Hat’s significant technological contributions to Linux in general, and the inroads they have made with Linux on servers, they never made much of a dent in the average Joe desktop market.
That, though, is where Canonical has focused its efforts, and made more progress than any other organization as far as I can tell — including Novell/SuSE, Redhat, and various other distros.
Ubuntu’s not my cup of tea personally, but I think Canonical deserves to be recognized for increasing Linux’s profile in the mind of average computer users. I hear the word “Ubuntu” roll off the tongues of my non-techie friends more than I would have thought, and certainly more than the name of any other distro.
Edited 2010-08-03 18:34 UTC
And they never wanted to, either. Canonical does, and should be applauded for that. When Ubuntu started out, there was no distribution that had the same role as Red Hat (the distro with that name) used to have. There was RHEL that was very expensive, and there was Fedora that was a bleeding edge distro for hobbyists. What was missing was a free production quality distro. SUSE was not free either. Slackware, Mandrake and others were just bad in their own ways.
Back then Debian was failing to release in pathological fashion, because there was no central authority – so it was pretty bad option for most desktop users.
Where did Ubuntu make a dent? Linux has the same 1% and 40% of those users are running KDE.
If Linux had doubled in share from the first release of Ubuntu I would agree but this is not the case and the fact that the KDE/GNOME divide still exists shows that Ubuntu has just been the most popular Gnome distro, not some new force making inroads within the mainstream.
Sure there are plenty of blog posts about switching Grandma to Ubuntu but there were plenty of those types of posts before Ubuntu existed. If you look at distrowatch you can see that Ubuntu is not that far ahead of the other distros.
There goes the troll again.
Well, he seems right about one thing: it’s now that the 1% grew, but that it is differently divided. Without Ubuntu, Linux probably kept the 1% marketshare, but SUSE and Fedora would be kings of the hill of Linux on the desktop.
Besides all of that, it paints to bleak a picture. GNU/Linux made a lot of inroads on the sever, and embedded devices (routers, access points, you name it), and increasingly in mobile devices. Linux on the desktop is not winning, but Linux on devices is.
Well, if you move outside USA, one will see that Linux is much more dominant on the desktop than OS X. On a global level Linux and OS X are estimated to both have around 5% market share with Linux being the larger one outside USA.
But even if we look at U.S. Linux numbers one will see that Linux has grown from close to 0% a decade ago to almost 2% now. And the *BSD’s are almost at 1% now (on the desktop that is) using the scewed numbers from U.S.A.
There is no doubt Ubuntu has done a lot to increase knowledge of GNU/Linux, and I can’t see that as a bad thing.
In the European countries I have been (including The Netherlands, where I live), you see far more MacBooks than laptops running Linux. I can’t imagine it’s any different on old-fashioned desktops.
So, please provide sources if you state percentages. Since, nearly every statistic for Europe shows the opposite. One typical example:
http://gs.statcounter.com/#os-eu-monthly-201007-201007-bar
Again, here are the Linux stats I posted elsewhere in this thread:
Finland 2.67
France 1.93
Spain 1.76
Germany 1.49
Sweden 1.2
Brazil 0.92
USA 0.72
Mexico 0.63
gs.statcounter is irrelevant. Like most other statcounters it is highly unreliable. The stats for China is enough to undermine its credibility.
Actually, it is well known that WinXP (many copies of which are pirated) rules the roost in China, especially when those Internet cafes are taken into account. All in all, those OS usage stats are hardly surprising.
So it is unreliable even though they provide their methodology:
http://gs.statcounter.com/faq#methodology
What are we supposed to use? Your own unsourced data?
What really surprised me is how tiny is the Linux bar in Asia or Africa. Being those continents supposedly poor (not all of Asia, though) one would expect the opposite.
Poor people pirate windows, or don’t have a computer.
Linux is favored by technically sophisticated users, which typically are not in the low income slice of the pie.
You can windows copy for less than 5$…
In south korea, you have to use Microsoft software or you cannot use internet based services.
That is very true and it had also come into my mind.
Confucius says: Man who wears retarded hat should not call others names.
