Slashdot Mirror
UserLinux Continues Debate Over GUI
An anonymous reader writes "Following up the earlier Slashdot item on this, LinuxWorld is carrying both sides of the discussion as to whether UserLinux GUI should be GNOME only, as Bruce Perens last week decided "by fiat," or include KDE."
They're targeting the enterprise market -- defined explicitly by Peren's whitepaper as (paraphrasing) the market that Red Hat gets $$$ in support fees for.
I just don't see companies who want that level of support settling for "here's our linux distro, and if you want support, uh... here's a list of 3rd party providers." Remember LinuxCare? They're still around, but only because they moved away from providing third-party support solutions.
I prefer gnome to KDE. And I see that KDE is more advanced.
;)
by Bruce's decision I see some hope for gnome to speed their development. That's why I think it is good.
I also hope that decision about mozilla and gumeric + OOffice will solve their greatest problem: its own widget set. (I prefer galeon to mozilla - mainly because of native widget set == less bloated)
to conclude: I think that Bruce made the best possible decissions, and I really hope it will be a great success.
(btw, sawfish is my WM, not gnome, which for me is too bloated
#
#\ @ ? Colonize Mars
#
So it's anti-competitive to prefer one option for doing the job that you want done? Was I anti-competitive for un-selecting the KDE checkbox when I did my minimalist install? As I see it, this is a prime example of competition that they can thrive on - they both have to prove their case and one gets included - if thats not competition then I don't see what is.
Please ! Stop to feed the troll !
.doc is the winner of the .sxw/.abw war !
The mailing list becomes really full of this old and crappy Troll...
There's no place for two DE in UserLinux... So there's a choice. If KDE was choosen, all gnome's fans will stand up !
I talk a lot about the Desktop and I continue to receive comments about that.
Well ! everyone has is own choice.. We don't have to speak so long about it !
Who is the winner of the Gnome/KDE war ? MS Windows of course ! Like
Ploum.net.
No, the real question is: why do any of us give a rat's ass what GUI a vaporware Linux distro uses? Are we suddently experiencing a shortage of prepackaged binary distros or easy-to-build source-based distros? Personally, if it had only KDE I wouldn't touch it, except if it were Knoppix. And Knoppix is already Knoppix, so I don't need UserLinux to be Knoppix for me. I use one K-program in my daily life, KMail, and even that doesn't really excite me-- I'd switch but the momentum for doing so is just not quite there. So I think your assertion that KDE beats GNOME hands down is not a "fact", rather an "opinion".
For my part, I think GNOME beats KDE by a long shot because it is based on C and not C++. The number of competetent C developers far outweighs the number of competent C++ developers. C is certainly more common on the rest of a Linux system. Finally, I am given to understand that in terms on internationalization support and various accessibility functions, GNOME is ahead of KDE (or was momentarily). I think there are a lot of factors that you may not have had to consider when choosing a GUI that someone attempting to make an "enterprise" distro would think of.
...until somebody makes a choice you disagree with, apparently.
We're all aware that the whole point of this "free software" exercise is that people are free to do whatever they want with it as long as they share, right? Even if other people think it's a bad decision?
Mike Hoye
The argument is this :
... and ...
SCO's bosses at Canopy controls Trolltech which controls Qt which controls KDE.
(do an nslookup of www.trolltech.com and www.kde.org to verify that last bit of logic)
The second contention that's a touchy subject is "Canopy controls Trolltech". Somebody is going to post a link to the trolltech site that says "only 8% of Trolltech is controlled by SCO/Canopy".
Then what the hell is Ralph Yarro (Darl Macbride's boss) doing on the board directors of Trolltech?
Link here for the skeptical.
The issues is real simple. If Canopy doesn't control Trolltech and Trolltech support Linux, then why haven't they
1) Come clean on exactly what their relationship is with Canopy
2) Voted Ralph Yarro of the board.
Trolltech should come clean. What is their relationship with Canopy? Does canopy have contractual rights to sit on the board? Do they owe debt to Canopy? Does Canopy have warrants on Trolltech? The silence is deafing. Speak Trolltech, tell us the truth.
The sad thing is QT is a good product. They could increase their respect and marketshare by telling the Canopy chumps to take a hike.
