There is a controversy in the Linux world. It doesn’t have to do with Microsoft, or anything overtly technical. It may seem, to the outsider, the open source equivalent of the question, “Boxers or briefs?” But it’s much more serious than that. More here.
Flipping the Linux Switch: the GNOME Desktop Environment
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSAlert.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Back in the day i was happy to use both Gnome and KDE. Over time though i have come to use the command line more, and by default KDE’s file manager includes a “open terminal here” keyboard shortcut, which i find incredibly useful.
Moving forward in time and Gnome’s file manager has no toolbar and insists on some crazy spatial concept. I am sure this can be changed but out of the box it just seems crippled. I have tried to use it for a week here or there but it’s a fact that i can do my work quicker in KDE.
So unless Gnome changes radically i for one will never be using it on a regular basis and cannot recommend it to anyone.
So, just because GNOME doesn’t suit your way of working with a computer, you can’t recommend it to anyone?
To counter the problems, whose solution would require Gnome to change radically:
All the Linux distributions I know do ship a plugin called nautilus-open-terminal. It lets me open a terminal into a directory I am currently looking at, by clicking “open in terminal” from a context menu. Currently it doesn’t have a very handy keyboard shortcut.
Gnome’s file manager doesn’t have to be used in spatial mode, and many distributions actually use the browser mode as default. In addition, at least I like to use the spatial mode. It isn’t crippled, just different.
I didn’t say it was no good to anyone, i just said i can’t recommend it.
I don’t mind things being different i just don’t see any benefits from it being different. If something is seemingly lacking to the point of being deficient in it’s functionality then perhaps crippled is an acceptable word to use.
I recently tried the new KDE 4 “start button” and though it was different it worked well.
I know lots of people use Gnome but i for one can’t fathom why
Although you better wait until KDE 4.1 because 4 is a framework more than anything,
A Start Button sounds more like Windows, Cassandra in Linux Mint 3.0 (which I’m using now and defaults to Gnome) and all the SUSE versions.
At least Mint provides multitasking through the Control Panels. But for full flexibility,
Try Arch or Slackware Linux if you have the patience and money to purchase any books on these distros.
Same goes with OpenBSD etc. which by the way supports GNOME to KDE, at least as far as I am aware of the free software license agreement of OpenBSD.
Although Richard Stallman will disagree with me. Just read last weeks spat between him and Theo de Raadt.
A bit off topic this is but you could possibly only support KDE through Linux Emulation on OpenBSD, which is another OS. Bit like GNUStep supporting Windows under KVM, or GNUStep being developed for Windows, that is de Raat’s point. As Stallman approves of GNUStep.
Did my English skills suddenly become that bad or is your English very bad, because I have not idea what you’re talking about.
What I mean is if you want to have any flexibility in your desktop environment but still keep it clean and neat, like you have said in your following posts, then install Arch or Slackware or your OpenSuSe environment.
Linux Mint has a SuSe/SLED like “Start Menu”. Novell thinks this will tie in with its deal with Microsoft to appeal to Windows users. So does KDE 4. But it is a development framework according to the developers, so if you are a user wait for 4.1.
But unlike these distributions, Mint does not close down the configuration tools window when every one is launched. You have experience this one in the GNOME environment surely. But it may work differently in KDE which you use as a better alternative.
Just a suggestion to the comments of the previous poster. Okay, no one’s English is perfect, even if you live in Australia, like me. Neither is my hyperlink skills, so you didn’t get my Archived last weeks (OS News) spat between Theo de Raadt and Richard Stallman. Both probably say the same thing about proprietary desktop environments.
Probably GNOME and KDE ought to learn a lesson from Amiga OS 4 or Syllable.
Edited 2007-12-27 07:42
I use Gnome because I can’t stand KDE. Gnome shows me only what needs to be there to get the job done. KDE shows me every option under the sun, and I hate that. Gnome is there managing my windows for me and KDE is constantly jumping up and down in my face reminding me that it is there.