People aren’t spending money on Linux on the desktop, developers need to get paid to do the best work.
You think Red Hat is paying for all this Desktop development based on its desktop sales? Shuttleworth has around $500 million to his name, so its easy for him to support a small team, but look at how he’s used it. They do basically nothing at any level of the Linux system, just stuff to set them apart outside of the important upstreams.
SUSE/openSUSE and Mandrake/Mandriva focused on desktop Linux much before Ubuntu existed. True, not all their releases were perfect, but many were.
For me Ubuntu has meant nothing but bugs.
Heck, even Lindows/Linspire was a great “Joe User” distro (before going ‘buntu)
Or Xandros.
Why did they die? Because Ubuntu taught people you don’t have to pay for a Linux distro.
Personally I was more than happy to pay for good distros, like the great, unrivalled Libranet.
You are very right. One of the things that contributed to the end of paid distributions was Ubuntu. I was part of the very small Libranet team, and I always felt this was one of the reasons of Libranet’s demise. It’s hard to compete with a multi-millionaire who ships stacks of free CDs to anyone requesting it. Still, for many years after its demise, Libranet’s installer and administration tools were far more user-friendly and useful than Ubuntu’s counterparts. This was all written by two paid guys and managed by another paid guy. Of course, Ubuntu was not the only cause of Libranet’s end, but certainly a contributing factor.
I feel very sorry for Mindriva and its users that it may have the same fate.
Edited 2010-08-03 20:58 UTC
Libranet ended when the major guy behind it died (I forget his name). It was a sad day indeed and a short downward spiral from there.
Disclaimer: I was one of the three Libranet employees. The loss of Jon hit Libranet hard, but this was not the only cause of its end. Before, it was difficult to sell enough Libranet 3.0 sets to be able to finance Libranet. Again, it is very hard to compete in a market where a millionaire throws around free CDs. And don’t be fooled, in the end most vendors, including Canonical need money to fund development, but they were in the position where they could easily do damage to many existing players, and happily did so. And this is my problem with Canonical: they are so focused on raising their own profile, that they forget that they thrive on the community. From Red Hat all the way down to Fluendo.
When someone points them to their weaknesses, rather than addressing them, Mark accuses others of tribalism. It’s not tribalism, but a simple message: contribute a bit more to the community.
Edited 2010-08-04 19:11 UTC
I don’t want to belittle Ubuntu’s accomplishments, because they really did raise the bar in terms of out of the box usability, and they made the rest of the Linux distros sit up and realize that mediocre would no longer cut it, but do you know that Red Hat isn’t even trying?
Seriously! They gave up on Desktop Linux around 2000, back when it was a much harder sell. Fedora may exist and get a certain amount of incidental backing, but there is no Red Hat Desktop Linux effort. The reason Ubuntu is better at this than Red Hat is that Red Hat are not attempting to be in that market because they don’t know how to profit at it. Which, as it happens, Ubuntu isn’t doing.
So let’s not compare apples and oranges.
So my Fedora 13 powered MacBook is not a desktop system?
Does it edit pictures? Check
Does it Play MP3’s ? Chek (add another repo you yum )
Does it work with 3G wireless Dongle ? Check
Wireless OOTB ? Yep.
So, unless I’m very much mistaken pretty well everything I need from a desktop is there.
Sigh.
Unfortunately thanks to Canonical’s viral marketing far too many people thing that Linux == Ubuntu and nothing else.
Believe me, I was once an Ubuntu fan. The aroun Fiesty, the lova affair soured. Buggy beyond belief. So I switched and never looked back.
I certainly don’t have ‘apt’ withdrawal symptoms.
Fedora has an explicit role as “hobbyist” distribution, with no long term stability promise for any release. It’s a victim of anti-marketing: no matter how good it is, it’s downplayed by the makers themselves.
If Fedora was established as production quality distribution right from the start (like Red Hat that it replaced), we wouldn’t have needed Ubuntu. But history happened.
