On Tuesday, developers from both the Debian and KDE projects will announce an ambitious initiative to provide a Linux desktop operating system aimed at the needs of large organizations. By working together the group intends to deliver an integrated solution for the desktop needs of government agencies, educational institutions and enterprises.
Linux Group Aims for Enterprise Desktop
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Almost every Linux article these days lately is about Linux on the Desktop.
It is good enough on the server side, so it is time to expand its market. I don’t see anything wrong or weird with that. Besides, this effort is about the Enterprise, not home desktops.
Its just an “Amazing” to see how the open source community is pulling together. On a side note:
“Currently under consideration is a port of the excellent and pre-existing Ark Linux installer frontend to underlying Debian technology such as debootstrap and debian-installer.”
I just found this of interest. They are not using the typical debian installer or the anaconda port. Please dont get me wrong, I have heard some fantastic things about Ark Linux. I look forward to seeing Ark’s stable release.
And I would love to see some pics of Ark’s installer. Since Ark is based on RH, does anyone know what their installer looks like?
I wonder if this is KDE’s answer to Perens’ UserLinux.
Yet Another Dsktop Initiative…
Desktop Linux:
1. Novell integrates it’s old product into Linux and renames all the pieces and makes changes. They should pay attention to ‘one product line’ and pick up marketing tips from MS and also Sun by observing that languages succeed based on support and clear and publically visible roadmaps for developers or anyone making a long term investment. And somehow mix in the hype, excitement, and more hype.
2. Sun is mostly doing the right things however they have to be even more supportive of Linux. It is Linux that will make Sun popular and will drive sales. I think that Java3d should be implemented for Linux and be part of the JDS.
3. UserLinux is in the inception phase, and I’d like to see them right away join Novell and make a real product based on Linux. They need all the resources they can get in order to compete with Sun and MS.
What is an “Enterprise Desktop” and how at all is it supposed to be implemented in the Desktop? These features sound more low-level to me.
Okay I read the article and understand it more, but I honestly hate buzzwords like “Enterprise Desktop”. It’s just that a lot of these features I have been wanting on my own desktop for the longest time.
Roaming support, whatever it means, sounds useless. Most places that use this have DHCP or WiFi.
The page is filled with lots of vagueness and buzzy-language.
“We are a group of KDE and/or Debian developers with a dedication to and interest in both projects.”
Sounds to me like just another distro, with a lot of talk going on. I mean I wish the guys luck, but honestly they aren’t changing much. We’ve already had so many distro’s with the same promises, such as Gobo and Xandros.
IMO not much needs to change for “Enterprise Desktop” to see Linux and other Free OS’s on business desktops you have many other obstacles to jump, more than “user friendly” issues.
OSDL should include the Linux desktop and not only the kernel.
Ark installer is based on the QT/KDE toolkit. I beleive some of the reviews featured on distrowatch show pictures of the installer. (you can play tetris or frozzen bubble during the install…)
Although I fully support ark their installer is still work in progress. I couldn’t install it because it has trouble with “complex” partitioning. Anaconda is nice but I think Mandrake’s installer is perfect… execpt it is GTK based…
We’ve already had so many distro’s with the same promises, such as Gobo and Xandros.
I haven’t seen these distros give much back. This project will be great has it comes from within KDE itself. I beleive these guys can really leverage KDE to it’s full potential.
This effort will fail, simply because anyone that wants to develop commercial applications for this platform, have to pay in advance (per developer) to develop for it.
When there’s another option, still debian, and backed by a more known and vocal organization (userlinux), without this license issue, the choice is clear.
> This effort will fail, simply because anyone that
> wants to develop commercial applications for this
> platform, have to pay in advance (per developer)
> to develop for it.
The actual price of the toolkit is nothing compared to the cost of a developer. That’s why every manager would decide to pay for the most efficient toolkit if it allows to reduce development time and if it increases the quality. That’s why so many companies are using Trolltech’s Qt instead of Gtk.
Most parts of the world simply can’t afford to pay Trolltech’s price for QT. That’s why GTK is a much better option.
Zen: “What is an “Enterprise Desktop” and how at all is it supposed to be implemented in the Desktop?”
An enterprise desktop is deployed with company interests and policies in view. E.g., you may not be able to install new apps or plugins, the user interface can be less polished and more “industrial” because training is available, usage could be “supervised” while not of personal nature, etc.