Its just like you can simply download and install mozilla/firebird/opera/etc on any windows machine - internet explorer would only be the default. You dont have to use it.
UserLinux is targeted at the enterprise community. It probably (and hopefuly) be used by non-geeky types who simply use what is installed on their computers. You can see KDE go the same way as netscape if UserLinux ever becomes popular.
I am not sure what the right approach would be though. It makes sense for an OS to have a consistent "face" to a non-techie user. Though choice is great (and it is the reason why I use linux), it can be a hinderance for someone who is not inclined to try out 5 different window managers before she decides which one to use!!
Don't Panic
Bruce writes:
It is possible for us to make our system entirely royalty-free for solution developers, both Free and proprietary. This dictates some software choices: GNOME and PostgreSQL rather than KDE and MySQL, simply because of the way those products license proprietary developers. This will support a large ecosystem of both Free and proprietary solution developers by lowering the financial barriers to entry all the way to zero.
So, let me get this straight - he wants to discourage the use of GPL'd code in UserLinux in order to have businesses create proprietary applications that can not in turn be included into UserLinux because they will not be free?
Sounds like an interesting one-sided ecosystem.
If he took this commercial-friendly argument to its logical conclusion he would dump the GPL'd Linux from UserLinux in favour of BSD. But then it would not be much of a UserLinux, would it?
Funny that the KDE developers' argument boils down to: "We have these proposed applications, most of them aren't finished yet, or haven't even been started, but they're going to be really great and integrate well with an enterprise desktop, so you should choose KDE and ditch Gnome."
Thing is, Gnome has the applications NOW, already integrated and complying to a well tested Human Interface Guide.
Further more, nobody ever said UserLinux wouldn't include KDE apps. There is no reason their "killer apps", should they ever reach a usable state, couldn't be included. The decision is to go with Gnome as the default desktop, and these KDE developers have shown no reasons why this isn't the best choice. There is no need to include both desktops, as a user can only be running one at a time anyway.
Finally, if you don't like the decision, don't use the distribution! There are plenty of other distributions out there, at least two of which still use KDE as the default desktop (Mandrake and Suse). Funny that you don't see Gnome developers demanding that these distros switch to Gnome.
Let the whining and flaming commence!
All that we see or seem is but a dream within a dream.
His arguments are well thought out, and I must admit that the lisensing issue had bitten me at least once with respect to Qt. I also realize what this will cost me personally to retool to Gnome, but in the end will likely be worth it. I would have made the same choice a couple of years ago if I had known that Gnome was not also based on Qt (I had already tooled up with KDE by the time).
Oh well... time to learn yet another GUI API...
As someone who have read almost all the mails in the list so far, I think the decision to go with GNOME was taken in a BAD way atleast. First there was a lot of discussion about the technical merits of these 2 and later it was decided to go with GNOME based almost solely on "proprietory friendly" license of GTK. If that was the only criteria for selecting the softwares, then what is the purpose of discussion? its easy to compile a list softwares with the above license. Also, as far as I understand(someone correct me if I am wrong in this please) QT has been adopted by the ISVs far more than the GTK. It boasts of a customer base including IBM, Adobe , Samsung and many many other high profile companies. So if all these companies have found the technical mertis of QT(better documented APIs, more powerfull, architectural superiority etc.) more than the "proprietory unfriendly"[sic] nature of the license, then why the same logic does not apply to Userlinux? . The same goes with Epiphany also. It seems to be (not finalized though) decided that Epiphany would be the default browser. There was a lot of discussions about browser choice also. Mostly it was agreed that Mozilla firebird was the best choice . But then at the end Epiphany was chosen(?) based solely on the "better GNOME integration".Again why there needs to be a lot of discussion about the technical merits if at the end some criteria like "proprietory friendly" nature of the license and "Gnome integration" are going to be the sole criteria. Is these 2 factors so important to give such a huge weightage over technical merits and everything else ? After carefully reading all the mails also, I have not found a good answer for these concerns. Perhaps I am missing something. But anyway I wish all the best for Bruce and his effort.
http://www.nasirudheen.blogspot/
It has a vague promise, but I would like to *see* KDE running on GTK+. The choice has to be made now, and there is no GTK-KDE *now*.
You don't understand. The KDE proposal sugests that UL include both Qt and GTK libraries, it says nothing about trying to "port" KDE to GTK, which is a completely ridiculous prospect.