Fedora is not a Red Hat system. It’s a community-run distribution. Red Hat does not put marketing muscle behind it at all, not in the way Ubuntu, Microsoft and Apple market thier OSes. Red Hat prunes technology and shares some of the base… but you cannto buy a Red Hat desktop system. You can get RHEL in a workstation flavor, I believe, but this is more of an afterthought and it’s made to complement the server version and not to stand alone in the home.
Canonical has done very little to actually market Ubuntu itself. Even the billboards etc, they were done by the users, not Canonical themselves.
Ubuntu is popular because of word of mouth, not marketing per say. To say Canonical contributed to that is a stretch.
Sending as many CDs as you wanted doesn’t matter a bit?
That isn’t marketing, though.
How many people do you think actually stuck to Ubuntu after getting those free CD’s? I know a lot of people from my former tech college that ordered it because its free, installed it once to see what its like, but removed it and went back to their previous OS.
It certainly hasn’t resulted in Canonical earning money, so its hard to say it was a successful endeavor.
A lot, in my own experience.
It could help a little.
At the RMLL 2010 in Bordeaux (France), I heard that Ubuntu-Fr doesn’t get money from Canonical at all…
Seems obvious to me that Canonical has been moving in on Red Hat’s turf (likely making offers to Red Hat clients) and while Red Hat’s response may seem like it’s directed at Canonical, it’s really aimed at the customers “See who is doing the real work and make sure you add that to the equation when you consider Canonical’s offers”.
Most important for Linux today, Red Hat obviously, since Linux on the desktop is not exactly big business. However, Canonical is in my opinion the first company really making an effort of pushing Linux on the desktop and Ubuntu has obviously made Linux more accesible for desktop users, so who knows what the future may hold.
After 6 years running Canonical, they are still not making money as far as I know. They haven’t added much of worth to the desktop either, despite getting all the credit.
Udev was done by a former Novell guy, working now with Google – that was the big one that Ubuntu profited from. They became “the first distro to ship project utopia” despite doing nothing to actually assist in that work.
Name anything else you use on a DAILY basis, or things that actually assist in making your life easier on Linux… nothing that adds to the experience of Linux on the desktop has actually come from Ubuntu, nothing. All they’ve done is added around the edges, and allowed proprietary drivers by default.
Well, I’m not pro Canonical or anything. I’m just saying that I think they’ve done a good job (as in better than anyone else before) in pushing Linux on the desktop. To compete with Red Hat for enterprise customers though, they will have a hard sell. Would you rather buy your services from a company that can actually fix problems for you (Red Hat), or one that has to wait for them to be fixed upstream (Canonical)? Red Hat’s comments on Canonical’s lack of upstream contributions underlines this for existing/potential customers, which again is what I believe was the whole point of Red Hat starting this debate to begin with.
To be clear, Red hat didn’t start this discussion, an ex-redhatter did, Greg no longer works with the company.
I hate Ubuntu much less than I used to simply because I realize what you’re saying is true, though. Canonical is attracting nothing of value to Linux. All they’re attracting are users, and very few that actually know what they’re doing stick around. Even less ever actually pay a dime to Canonical. In fact, I think the only source of income to speak of for Canonical is coming via the development deal with Google.
Anyone actually looking to do anything serious with Linux is still going to Red Hat with their business, which is definitely a good thing.
Ubuntu is widely used as a development workstation. That’s somewhat serious stuff too, in the sense that people get paid actual money to do it.
I felt sad everytime I saw my colleagues pressing the Up key dozens of times to retrieve a command, when they could have just pressed the PgUp key once if Ubuntu ships with a developer-friendly /etc/inputrc.
I felt sad when Ubuntu decided to be the first distro to ship the shiny but untested distribute in place of the old but steady setuptools in an actual release. A LTS release, no less.
I felt sad when trying to install Greenplum on Ubuntu 10.04 and couldn’t get it to work without some funky workaround. I have always wondered why vendors would only list RHEL and SLES as supported OS, when their products always work fine on Gentoo. Now I know why.
I felt sad when using Ubuntu on production servers became the logical next step after using it on development machines. The apache version on Ubuntu 8.04 LTS is still stuck on 2.2.8 with many unfixed mod_proxy bugs. The memcached version is still 1.2.2 with a serious connection bug.