Gnome looks clean and KDE looks cluttered. KDE 4 looks a little better, so I’ll give that a try at some point.
I don’t like KDE’s file manager either. Nautilus suits me better.
To each their own.
I agree. KDE now is just a mess of options and so many apps installed even in the smallest KDE install, it’s mind boggling and insane, so much clutter.
They need to modularize KDE like they do Gnome; you can have a very bare bones install and start from there.
And you can’t have that with KDE? Wow, that’s news to me. My KDE installation on openSUSE has almost no applications besides the basic stuff. Basically every app has its own package. Sometimes YaST bitched about dependencies when I removed certain KDE applications from the default openSUSE installation, but I had ignored them and I still have a stable desktop. If you have experienced that each KDE package installs a huge number of apps then it’s a packaging problem with your distro.
EDIT: *deleted*
Sorry for the doublepost. The first comment didn’t show up in OSAlert4.
Edited 2007-12-26 20:55
IMHO it’s the distributor’s responsibility to configure KDE in a sane way. I use KDE 3 on openSUSE 10.3 an I prefer it over GNOME, even though I’m mainly a Mac user and OSX & GNOME are supposedly designed in the same spirit.
Maybe I had different thoughts about KDE if I only knew KDE from another (unpolished) distribution.
Well, the default settings of KDE seem to be sane ones.
When I installed Debian 4.0, I had no gripes with KDE at all, it was clear and concise. What probably brings problems into KDE’s usability is the “gnomification” RedHat did in the days of RHEL3 (I don’t know if they still gnomify KDE nowadays).
So probably the Distributors should try to NOT configure KDE too far away from it’s rather sane default settings (because these dafault settings often already have quite some user feedback included).
And for KDE’s clutter: No matter in which mode Konqueror works, it never displays more than 10 Icons (3 more than Firefox 1.5.x), but it has 10 times more possibilities. I can live with 3 Icons more when it means that I no longer have to encounter a missing feature.
But the most severe lack I encounter when using GTK applications is the “file save as” dialogue. It just is not made for my way of orienting myself in the file system. Probably I am especially dependent on visual representations of locations, but i badly need a tree view. And KDE does just that, it shows me a tree view. GTK insists on not remembering that the last 100 times I had to “save as” something, I every time had to click on the “show folder browser” button (or whatever it is called in English) to orient myself.
I don’t know what the GTK usability guys where thinking when they decided to hide the tree view, but it seems they wanted to force people to build up the image of the tree view in their brains during the reading of the path line, instead of just saving the users this work by simply displaying a tree view.
If the only thing anyone can find fault with KDE is its looks, and you will find words such as cluttered or “not polished” being thrown around, then I would say that it is doing quite well.
KDE is simply a very mature set of technologies. Each network transparency is unmatched as is its speed or the richness, depth and variety of applications.
But hey, keep saying things such as it’s too cluttered if it helps you internalize the choices that you have already made.
Indeed, that happens a lot
Sure, KDE 3.2 was quite a mess, but that’s also quite some time ago… KDE 3.5 has seen much usability work, and while that work certainly isn’t done, KDE doesn’t fare worse than MS Windows. Combined with the familiarity many windows users seem to experience with KDE, I don’t see how anyone can say it’s not the best choice for ppl willing to switch OS.
And did I mention KDE is the future of the Free Desktop [1]
http://nowwhatthe.blogspot.com/2007/10/on-vision-and-future.html
The author of that blog is to optimistic for KDE, it doesn’t see the two main problems, the GPL license and the expensive cost for comercial developers.
All the hype and words KDE4 is having about beating to dead MS and MAC were present when KDE3 was released, the same soup about new technologies etc. And look at it now, It can’t even beat GNOME.
The solution is simple, be more independent from Trolltech. Period.