If I was to start developing a closed source application (I could, if I lost my day job or whatever) Qt would not be an option for me. Only the big/medium size ones with steady income can afford it.
Oh, who cares? If you want to develop and distribute closed source apps, why don't you do it on Windows? In case you didn't notice, the Linux culture values freedom. [Note that this does not include most "enterprise" software development, which is typically for internal use only. In this case, they can simply use the Free Edition of Qt. As long as they never distribute the code, they don't have to abide by the GPL.]
Liberal (adj.): Free from bigotry; open to progress; tolerant of others.
I can understand the business decision to go with one DE over the others... this makes good sense. What I *don't* understand is why Perens refuses to include the KDE libs by default. Including the libs and a few choice applications to fill the gaps where kde apps outperform what gnome offers (k3b and others) would do a lot for improving the total package of UserLinux and keep some of the critics at bay.
If I can put on my tinfoil hat for a second here, the best reason I can think of for not including the KDE libs is to stir up the traditional KDE/GNOME debate and get more coverage on slashdot and other sites. Trolling for media coverage, it's the wave of the future!
501 Not Implemented
I am not denying Qt licensing issues. They are real. And since KDE is based on Qt, it goes against them.
The point I wanted to make was why the KDE team was so concerned about not being included in UserLinux. If Userlinux is ever to become as popular as they hope it would (and it does not include KDE), it would be a big loss for KDE.
Don't Panic
Lord, this arcane bickering seems pointless. From my reading Perens' goal is to develop a distro that "just works" for the average corporate user.
I just wrapped up my semi-annual "download a linux distro and see if the damned thing will work" exercise, and once again I'm falling back on Win2K, which does everthing that I need, and does it with a minimum of BS.
Over the last few years I've tried RedHat, Mandrake, and a half dozen other distros, on both bare machines and dual boot systems. In every case I hit the wall when trying to implement some simple or essential feature. Every time I found myself led into that arcane and recursive hell known as man pages and how-tos. (and mailing lists, and discussion groups and...)
I'm good with hardware, and can make Windows do anything I need. I have managed to troubleshoot some exceedingly obscure problems in the past. Still, once again I've abandoned Linux because I can't afford to invest weeks of obscure research just to do day to day work.
I really want to be rid of Microsoft products. I find them more irritating than useful, and surely don't like MS' business practices.
I love OpenOffice and have pretty much abandoned MS Office. I like Mozilla, and use a wide variety of freeware and open source products.
I would in a minute abandon MS Windows if it were practical, but to do that I need a distro that will do what Windows does:
Find and configure all my hardware, set up internet access and networking to allow all of the computers on our system to share files, and easily allow others to use the printers attached to my PC, easily set up my two video cards and monitors, set up to sync my PalmPilot.
And have decent looking fonts.
So far every distro I have tried has blown at least two of these basic goals and has offered no easy way to achieve them.
Again, I cannot afford to spend days or weeks tracking down the obscure solution to make something like HotSync work.
It does not matter to me whether this happens with GNOME or KDE. If I can boot from CD and have all of these things come up working, I'll buy it.
If including only GNOME allows Perens' the time to make a truly reliable installer, then I'm for it.
Three Squirrels
You make a good point about GNUStep - by selecting it, you open up the work of Mac software as well. Some not-so-drastic changes to the code and a recompile later, you've got an x86 version of PPC application.
Perhaps GNUSteppers should start THEIR own distro.
Heck, with the licensing options, all the major desktops should have a distro. Let the market decide!
I'm not kidding.
To celebrate the occasion of my 1000th post, I will post no more forever on Slashdot. Goodbye.
Nice story, but it's bullshit. Gnumeric couldn't have borrowed code because there was no one to borrow from. Neither KSpread nor OpenOffice Calc were around: Gnumeric was the first of the modern Linux spreadsheets. The only options around at that time Oleo and Siag. You don't get an app of Gnumeric's quality (it's probably the best OSS spreadsheet around, including OO.o) by copying all your code from Siag and Oleo.
It's hard to be religious when certain people are never incinerated by bolts of lightning.