For an analogy, think about cars a company buys for field operations: they’re usually more utilitary, simpler and sturdier versions of conventional end-user cars.
> Most parts of the world simply can’t afford to pay
> Trolltech’s price for QT. That’s why GTK is a much
> better option.
If you GPL your application (and thus giving back to the open source community) you don’t have to pay for a Qt license, even if you sell it commercially.
Also, TrollTech has special arrangements for startups that cannot afford the Qt license right away. If you actually start make a profit of your app you can then easily pay back the Qt license, especially considering the much lower development time compared to Gtk and other toolkits.
“KDE Core Enterprise Enhancements:
Restrict desktop, application, and printing actions.
Restrict internet access on a URL basis at a desktop-wide level.
Restrict desktop resource customizations.”
Reduce the functionality of Linux and OpenOffice.org and Mozilla. Yeah, THAT will help.
I know I wouldn’t want to have to use it. I use Linux at work for things like hosting shell and perl scripts, snmp tools, system administration, and the wide veriety of network tools available (host, dig, nmap, nc. etc) for troubleshooting problems. Almost every other person I have ever met using a Linux or UNIX workstation is using it for all the same functions.
Bear in mind that most of these people are using Windows as a primary workstation. I simply would not want to have to be a network administrator at any place using Linux workstations. My personal Linux boxes waste enough of my time and I am su’d as root 20% of the time on my boxes. The though of having to do this for 100 people just to get around using a microsoft product is laughable.
Besides, everything good about Linux can be found at a bash prompt. You can pry my windows workstation from my cold dead fingers. I would rather buy a copy of windows for the company I am working for than get stuck using Linux as a primary workstation.
should read “and add* OpenOffice.org and Mozilla”
>Also, TrollTech has special arrangements for startups that >cannot afford the Qt license right away. If you actually >start make a profit of your app you can then easily pay back >the Qt license, especially considering the much lower >development time compared to Gtk and other toolkits.
You’re still going to have to pay it back to Trolltech eventually. How much does a programmer in India or Eastern Europe, etc earn? QT’s pricey even for small companies in the west.
As for efficiency, GTK supports a more modern version of C++ than QT’s – as well as C#, C of course, and numerous others.
Most of the best free software apps are written using GTK.
Why should the programmer have to humiliate himself by going begging to Trolltech when there are a number of excellent free toolkits available?
This is wonderfull. Finally, someone will ensure that the kde desktop will be pushed to its potencial. Congratulations to all, I hope to contribute.
> Why should the programmer have to humiliate himself by
> going begging to Trolltech when there are a number of
> excellent free toolkits available?
For developers such as shareware writers that don’t want to invest money GTK will also be a supported option.
> Most of the best free software apps are written using GTK.
wich ones would that be?
Hi
Those who argue that QT license wouldnt satisfy the needs of enterprise developers is not aware of reality. There are whole lot of people who are paying trolltech because other lgpled or bsd licensed tools dont satisfy their need.
Just list the gtk based proprietary tools and compare them with Qt based proprietary software. If you are not willing to do the reality check just search the userlinux archives.
Regards
Rahul Sundaram
> You’re still going to have to pay it back to Trolltech
> eventually. How much does a programmer in India or
> Eastern Europe, etc earn? QT’s pricey even for small
> companies in the west.
Do you realize that companies in those countries are charging rates for Qt development comparable with what companies in the US are charging, perhaps only slightly less? That they already therefore own the Qt licenses you claim they cannot afford? Probably not, because you have no clue about what you are speaking.
> As for efficiency, GTK supports a more modern version of
> C++ than QT’s – as well as C#, C of course, and numerous > others.
That not only makes absolutely no sense, but it’s plain wrong too. Qt can be used from all of those languages, and well, what you mean by “a more modern version of C++” shows that you have no clue what you are talking about to begin with.
> Most of the best free software apps are written using
> GTK.
Oh yes, all of those ones. Of course.
> Why should the programmer have to humiliate himself by
> going begging to Trolltech when there are a number of
> excellent free toolkits available?
Yes it’s very humiliating to have to pay for a great software development kit to use to make commercial, proprietary software when you can pay Microsoft, or better yet, you can use a sub-standard free package that’s out there and not contribute back to the community. So humiliating to pay for the corporate support that is generally demanded and paid for by these ISVs anyway. So humiliating to support a company that produces a GPLed toolkit and puts their revenues back into development of that product. Oh my, what were we thinking.