Ubuntu is a sad distro for developers.
How many of those developers are actually paying to utilize Ubuntu? Very few would be my guess.
When I said “serious” I meant people doing mission critical jobs. No disrespect to developers, but other than losing your unsaved work, there are few risks to a companies bottom line if your system goes down for a while.
Attracting users to Linux is very valuable, it encourages developers to consider Linux as a viable platform.
Redhat and Ubuntu are productive in different ways.
Redhat and Ubuntu have different approaches.
Redhat and Ubuntu are not SCO
Edited 2010-08-03 22:04 UTC
First off this is a silly argument. Red Hat makes their code freely available under the GNU license. They can’t complain because someone else packged it better than them. They still get the credit in the source.
Also, i’ve seen a lot of FX houses who use Ubuntu heavily, because its generally stable, free, and they already have all the support they need in their highly paid Linux admins. Ubuntu has better hardware support out of the box, the installer and updater are damn near seamless compared to how things used to be and the biggest reason why Ubuntu took off, its one freaking CD. You used to have to download a whole DVD full of crap just to get a halfway decent distro. Ubuntu focused on keeping everything to one CD and let the user use their repos to add whatever they need.
That’s one of the main reason why I see developers or FX houses use Ubuntu, you get most of the benefits of Red Hat with the benefits of Debian, all in a nice package. I used to be a redhat user and moved to Fedora when redhat decided the deskotp market wasn’t worth it. Every release of Fedora has treated their users the same, as guinea pigs. Buggy installer, buggy software releases. Ubuntu at least TRIES to have a modicum of stability between releases and anyone who calls ubuntu’s releases buggy have either not used Linux long enough or have really short memories. I have horror stories of so called desktop distros like Mandrake, and Red Hat. Ubuntu is rock solid in comparison.
I think a lot of the hate for Ubuntu stems from jealousy. I don’t mean to sound liek I drank the kool-aid or buy into their marketing but the distro really i Linux for regular people. When Ubuntu first came out and forums first came up the forum was probably the friendliest linux based forum anywhere at the time. The users on the forum were genuinely trying to be helpful and newbs were free to ask questions without having some deb guru berate them. I couldn’t really say that same about Fedora. I went on the forum to ask a question about an installer bug that was their for four consecutive releases, asking for a workaround and got ignored, and saw that others with the same questions on the forum got ignored. Interestingly the answer was in the ubuntu forum as the forum members didn’t seem to mind if you posted about another distro, they would still try to help you. Ubuntu is where its at because that kind of attention is what users want. Word of mouth is what got Ubuntu where its at, the rest is just gravy.
Anyone who has gone through the Intel driver debacle with Ubuntu 9.04 will disagree with you.
So users have no value if they don’t know what they’re doing, then? Microsofts accounting department would disagree with this.
Think of the developers that could be hired with a tenth of Microsofts revenue from Windows? Of course that would require a user-oriented mentality to create something users would want to actually pay for in the first place.
Chicken and egg problem. This kind of viewpoint hinders Linux’s acceptance. Canonical’s more “user” oriented approach advances Linux’s acceptance (which btw, also helps Red Hat).
Microsofts sales of Windows to normal “users” dwarfs Red Hats pathetic (relatively speaking, of course) revenue. Something to think about before jumping on the non-developer-users-don’t-matter bandwagon.
Don’t be too quick to dismiss things “added around the edges”. It only takes a couple buttons not working or a few mysterious errors (easy to correct if you launch from a terminal and read stdout, or know which config file to tweak!) to drive Joe Average away from Linux, possibly for good. It is precisely dismissing this kind of low-value problem that keeps Linux from desktop penetration. The only distros I know that have cared about it are Ubuntu and SuSE, of all people.
Greg KH still works at Novell.
I guarantee you that if you ask Joe and Josie Computeruser to name a Linux, they’ll say “Ubuntu.” If you ask a CIO or CTO, they’ll name Red Hat and possibly Novell’s SLES (SUSE Linux Enterprise Server). But Canonical has done more than all the other Linux companies and groups put together to popularize Linux with ordinary people.