Imho, the GPL is a big advantage of Qt over GTK. I prefer a world with Free Software, and if proprietary companies want to make money making non-free software, let them pay. Their money helps trolltech to improve KDE, so at least FOSS benefits from some of the money.
And there is no excessive costs, Qt isn’t that expensive for a normal developer.
And how can’t KDE beat Gnome? Gnome still can barely hold its ground against KDE 3.5.x, 2 years old KDE technology…
And we’re as independent from TT as Gnome is from Novell, Sun and/or Red Hat. I actually believe for sure these three companies have a lot, if not too much to say about what happens in Gnome. Remember who started the big ‘cleanup’ of gnome, driving away many users?
There is the Free Qt Foundation, which protects KDE/Qt from abuse/bad stuff from TT, and even then – it’s GPL, we can fork it anytime. Meanwhile, TT is working better and more closely with us lately, so things are going the good way, not badly.
Imho, the GPL is a big advantage of Qt over GTK. I prefer a world with Free Software, and if proprietary companies want to make money making non-free software, let them pay.
The same extremists and discriminatories ideas that will hold back KDE. Mac, Windows, GNOME all offer for free cost the toolkit it is a shame KDE fails in something as basic as that.
And there is no excessive costs, Qt isn’t that expensive for a normal developer.
Maybe you are on crack but it is to expensive, but you are not even a developer anyway, what would you know?
And how can’t KDE beat Gnome? Gnome still can barely hold its ground against KDE 3.5.x, 2 years old KDE technology…
The user base said the contrary but kid your self if you want. KDE has the technology no one asked or needed in the first place, it has the charm to turn not needed features into bloat.
And we’re as independent from TT as Gnome is from Novell, Sun and/or Red Hat. I actually believe for sure these three companies have a lot, if not too much to say about what happens in Gnome. Remember who started the big ‘cleanup’ of gnome, driving away many users?
Another lie, KDE is so tied to Qt and TrollTech that is not even funny anymore, if TrollTech screams “Frog” KDE will jum at the moment.
There is the Free Qt Foundation, which protects KDE/Qt from abuse/bad stuff from TT, and even then – it’s GPL, we can fork it anytime. Meanwhile, TT is working better and more closely with us lately, so things are going the good way, not badly.
The free Qt foundation is just a fallacity, it is worthless.
KDE is and will be doomed because is TrollTechs pet project, simple as that.
Edited 2007-12-27 02:32
The fatal flaw in your argument, of course, is that Windows and OSX developers still choose to pay for the ability to develop in Qt. Far more than they do in linux.
The other fatal flaw is that developers pay a license already to develop in Windows or OSX, but it’s absorbed in the cost of the overall license through the magic of economies of scale. Since linux is “free”, it requires a different model.
The long drawn out “Qt isn’t free!!!” argument becomes really, really tiresome considering that Trolltech continues to grow quarter over quarter. People pay for Qt despite the fact that they can develop for “free” on Windows and OSX, and as they continue to, it erodes the whole argument. Tt are clearly doing something right, and maybe the armchair quarterbacks posting otherwise don’t really understand the way business works, or the concept of value and ROI. Here’s a hint: free price != cost savings.
The same extremists and discriminatories ideas that will hold back KDE. Mac, Windows, GNOME all offer for free cost the toolkit it is a shame KDE fails in something as basic as that.
Sadly, no one is going to use that zero-cost toolkit because it simply isn’t good enough when put up against anything you can use on Windows or Mac OS – and development tools is still a ten digit market for Windows and Mac OS.
That’s the kind of flawed, and pretty desperate, logic we’ve been dealing with for quite some time now to justify why certain things are not good enough in the free desktop world.
Visual Studio is not free and still most commercial development for Windows happens in Visual Studio.
And Novell does everything to turn GNOME into a puppy of Microsoft. Today the GNOME Foundation promotes the “excelent standard” (quote from Miguel de Icaza) Office XML, most new GNOME software is written in .NET, and so on.