>>Most of the best free software apps are written using GTK.
>wich ones would that be?
Gimp, Evolution, Gaim, Gnumeric, Gnucash, Gnomemeeting, Pan, Galeon/Epiphany, Sodipodi/Inkscape, MrProject – just to name a few
Restrict desktop, application, and printing actions.
Restrict internet access on a URL basis at a desktop-wide level.
Restrict desktop resource customizations.”
Reduce the functionality of Linux and OpenOffice.org and Mozilla. Yeah, THAT will help.
The point isn’t to restrict it for everyone, it’s that the software will be in place so that administrators can restrict it – this is incredibly useful in a corporate situation, I’m a sysadmin, and at this point wouldn’t consider moving my workstations to Linux, because there is no way to lock them down so that users can’t mess up their settings, leaving me to go and fix them.
Yet Another Dsktop Initiative…
I hate to say it, but this is true. I think a Debian/KDE soution would be awesome, but doesn’t anyone else agree that 26 simultaneous desktop projects might be SLIGHTLY counterproductive? Cooperation got Linux this far, I don’t know why so many are abandoning it now.
Could the level of discourse in the OSNews forums go any lower? I don’t even agree with this proposal, but here I am defending it, because all the evidence above seems to indicate that I’m the only one who read the thing.
Responding to all sorts of comments from various people:
>I wonder if this is KDE’s answer to Perens’ UserLinux.
>Yet Another Dsktop Initiative…
>Cooperation got Linux this far, I don’t know why so many
>are abandoning it now.
>When there’s another option, still debian, and backed by a
>more known and vocal organization (userlinux), without this
>license issue, the choice is clear.
There’s no need to wonder if you read the article.
It’s spelled out pretty clearly what this is.
This is a proposal by KDE/Debian people to use a KDE-centric solution in UserLinux. It is neither “yet another desktop initiative” (it’s simply a proposal trying to influence an existing project) nor is it an abandonment of cooperation.
>Reduce the functionality of Linux and OpenOffice.org and >Mozilla. Yeah, THAT will help. I know I wouldn’t want to >have to use it.
The proposal was talking about lock-down ability for enterprise desktops, and anyone who has any experience in that area will tell you this is an important and necessary capability.
>This effort will fail, simply because anyone that wants to
>develop commercial applications for this platform, have to
>pay in advance (per developer) to develop for it.
The proposal specifically talks about providing excellent support for GTK. The multi-toolkit desktop is already the norm and the reality.
>What is an “Enterprise Desktop” and how at all is it
>supposed to be implemented in the Desktop? These features
>sound more low-level to me.
You’re right. From the proposal: “We will enhance KDE to a level that, while not currently practical with multiple targets and the endless variations of the Linux/Unix platform, are made feasible by a focus on leveraging core Debian technology.”
So their developing a workstation, right? Thats what it seems like. Thats not an easy task, workstations have a lot of specialised software that needs replicating on the linux side or atleast support for it. One of the bigest things it’ll need to have is Microsoft Office support. OpenOffice does this a little bit, but with version 1.0 it doesnt save or open office documents correctly, at least the more complex ones. Hopfuly that’ll change with MS XML format being used more?
things like SMB are already done on linux and comes standard with some GUI stuff. Front ends for cvs exist and it’d be good to see that in there too. Big job ahead of them i would think. I think desktops on linux are usable, but as a workstation im not so sure.
“You’re still going to have to pay it back to Trolltech eventually. How much does a programmer in India or Eastern Europe, etc earn? QT’s pricey even for small companies in the west. ”
Are you new here? If you don’t GPL it, yeah, put up the bucks.
The actual price of the toolkit is nothing compared to the cost of a developer. That’s why every manager would decide to pay for the most efficient toolkit if it allows to reduce development time and if it increases the quality. That’s why so many companies are using Trolltech’s Qt instead of Gtk.
I’m sure I’m saying this simply because I know them better, but I prefer the GTK and wxWindows libraries to Qt. I think they are both easy to use, look a lot better, and are free.
>One of the bigest things it’ll need to have is Microsoft
>Office support. OpenOffice does this a little bit, but with
>version 1.0 it doesnt save or open office documents
>correctly, at least the more complex ones.