Despite the “great work” Canonical has done, if you ask random computer users to name a Linux distro, they will still look at you with the same confused stare they did 10 years ago.
If you don’t believe me, right now, go to any public place and just ask 10 strangers. I guarantee none know what you’re talking about. If that’s the huge difference Canonical is supposedly making, they’re failing worse than even I thought.
Personally I don’t understand all the praise heaped upon Canonical for Ubuntu. For me nearly every release had some sort of catastrophic bug that forced me to dump it in favor of something else, usually Fedora. Case in point, 10.04 suffers from a frequent issue on laptops with automounting USB volumes after boot. The user forums were practically useless in diagnosing the problem, most of the lame-brained suggestions for a workaround were silly at best. And of course none worked to resolve the issue. Please note that I managed to determine that EVERY distro based on Ubuntu 10.04 (including LinuxMint 9) suffered the same problem. My ultimate answer was to go back the Fedora world where everything really does “just work”.
I have NEVER had a Fedora update to break things that worked at the original install, but Ubuntu updates broke things regularly. Don’t take my word for it, read the forums.
The bottom line, Fedora works because REAL software engineers and experienced coders put out a stellar product. Ubuntu is simply more flash than dash. Even release versions feel like beta or RC versions since they are so buggy, and I truly do not believe that Canonical has the resources to fix them.
Please don’t misunderstand, I desperately wanted Ubuntu to work for me since package selection is second to none. I just can’t tolerate the broken functionality.
Finally, this was posted using Chromium in Fedora 13 running on a Compaq EVO N610C with full Compiz. Rockin’ baby.
Red Hat does do more for Linux then Canonical. However, Rad Hat is also making money on Linux when Canonical isn’t. Also they are 2 very different companies which is good because if Canonical tried to copy Red Hat they wouldn’t standout and they would most likely fail. Its not like Canonical isn’t doing important work that does help Linux.
PS
If you want to bash a distribution for taking advantage of the hard work that is being done by Red Hat without giving enough back then why isn’t anyone bashing CentOS?
Just my .2 cents
Dan
I just want to say that Ubuntu’s success has less to do with it being a good “marketer” and more to do with it producing a distro that works better than the others (meaning simpler, less bugs for everyday people, etc).
People who think that Ubuntu is popular just because they market their product well are really missing the point about getting linux mainstream-nongeek acceptance.
It’s about the little things, not huge (often useless) features full of minor issues that are never fixed for 5+ years, all the while leaving older very annoying (for mainstream-nongeek users) bugs unfixed.
Papercut’s is a good example of what I mean.
You know what else makes Ubuntu/Canonical different?
It’s the ideological thrust. Ubuntu is the only distro that still aims at world domination on desktop, with no regard to cynical realism. It resonates with those of us who think it’s a reasonable aim, provided that there are those attempting to do it. For example debian never attempted to be popular (they seem to be mostly interested RMS-purging the desktops that are already there, or servers).
Ubuntu also has a strong leader/jesus figure and it’s not democratic. This guarantees it won’t get stuck in internal struggles (as happens with debian on occasion). It’s the distro that will the most succesful one in the future as well, in terms of mindshare if not financially.
If Canonical ends up not contributing directly to upstreams while doing it, so be it. This may come as a part of not being democratic – Ubuntu does its own thing regardless of whether the upstreams agree or not. Upstreams don’t slow it down by stop energy.
Except that Ubuntu is heading in the wrong direction IMO and no one is going to be able to change whatever decisions made by the S-dictator. Some developers have actually left because of this. Any wonder that quite a few Ubuntu-based distros are silently moving away from it?
Remember times in school, university or at work – who are the best known and sucessful people? Those, who have good communication skills and shouts loudest. They may be no good at some skills, but they are known and people listens to them. In opposite, many engineers and other people with high competences are often called nerds and many of those, even geniuses end up working for those having good talking skills.
In Linux world, most development is done by RedHat and Nowell. Ubuntu develops some software, some integration tools, etc, but they show Linux for broader audience. They may be not contributing so much code – this is done by those silent engineers, but they are makeing people talk about them.