Trolltech OTOH has almost no influence on KDE except Qt. If KDE was a pet of Trolltech, KDE had jumped to Qt4 much earlier. KDE3, after all, is using legacy technology in Trolltech’s eyes.
If you take a look at http://commit-digest.org/mappy/?view=2007-12-23&scheme=map_default&… then you’ll see that only 0-1% of all KDE Subversion commits come from Norway. Developers from the Netherlands commit more code to KDE.
The difference is that you have different toolkit for gnome. (I don’t agree with the former post though)
The force of mono is that they provide a productivity wise framework, a desesperate IDE for gnome and the result shows good application are delivered faster. Nothing happen for no reason.
Qt tools are fun to use and powerful, and it is probably the most efficient influence over the KDE project.
I thought Visual Studio Express was free. Visual Studio 2003 (the compiler) was free if I’m not mistaken.
How many *commercial* developers use the freeware editions of Visual Studio?
That’s classic, GNOME problems become KDE remedies. Trollish.
And look at it now, It can’t even beat GNOME.
Ha Ha… Now I’ve heard everything. KDE has about three times as many users as Gnome, as verified by every independent poll I’ve ever seen. This gap will widen considerably as KDE4 develops. Gnome is as good as dead.
Actually no, GNOME is winning and KDE developers know it, but hey, juts keep living your lie.
Dream on. You can find Gnome itself agrees it has approx a 30% marketshare… http://live.gnome.org/10×10
Do you have a link supporting this? Since there are so many of these independent polls as you say it shouldn’t be hard to find a good credible source.
Do you have a link supporting this? Since there are so many of these independent polls as you say it shouldn’t be hard to find a good credible source.
Over 70% of OpenSuse users are using KDE, so says OpenSuse’s own survey on this:
http://files.opensuse.org/opensuse/en/6/6c/Opensuse_survey_102_data…
This is despite every review I have ever seen of OpenSuse telling us that Gnome is the default, Gnome is Novell’s default desktop on all their distros, Gnome is where all their development input is going etc. For some reason all that seems to be passing their users by.
It’s just another case of people wanting to tell us one thing, and the quiet reality being something different.
Hmm. Funny enough because I found this
http://www.desktoplinux.com/news/NS8454912761.html
A more impartial source than openSuse might be better. OpenSuse users are usually old Suse users who have used KDE since whenever the distro started packaging it. Suse is still considered one of the better KDE distros around so it stands to reason that Suse would have a higher amount of KDE users since they may already gravitate towards that distro. using one distros survey to convey the whole market is flawed. Do you have another link that could support your claim?
If you laugh at GNOME thinking that it’s dead then you better use another OS, opensource is about freedom, not killing projects.
The author of that blog is to optimistic for KDE, it doesn’t see the two main problems, the GPL license and the expensive cost for comercial developers.
The GPL license isn’t a problem for KDE, just as it isn’t a problem for the Linux kernel. What the Linux kernel and the GPL has given us, certainly, is more open sourced drivers shipped with the kernel itself.
As for commercial developers, well, most have consistently demonstrated around here that they have no idea what commercial developers buy in their day-to-day work. Even if developers were suddenly to start using GTK tomorrow, they would still buy libraries and tools from proprietary tools companies – and GTK’s license allows that – paving the way for all kinds of extensions and lock-in. Either you have an open source desktop, or you don’t, and you need to keep its integrity.
It’s exceptionally sad that people have regurgitated the whole licensing ‘zero cost’ thing over the years as some sort of advantage, and it smacks of the days of Unix desktops, OS/2 and Windows when Unix desktop people thought they could tell everyone that their desktop was the standard and that would be good enough.
It’s really, really sad that people cannot say that Gnome is better than KDE because of a, b, c features and x, y, z APIs, because that’s all they have. Unfortunately for you, the arguments you’re coming up with are not what the real world wants to here, and they weren’t when Unix desktops were universally rejected in the early nineties because they were crap.