OpenOffice is at 1.1. Word compatibility is excellent at this point, though of course not perfect. I use it with pretty complex documents all the time and haven’t had a problem in ages.
If you really need MS Office, you can just run MS Office. This is one of the things Wine (or even better, Crossover Office) does extremely well. Office 2000 runs nearly flawlessly. Photoshop now works also.
UserLinux has to try to build an ‘Enterprise Desktop’ because all of their funding comes from enterprise interest groups. I think that an open source knowledge base is needed more than anything, and that it would drive commercial interest in Linux which would in turn keep prices competitive for Enterprises.
>I know I wouldn’t want to have to use it. I use Linux at
> work for things like hosting shell and perl scripts,
> snmp tools, system administration, and the wide veriety
> of network tools available (host, dig, nmap, nc. etc)
> for troubleshooting problems. Almost every other person
> I have ever met using a Linux or UNIX workstation is
> using it for all the same functions.
this isn’t about limiting the functionality for everyone, it’s about allowing people to choose subsets of this functionality as they see fit. this is primarily for corporate desktop deployments. the infrastructure is already there:
http://www.kde.org/areas/sysadmin/
the tools to make this usable when managing 100s or 1000s of desktops and without requiring intimate knowledge of KDE is the part that is being addressed.
> Bear in mind that most of these people are using Windows
> as a primary workstation. I simply would not want to
> have to be a network administrator at any place using
> Linux workstations. My personal Linux boxes waste enough
> of my time and I am su’d as root 20% of the time on my
> boxes. The though of having to do this for 100 people
> just to get around using a microsoft product is
> laughable.
this is exactly what the new set of tools currently in development help you avoid. some of us understand the issues intimately and are working on solutions. this time next year, this issue will be largely erased.
there are also other aspects of making KDE enterprise deployable that don’t include configuration management that are being worked on as well.. and it’s all open source. go ask Novell for their Free Software version of NDS and Red Carpet, or Sun for their Free Software deployment tools. 😉 exciting times.
YES you can program in GTK while running KDE, and YES you can program in QT while running Gnome.
Is their a point to this stupid argument.
OSS innovates in the shadow of others yet again.
Where were these initiatives before Sun splattered the world with Mad Hatter and JDS press releases?
What has changed in these organizations that they couldn’t have done this before?
IBM isn’t interested in the desktop, RedHat isn’t interested. Sun comes out and shows that there is a market for this, and provides a solution to fit the bill (however imperfect), and then starts setting up large clients for this endeavor.
Of course, the one thing Sun also did was wipe KDE from it’s realm of existence. Install JDS on a zillion desktops, and there’s a zillion desktops that, at the moment, can’t run any KDE apps.
Add to that the fact that Novell bought Ximian, (whose apps WILL run on a JDS), and that’s two big blows to KDE. All of a sudden there’s a lot of commercial support for Gnome compard to KDE. Sun didn’t wishy washy the desktop issue.
They simply picked the one they liked and stuck to it. This makes it easier to support, easier to package, easier to test, easier to develop for.
There is no reason why Debian/KDE could not have done this kind of “initiative” last year. But, nope, they didn’t.
Seems to me it’s the gasp of a hammered KDE rather than any impassioned effort to get on the Corporate Desktop. A “Me Too”.
Because, if they were serious, if they really wanted to support “Enterprise Desktop”, they too would have punted on KDE and just leveraged Gnome, helped ACCELERATE the process rather than reinventing the wheel One More Time and holding the effort back with wasted manpower.
It was only a matter of time before someone Picked a desktop environment and shunned the other. For whatever reason, Gnome seems to have won that race, especially with the pressure of Sun and Novell behind it.
But, as the community is wont to do, it fragments Yet Again.
Be assured they will or better already do.
Why accelerate something that you consider less advanced? And about fragmenting the community, you are wrong. The correct front unify the desktops is the _standards_ front. All toolkits should be able to live happly together.
Want to help? Join freedesktop.org, as meny people do.
wasted manpower is not leverage the millions man hours and the wonderfull apps that are part of the QT/KDE framework. Cooperation is the name of the game, that is what they are doing.
So why are you complaining again?