Reading through some of the links, the initial cause of all the hullaballoo was the Gnome Census which showed Redhat making significantly more commits to Gnome than Canonical. However, this is misleading since it’s counting commits over the entire course of the Gnome project while Canonical has only been around for, what, five years? If you looked at the commits made in 2009, then the Redhat and Canonical numbers would be similar. And the ragers would have to find something else to scream about.
Running Fedora 13 with KDE4.4. I don’t have a dog in this fight, but it just strikes me as an Internet argument based on numbers taken out of context.
Not really, no. Canonical has been around about 6 years. Gnome has been around 11. That means that Canonical has been around half of Gnome’s life. As such, if they were anywhere near Red Hat in commits per year, they would have half as many as Red Hat. Canonical has 4500 commits to Red Hat’s 71000. The math doesn’t lie. Canonical just doesn’t do very much development. Maybe people would be less hostile if Canonical occasionally thanked Red Hat in public for all the hard work they do instead of just taking all the credit.
Actually, it is worse than zero-sum:
http://google.com/trends?q=linux%2C+ubuntu
I’d really like to see them contribute a lot more. Instead they ship broken Plymouth, Intel drivers and various other bits.
Canonical should help fix the lower stack, specifically; Intel drivers like i915(they blacklisted some for Lucid while Fedora shipped working ones), the Nouveau driver, Plymouth.
They should have stuck with Splashy instead of trying to get news and marketing with the other cool toys and breaking them.
Cannonical’s marketing SUCKS!
Brown? BLEGH!
Buttons on the wrong side? Why?
Linux Mint is twice the OS and has a much friendlier community, better default layout, and over all a much more logical choice on pretty much every level.
Mint FTW!
Funny then how Mint is entirely based on…Ubuntu.
It’s painfully apparent, after reading comments over the last week from coders and community members alike, that this will never be resolved and the best thing that Ubuntu can do is move on and do what they feel is best for their distro and their community. If they want to continue to use Gnome code, then either overload the gnome project with commits (even if none are accepted) to shut them up about it, keep using and contributing to it as they do now, or fork and forget those crying because others are not playing the game well enough for their taste. The same goes for RedHat.
My opinion (not for the easily offended)–
Lee Iacoca once said that “you can lead, follow, or get out of the way”. Ubuntu and RedHat need to decide where they are in that line. No one is forcing anyone to code or use open source projects and though I’m no expert on the GPL, v2 or v3, I dont recall anything that says you must contribute x amount of commits in order to use code from this project–including Gnome. For real, what is anyone going to do about it? Anyone have the balls to take their GPL code and prohibit another group from using it in accordance to the license agreement because they dont like them? That’s right—No one. So for those bickering over this “who’s better than who” needs to shut up and grow up. Dont like it? Code for Apple or Microsoft or join their group of fanbois.
I would LOVE it if Ubuntu decided to fork everything, then you’ll see how strong their expertise really are. That would be hilarious to watch.
Edited 2010-08-04 21:15 UTC
Canonical already started from Debian repository (universe and multiverse) to Gnome called Ayatana (https://launchpad.net/ayatana).
If that is the way to contribute to Linux ecosystem, I will say: no thanks. Instead of recognizing their wrong doing and actively participate to the community, Ubuntu founders keep shifting the blame.
If that means to be attacked by fanboys, so be it.
Ayatana isn’t a fork of anything, it is just a rather poor implementation of the notification system that will be going into gnome-shell.
It looks like Canonical has decided that, while the rest of us move on the Gnome 3.0, it wants to maintain a desktop that is more similar to the current state of affairs. If you look at their Unity mockups, you can see what I’m talking about. < http://static.arstechnica.com/assets/2010/05/shell-desktop-1-thumb-… >
Instead of contributing to either MeeGo or Gnome-Shell, they have decided it best to maintain as much status quo as possible moving forward. They are trying to separate themselves from the rest of the community similarly to how Novell has in the past (slab and choosing mono apps over everything else.) It didn’t work then, and I hope Canonicals attempts are met with just as little momentum.
Here is a simple way to answer this question:
If both Red Hat and Canonical went away tomorrow, which would be most missed?