All the hype and words KDE4 is having about beating to dead MS and MAC were present when KDE3 was released, the same soup about new technologies etc. And look at it now, It can’t even beat GNOME.
You’re going to need an itemised list sweetheart, otherwise that is utterly meaningless and rather sad. You’re going to lose on features if you do.
KDE is the first free desktop that will be able to adequately compete with the likes of Vista and OS X in terms of visual effects and quality, as they are built into the toolkit, the window manager and the desktop through various APIs from the ground up. It’s a complete infrastructure, and not the unreliable, ad-hoc opt-in mess that exists currently in the form of Compiz, GTK and Cairo.
The solution is simple, be more independent from Trolltech. Period.
GTK development is not independent from what Red Hat wants, and Gnome currently sways in the way Red Hat, Novell and Sun wants it to, so I don’t know what you mean there. Novell’s contribution to Gnome has currently bogged down with their efforts on Mono, many Gnome developers have actually left, Red Hat is wandering off producing an online desktop and Sun is doing, well, I don’t know what. The best stuff is coming from more independent developers.
As for KDE, most KDE developers are not in the employ of Trolltech, and KDE uses Qt because most developers don’t want to fanny around debugging their own development tools. KDE benefits an awful lot from Qt and a good relationship with Trolltech. Trying to slow KDE down by asking for that to be severed is pretty weak ;-).
>The GPL license isn’t a problem for KDE, just as it isn’t a problem for the Linux kernel. What the Linux kernel and the GPL has given us, certainly, is more open sourced drivers shipped with the kernel itself.
1. you can’t compare a API with a program.
2. Linux is a really bad example. The Linux you get from kernel.org is non-free. It contains binary firmware for special hardware! gNewSense and Debian remove this binary blobs from their version of Linux.
3. If we talk about promoting free driver Linux is again a very bad example. The Linux developers has accepted non-free driver like NVidia and uses/promotes wlan drivers with a non-free hal (e.g. madwifi) while the OpenBSD guys produces a real free driver (e.g. real free atheros driver). If it comes to free driver especially in the network field OpenBSD does a much better job than Linux.
Edited 2007-12-27 15:04
and uses/promotes wlan drivers with a non-free hal (e.g. madwifi)
Just to butt in the conversation: they are actually working on a completely free hal, and it is included in the recent rc versions of the kernel. It is still very much under development but when it is finished it’ll beat any of those older implementations.
>>and uses/promotes wlan drivers with a non-free hal (e.g. madwifi)
>Just to butt in the conversation: they are actually working on a completely free hal…
You should have quoted/read the whole sentence:
“The Linux developers has accepted non-free driver like NVidia and uses/promotes wlan drivers with a non-free hal (e.g. madwifi) while the OpenBSD guys produces a real free driver (e.g. real free atheros driver).”
It was OpenBSD who created the free hal. The “Linux guys” ignored it for a long time and now finally port it over to Linux. The driving force for this driver/hal comes from OpenBSD while the “Linux guy” where quite happy with their madwifi driver + non-free hal for a long time.
Edited 2007-12-27 16:01
“””
“””
On both sides. “Why are there sides?” is the interesting question. Why do people feel threatened when others make different choices? That’s not directed at you. It’s just a general question that suggests itself to me whenever I run into a desktop wars thread.
“””
“””
Don’t count your chickens before they’re hatched. I think that it is pretty clear that KDE will be a strong player, and an excellent choice for many. But desktops are obviously not a one size fits all proposition.
And on that topic… be careful what you wish for. You don’t *want* KDE to be “The Future of the Free Desktop”. It would have to become something that most current KDE users would probably dislike.
Hey, I feel the urge to say you’re right
Indeed, if KDE is the future of the free desktop, that doesn’t mean many other projects can’t co-exist… There will always be choice. Luckily.
KDE doesn’t fare worse than MS Windows.
That one made me smile, because I have to work in Windows at my job, and I use KDE at home.