> But, as the community is wont to do, it fragments Yet
> Again.
the article actually talks about enhancing the _combined_ presentation of KDE, OOo, GNOME, etc as an important part of things. personally, i find exclusionary myopic worldviews to be rather unproductive, as they end up causing wheel reinvention and the presentation of a weaker product by not taking advantage of all the strengths available. it’s really too bad that some projects, such as Sun’s JDS and Bruce Peren’s User Linux, are marching down such a path.
> OSS innovates in the shadow of others yet again.
as opposed to all the innovation OSS has done on its own. you can’t always be first, but OSS isn’t always second either. all the large tech vendors have at some point innovated in the shadows of others, from Sun to IBM.
> Where were these initiatives before Sun splattered the
> world with Mad Hatter and JDS press releases?
be content in knowing that Sun’s JDS press-o-rama had nothing to do with this paper being written.
> What has changed in these organizations that they
> couldn’t have done this before?
a lot of things. for one, we’ve only recently come to the point of focussing on enterprise management tools. this is because it’s taken a few years to get the foundation technologies in place (from kio to kparts, kdelibs to DCOP), the infrastructure pieces (e.g. KDE’s sysadmin “Kiosk” stuff), the various desktop applications written that one needs to make a desktop useful, etc, etc… another thing that has changed is the level of interest and partners KDE has in this area; this has grown quite a bit in the last couple years.
in other words: you need the cart first before you can hitch the horse to it.
> Because, if they were serious, if they really wanted to
> support “Enterprise Desktop”, they too would have punted
> on KDE and just leveraged Gnome, helped ACCELERATE the
> process rather than reinventing the wheel One More Time
> and holding the effort back with wasted manpower.
KDE has a lot of technology in it that GNOME does not that can be critical when looking to roll out desktops into the enterprise, or even moderately sized installations. there are other advantages besides purely technological that KDE has, as well.
you know, sometimes accelerating the process means not helping reimplement the world somewhere else just because a few others have decided to do the same. GNOME was, after all, started after KDE and still lacks anything resembling KXMLUI, kiosk, kprinter, DCOP, a good SDK and more…
> reason, Gnome seems to have won that race, especially
> with the pressure of Sun and Novell behind it.
since when was Sun a desktop force? right, never. since when was Novell? and you happily neglect mention of SUSE with regard to Novell. =/
i’d also remind you that desktop Linux is only 1-2% of the desktop market right now; that 98% open field is a lot of playing space. there’s room for more than just KDE or just GNOME out there, and with such a tiny slice of the current market there is no de facto standard, let alone a winner to a race that hasn’t even begun.
QT is cheap for a company, you do not need to pay if you need to learn or share software.
All the guys saying that GTK is better should explain why to the people who is using QT.
Well, it looks like a nerve was struck there
>Do you realize that companies in those countries are >charging rates for Qt development comparable with what >companies in the US are charging, perhaps only slightly >less? That they already therefore own the Qt licenses you >claim they cannot afford? Probably not, because you have no >clue about what you are speaking.
I’m not talking about a few big companies doing offshore work for Americans. You’re surely not claiming that sales of QT there are more than a fraction of that in the west. I remember a thread in the kde lists where the price of QT was discussed. The KDE developers saw no problem with it, but a programmer from eastern Europe wrote to say that they could not afford it there. One of the KDE guys suggested he contact Trolltech to try and negotiate a reduced/deferred price, to which, understandably, there was no response.
>> As for efficiency, GTK supports a more modern version of
>> C++ than QT’s – as well as C#, C of course, and numerous >>others.
>That not only makes absolutely no sense, but it’s plain
>wrong too. Qt can be used from all of those languages, and
>well, what you mean by “a more modern version of C++” shows
>that you have no clue what you are talking about to begin >with.
This is from gtkmm’s site: “Qt was written when C++ was not standardised or well supported by compilers. Its design today is still based on choices available at that time, so it does not play well with more up-to-date code.”
I was saying that efficiency has more to do with the language selected eg the more modern C++ in gtkmm, c# (GTK# is more mature than QT#) The programmer also has a vastly greater range of languages to select from, for use with GTK.
>So humiliating to support a company that produces a GPLed >toolkit and puts their revenues back into development of >that product. Oh my, what were we thinking.
You work for Trolltech, I assume.
If you’d read my message carefully, you’d have seen that I was refering to programmers who could not afford a QT licence. No need to approach Trolltech for charity when GTK and other toolkits are LGPLed.