One argument that I keep hearing from the KDE/Qt camp is `If you are going to make commercial applications, you can afford to pay the price of a license'. There is even a study that shows that for a team of 10 developers at 75,000/USD year for two years, the price of Qt is peanuts compared to the total investment.
The flaw I see with this analysis is that it is not in par with the rest of Linux. You do not see sysadmins paying for Apache, Perl, MySQL, Postgress, Python, Slashcode, PostNuke/PhpNuke or any content management system, and I am sure anyone running a porn site, or using Linux in an eneterprise settings is probably mkaing more than the 75,000 per developer listed above.
Bruce's position is consistent with the fact that there are no strings attached, no special royalties, no special treatment of users: everything is free, and you can build entirely fre e systems with no strings attached with GNOME as the platform.
Also, 2,600 dollars per license might not be much in Britain or the US, but it is sure a hell of a lot of money for a startup anywhere else where Linux is being deployed: Malaysia, China, India, Brazil, Peru, Colombia, Chile, Argentina.
Hey, 2,600 dollars will get you a Windows Server edition, Visual Studio and quite a few extra goodies from MSDN, am just not sure that telling commercial developers to ponny up more than they have to for MSDN is a winning position here to promote Linux over Windows on the desktop.
Hey on Microsoft you can even get the Platform SDK and the Microsoft.NET development tools for free and use SharpDevelop instead, and that my friend is a lot cheaper than Qt is.
I am not trying to be a smart ass, but I would really appreciate it if you would enlighten me on this issue. I may be victim to the slashdot-fiction as IANAL and I have not studied either license in detail.
My understanding from what I have read is that if you are going to make a commercial product with QT, you need to buy a royalty license. Is that not the case?
"My understanding of Bruce's goal is that UserLinux will be the one distribution that all comercial software and hardware makers will chose as a point of reference."
If hardware makers support it, then, because it is just re-packaged Debian, that support will be available for any Distribution out there.
If software makers support it, then, because it is just re-packaged Debian, the KDE people can add it to KDE.
"I fear a future in which all comercial developers will only support UserLinux."
Why? It's just re-packaged Debian. What are you afraid of?
"I believe this is what the KDE developers fear also."
Afraid of what, SPECIFICALLY?
The GPL is incompatible with several open source licenses including Apache, Helix, and Eclipse. By relying on GPLed *libraries*, you're saying that the only open source worth writing is GPL-compatible open source.
BTW, at least half the companies you mention also produce products based on Gtk+ (WebSphere from IBM and Realplayer are probably the most prominent of the list). Every company I've been to produces lists like these. If one person in a company buys your product, you can automatically say "This company uses our product". They may just be buying it to test it or they may discover that it doesn't suit their needs. It doesn't matter, as long as they bought it, marketing can get away with saying that they are a customer.
And let's not forget the small company. Thousands of dollars is *a lot* of money. Money is tight, and I don't see anything particularly amazing in Qt that would cause it to pay for itself. Small companies don't often care about portability (so they stick with one platform), but even when they do, Qt isn't the best choice. WxWindows or Swing or SWT or pyGtk+ or PhpGtk+ are far cheaper than commercial Qt (their free) and they're at least as easy to use.
Apple and Microsoft charge a license fee for their runtime/desktop, so you aren't comparing apples to apples here.
First you presuppose the KDE developers didn't consider this issue, which is a big assumption. Second, you present a scenario where there never would have been the quality KDE we have today. I say the KDE/Gnome competition have been a good thing for both desktop environments.
SCO's bosses at Canopy controls Trolltech
Not according to Trolltech's investor list, which claims that the employees own nearly two-thirds of the stock. Even Borland owns more than Canopy and SCO, which put together control less than 6 percent of Trolltech.
Of greater moment is the choice of database library. PostgreSQL is enormously more complete and standard-compliant than MySQL, and (for years, now) faster, and is committed to the more liberal license that Bruce has demanded. MySQL (like KDE) is straight GPL in release 4.x, and lacks many important enterprise features.
Odd, isn't it, that we don't see flame wars over the database? Probably they will switch quietly to PostgreSQL once they get around to the matter, and nobody will make a fuss.
Because it doesn't make sense. It sounds like he's more interested in proprietary, commercial development than Open Source or Free Software, or even, (god forbid) choosing the best software for the job.
"I assumed blithely that there were no elves out there in the darkness"