If Ubuntu went away, nothing changes, everyone just goes to a different distro, and they are forgotten in time like the likes of Libranet and Linspire before them.
If Red Hat went away, the entire Linux community would crawl to a standstill trying to pick up the slack. The community would lose its best contributors, its leaders, and would likely never really recover.
I would say this is checkmate on this argument.
I honestly don’t like that this is true, but that is the repercussions of Mark investing his money in the areas he has chosen to.
He isn’t actually benefiting Linux as a whole. Launchpad is competing with the undisputed champions in several categories: git, bugzilla, transifex (and gitorous/github, not to mention openBuild – the OpenSUSE build system used by Linux foundation and Meego). Unity is trying to compete with Meego using parts from gnome-shell. Everything he’s doing is really going to waste within the community at large, they aren’t the best options in any area.
The likes of Red Hat and Novell are much more essential to Linux as a whole because they’re investing in things that benefit everyone. Mark is concentrating on things that make his distro different instead of being a good community citizen.
EDIT: I hope Mark reads this, and adjusts where he directs his developers.
Edited 2010-08-05 03:27 UTC
This is just not true because:
– It’s all open source
– People would be hired to continue their work in a heartbeat. Do you think they would need to look for a job in a different sector, really?? (PS: Nokia is always hiring Linux gurus 😉
– KDE is an independent alternative to Red Hat-led Gnome.
Yes, you’re right.
There have been similar cases (well, more or less) where alternatives have been developed for specific needs. One of the most famous cases is Git.
BUT (and this is one big BUT):
http://fedoraproject.org/wiki/RedHatContributions
http://sources.redhat.com/projects.html
http://www.redhat.com/truthhappens/leadership/osdevelopment/
http://et.redhat.com/page/Main_Page
Etc …
In the event that Redhat stopped support Linux Development (with code, developers and infrastructure) the whole ecosystem would suffer to the ground. Of course, other companies or individuals would continue to work but the transition would be a real *****pain in the ass*****.
Could you say the same about Canonical?
.
And the little or much success that Ubuntu may have today is due primarily to ***Debian*** and ***Redhat***, but above all, due to this sentence that Ubuntu community seems to have forgotten:
“If I have seen further, it is by standing on the shoulders of giants”.
See, but Red Hat is also the leading developer in the Xorg stack, the most active at freedesktop.org, the leading kernel developers. The most active throughout GNU projects. Red Hat pays developers handsomely compared to people like Canonical, so there is no guarantee that they’d stick to Linux development if Red Hat was no more.
Also, companies like Nokia and Google would significantly change the focus of Red Hat engineers they’d hire should Red Hat cease to exist.
EDIT: (GNU projects Red Hat maintains include glibc, gcc, gettext, coreutils)
Edited 2010-08-06 02:11 UTC
I’m both a corporate sysadmin, and a home user of linux. At work we’re using CentOS and at home I use ubuntu. Frankly, Red Hat RPM based package management sucks. It always has and it always will. RPM is Redhat’s greatest weakness and half the reason Debian survives as a distro.. if Redhat switched over to apt-get and .deb packages it would probably deal a large blow to ubuntu’s market share. Then again, having everything in/usr instead of /usr/local is very convenient. Sure directory listings can take a while but seriously that’s what tab completion and other cli tools were made for. In terms of ease of use Ubuntu is dominating in this area at the moment simply because rather than making a custom solution they are creating/using open tools that are being adopted as the standard. Things like network-manager were adopted into Ubuntu first. Rather than write a crappy perl script like mandrake or their own configuration utility. They backed the standards through freedesktop.org. This is the way the entire stack needs to be done. Red Hat should have jumped into this stuff years ago and not tried to force everyone into using their package management system. (for the record I’ve used RPM distros, .deb distros, and gentoo/ebuild script based distros.) .deb has been the easiest/least prone to screwing up in my experience. Your Mileage May Vary.
RPM got a bad rap many years ago because everyone compared it erroneously to APT, rather than DPKG. I personally find YUM to be a delight compared to APT-GET, so I’m not sure where you’d get this opinion from today?