The case for QT, today, seems much diminished, given the rise of gtkmm, Java, and C#.
or are the kde-folks really frightened to be left behind (sun, novell, and likely userlinux all betting on gnome)-this attempt has definitly something of a panic reaction!
seems a bit like the last big effort to me not to loose the game (vs.gnome).
of course with that view, there’s nothing said about kde being somehow inferior-just that the tables seems to have turned against it.
Too many enterprise initiatives in OSS – how about something nice and easy for Mom and Pop? Like a really polished toolkit for OEMs to sell basic OSS based systems for point of sale, inventory, and keeping books? Keep it nice and open folks!
> sun, novell, and likely userlinux
There are a lot of rumors that Sun is the secret & dubious sponsor behind UserLinux.
Novell bets on KDE as well as Gnome according to their Vice Chairman Chris Novell
http://consultingtimes.com/chris_stone.html
The rumor that Gnome would be preferred was just spread by Nat to abuse his own position at Novell for shameless Gnome marketing. Guess why Nat has been so silent recently — he got his a.. kicked for doing that.
So there’s really nothing to be worried about.
I think it has become clear to anyone who has been on the userlinux mailinglist recently that Bruce lies and tries to enforce his agenda by dictatorial means
** Bruce Perens doesn’t reveal the names of his sponsors which seem to be clouded in secrecy. Who would ever work on a project where you don’t know who the instructing party is?
** Bruce Perens doesn’t reveal the requirements specification which obviously does exist and which ties the 1 Mio Dollar payment to using Gnome instead of KDE for example. If you browse the archives you can clearly see that the decisions which were made in the whitepaper are mostly not related to the discussion which happened on the list.
** Bruce Perens claims that the 1 Mio budget is is tied to using Gtk – while on the other hand he claims that he has no financial interest in Gtk (his claim of having a financial interest concerning the Qt book just sounds like hypocrisy to me!)
** On one hand Bruce insists that the reason for excluding the Qt toolkit (and therefore KDE) is justified by the decision that only *ONE* toolkit should be used in UserLinux, while on the other hand he has already approved *THREE* other toolkits for UserLinux!
** In spite of the fact that the majority of the vocal list has already clearly stated that KDE is much better in technical terms, much better suited as a development platform and much better in terms of cost (as the time to develop a Qt based product is lower and support/quality is much higher) he spreads a deliberate misrepresentation.
Obviously someone has a financial interest to drive his agenda and tries to muzzle others under the “banner of freedom”.
>or are the kde-folks really frightened to be left behind >(sun, novell, and likely userlinux all betting on gnome)-this >attempt has definitly something of a panic reaction!
I think you’re right. There are a great many KDE developers on the userlinux mailing list all pushing for KDE on Userlinux. Given all the recent announcements of huge Gnome deployments, and the Suse buyout, they must be getting a little desperate.
Bruce, however, has to choose solely on the basis of what’s best for Userlinux, rather than what’s best for KDE, as the KDE hackers naturally are.
The support costs for a second DE would be needlessly high. Gnome’s accessibility and LGPL libraries make the decision for it’s inclusion a no-brainer.
Read Bruce Perens rebuttal at http://www.userlinux.com and see if it explains _anything_. After saying that kde people should form a “subgroup” he is kicking them out of User(Gnome?)Linux. The kde proposal was very reasonable will-work-for-food type: many fetures and excellent support for gtk+. So the argument about the LGPL kit does not cut: if you have excellent support for a LGPL toolkit, it does not matter that your toolkit is GPL.
He did not accept them to try it out. He banned _all_ qt _apps_ from his distribution. I can even find a rationale for choosing a LGPL _desktop_, but banning the _apps_ is ridiculous! The GTK _apps_ are also GPL.
There only one explanation: the guy is nor sincere and he has a political anti-qt agenda.
Because the KDE proposal was just too good.
Millions man hours have been given to KDE apps, like Quanta Plus, Kdevelop, K3B, etc… Given the KDE framework it would be trivial to make them _integrate_ to any Gnome desktop. You can ever revert the order of OK Cancel buttons if you want, change the menu layout in xml, etc… So why accept 5 or 6 different development frameworks, like mozilla, OOo, GTK, Mono, mxwindows and others. They may look the same, but they are different frameworks. So what is the rationale to ban apps that can integrate very well to the desktop, when you already have a mixed framework, even when they are the best apps? Is there any app like Scribus in Gnome?
Gnomers, you can gloat. But this is sad, ilogical and unfair.
Bruce does not deserve to be called a leader. A leader unite people. There is a way to integrate everything: the standards path. There is _no_ technical reasons why all toolkits cannot live together happy as they do in many enviroments, and this is the freedesktop.org way.
He is going the SUN way.
A lot of KDE people are afraid. I know I am. If Linux becomes GNOME-centric, and KDE is no longer viable, I’m back in search of a desktop OS. I just plain don’t like GNOME. I think the UI is pretty polished, but that’s its only saving grace in my eyes. So if GNOME wins, than it will be yet another example of a worse technology beating out a better one because of business concerns rather than technical ones. I think there are many other KDE users who also feel this way. The KDE brass doesn’t seem worried about it yet, but the rank-and-file are getting a little jittery.
If you GPL your application (and thus giving back to the open source community) you don’t have to pay for a Qt license, even if you sell it commercially.
http://www.trolltech.com/products/qt/freelicense.html
Pay particular attention to:
The Free Edition licenses do not allow the development or distribution of commercial software.
I also seem to remember reading on Trolltech’s site (can’t find a link at the moment) that you can’t use the free version to develop and then buy a license later.
I once asked KDE about the licensing of their libs which are LGPL but are derived from QT which is GPL and never got an answer as to how they can do this. Surely if something is derived from GPL then KDE’s libs should be GPL?
i like the way “mozilla, OOo, GTK” were listed as different development frameworks, while this is the case, all of the above are using(or starting to) GTK on linux at the moment
as for licensing, KDE’s libs indeed should be GPL’d if they’re the one’s that just build on GPL’d Qt, if they’re something new and interesting, and not based on the same code as TrollTech’s they can be LGPL’d. however, if you write a commercial application using Qt, you have a license from trolltech, and want to port to KDElibs on linux, problem, Trolltech can’t license back a copy of KDE’s work to you as non-GPL, certainly not without asking permision from every hacker who’s ever worked on the KDE core at any rate. so commercial apps build on KDE are almost certainly in breach of the GPL.
as for KDE users getting jittery, and users needing to find a new OS because of a dislike for GNOME!? why, what’s to stop KDE developers moving on to GNOME, fork it, and build a KDE-alike version, but i personally doubt that KDE’s in as much trouble as you think, and maybe trolltech would LGPL Qt if things started looking really bad for KDE, i mean, no Desktop Environment using Qt means little incentive to use Qt on desktop linux at all!
the other place that Qt licensing annoys me is win32, where it’s not even remotely free, this has hurt a few projects who could have been very useful in crossplatform compatibility, oh, and does the GPL allow you to links your GPL project to non-free Qt libs on win32? i doubt it, so you can’t GPL’d programs and port them to win32 without a license change.
anyway, i am _NOT_ anti KDE, but i really don’t like it’s reliance on Qt.
You can write GPL software that uses the commercial library. Just that you must add an exception to the license that it may be linked with that library. Do you think that there is no GPL’d software on Win32 yet? 😉 Hint: the win32 libraries are proprietary even without Qt…
Regarding “Issues with GPL free QT” and the link to the Qt FAQ: I would say that choose GPL as the Qt license (as you know the “free” Qt on X11 is licensed under GPL, QPL you can pick one of them) and follow what GPL says, and not what Trolltech says in the FAQ. Guess what it says: you can develop commercial applications. You can even distribute them as long as you also distribute the source code. You can develop whatever you want, including inhouse software and if you don’t distribute, you don’t break the GPL license. If you do and you don’t distribute the source, you break it and you have to buy a commercial license. Read also the GPL FAQ. Of course I’m not a lawyer, but I would more trust the GPL FAQ in this matter as that was created by the same people who created the license itself.
> DE’s libs indeed should be GPL’d if they’re the one’s that just build on GPL’d Qt
As Andras already wrote, QPL allows you event BSD style licenses, so KDELibs are mostly LGPL, some things under artistic license IIRC.
> f you write a commercial application using Qt, you have a license from trolltech, and want to port to KDElibs on linux, problem
No problem, KDE libs are LGPL, Qt has a license for closed source applications (assuming this is what you meant with commercial)
> so commercial apps build on KDE are almost certainly in breach of the GPL.
Nope, see above.