As for NetworkManager, umm, Red Hat wrote it. As for general usability, well, Ubuntu actually does very little of it themselves. Their package management GUI’s all use synaptic at the backend, that has been around for close to a decade longer than Ubuntu. I honestly can’t think of anything else you could be talking about?
You did not clearly explain why RPM Package Manager (the actual acronym name) sucks. Speaking as one of Fedora packagers, it appears your assertion is based on the use on apt-get, used to be proclaimed as über. As sysadmin, you should know about DPKG, Debian lower-level package manager equivalent to RPM.
RPM is actually a strength for Red Hat. Writing a spec file to build source RPM is not hard with some assistance. RPM was not been maintained but with Mandriva and OpenSuse gathering altogether with Red Hat, it is getting better. Remember, RPM is LSB standard which why Debian folks have made alien for converting it to DEB and vice versa.
Perhaps you shall start using Ubuntu Server at work then, and see if your view changes. I am sure you will have a lot of fun with the user-friendly apt-get/deb combo that:
– automatic starts a service for you right after you install it (i.e. before you get to configure it)
– requires a service to stay up before you uninstall the package, so that it can perform the service stopping for you
– changes random config in php.ini/postgresql.conf/etc because Ubuntu is so much better than upstream
Thanks Professor Obvious for pointing out that the company that is many times bigger, with many more engineers and with the much larger revenue contributes more and is more important. We’d never have figured that out on our own.
Thing is, even the developers Canonical do have aren’t working in significant upstream projects, so this is really a pointless statement.
Most of the contributions Canonical has made to things like GNOME are simply bug fixes. While I’m not understating the importance of such things, without the original code, there isn’t much to fix. Projects like Debian and Gentoo manage to contribute more despite being entirely volunteer based, that is pathetic.
So what? There’s no clause in the GPL that says you have to contribute X amount of code. In fact, if you look at the whole picture most distros contribute far less than Canonical. To be honest, I don’t care who contribute what, Ubuntu satisfies my needs and that’s why I, and many others, use it.
No distro that isn’t volunteer based contributes less than Canonical, and several distros that are entirely volunteer based contribute more.
My only point is that its a sorry state of affairs when you do next to nothing, and yet garner most of the praise. I have yet to see a review or anything similar that actually praises Ubuntu for something they’ve done. The only mention of anything Ubuntu-specific in most reviews in fact relates to its theme, which consistently is discussed negatively.
I’m not sure when theming became a major discussion point, they are probably the easiest thing to change, but many consider distros like Fedora old fashioned and similar because it uses the upstream default. In reality, you’re getting most of the things you praise Ubuntu for 6 months earlier.
Red Hat developed the underlying functionality that provides windowing effects – AIGLX. Red Hat developed NetworkManager, which has greatly improved network configuration on Linux. Red Hat developed system-config-printer – hell, it even maintains the naming construct of all Red Hat system management tools. Red Hat used to hire most of the guys doing Linux specific Firefox work. You name it Red Hat is probably responsible for it.
Canonical needs to do more, or give credit where it is due.
a, i like their color scheme. Not everyone share the western worlds preference for blue.
b, if it’s so bad why is ubutntu so popular?
Yes, that’s already been said by Canonical too
Canonical hasn’t taken credit for anything they havent done.
If people had actually read the cencus and listened to the guy who made it we wouldnt have had these pointless flames.
Edited 2010-08-07 01:31 UTC
So what? There’s no clause in the GPL that says you have to contribute X amount of code. In fact, if you look at the whole picture most distros contribute far less than Canonical. To be honest, I don’t care who contribute what, Ubuntu satisfies my needs and that’s why I, and many others, use it.
One more kick on this dead horse –
I already tried the new opensuse and rejected it as broken.
I have been running fedora 13 for a few days. I was *this* close to saying it’s actually a pretty good distro. Now, though, I switched to compiz and alt-tab stopped working. It acts the same as “tab” on my system. Switching off compiz fixes the problem.
Ubuntu (esp. Lucid) is still the best distro among the ‘name brands’, at least on my computers.
I guess I’ll try debian squeeze on this hobby computer next…