And KDE is in fact like Windows, but with the concrete walls removed which I constantly hit whenever I have to use the Windows desktop’s functions.
There are just 2 things which Windows really lacks compared to KDE:
– Multiple Desktops.
– A tile-able, tab-able file/network browser which also can show PDFs and is not cluttered.
Well, the second one is also missing from GNOME
.
He cannot make a recommendation on Gnome, as a recommendation is your own opinion that other people use to assess whether something would be good for them or not.
Have a look here to see my personal recommendations for a few things.
http://www.ciao.co.uk/reg.php?FriendId=5964893
The Ubuntu one got slated though :/
Remove that post. Bloody personal recommendations my ass…..
Has nothing to do with anything, if yuo want to get free clicks and stuff. please post on myspace or some other turd site… to do so in a serious site as osnews is just an insult.
Yes, indeed.
I posted the incorrect link. The correct link was this.
http://www.ciao.co.uk/Member__donley_5964893
Spatial mode is terrible. This is maybe the only thing I hate about Gnome. It makes your desktop a real mess and you have to close each new window individually…
http://www.gnome.org/learn/users-guide/latest/nautilus-spatial-mode…
Then use browser mode!
I don’t use it, either so, I just use Gnome Configuration Editor(gconf2) and under “apps->nautilus->preferences”, I check the “always use browser” option to turn off spatial browsing.
If I’m correct, some distros (Ubuntu, for example) do this by default.
In the end, this all boils down to preferences. I used to be a die-hard KDE person myself and I went to Gnome. I guess, to some, that would make me backwards.
Well, I am just happy to be reminded that there are desktop environments that cater for different users. Thumbs up for diversity!
“””
“””
Indeed. I’m not sure what the inflammatory opening paragraph of the article really had to do with the rest of the article. But only believers in the “One True upper($fill_in_the_blank)” could be overly concerned about Gnome vs KDE.
I tend to subscribe to the 35-35-30 rule. Competition is at its best with two strong contenders at ~35% each, with a mix of other players making up the remaining 30%. You can adjust the numbers to taste; there is nothing magical about 35-35-30.
Two contenders making up 70% of the market means it’s *choice* but not *chaos*. The two players have strong incentive to compete with each other… but also to interoperate well. And you never know when a new titan is going to emerge from that churning cauldron of other players.
I’d like to see the same in the OS market.
Ive heard that for instance SUN doesnt bundle KDE with solaris because KDE has no stable API as Gnome has. Therefore it is much harder to support KDE, than Gnome. Can someone confirm this?
Seems like political excuses in my opinion. Sun sponsors GNOME and they have invested a lot of cash and developer time on various aspects of GNOME, so I am not surprised they ship GNOME and not KDE. I don’t think it has anything to do with the KDE API, it just seems like they prefer GNOME.
No, every bit of software that is integrated is evaluated on a large set of criteria. API stability, licensing, developer responsiveness, etc. are all considerations.
While at this point Sun may continue to choose GNOME due to their past investment, their initial choice of GNOME over KDE was for different reasons.
At least the KDE 3.x series has a stable API. That’s the reason why KDE 3.5.8 is still stuck with ARTS. GNOME OTOH replaced their old sound system (whatever that was) with GStreamer.
At least the KDE 3.x series has a stable API. That’s the reason why KDE 3.5.8 is still stuck with ARTS. GNOME OTOH replaced their old sound system (whatever that was) with GStreamer.
GStreamer is _not_ a sound server. GNOME still uses the Enlightened sound server, although I hope not for long.
Because esd is just so much better….
No they didn’t. Certain apps uses Gstreamer but so can certain KDE apps (like Amarok and Kaffeine).
The GNOME system sounds still uses esd.
No they didn’t. Certain apps uses Gstreamer but so can certain KDE apps (like Amarok and Kaffeine).
That’s because GStreamer isn’t a sound server like Arts or ESD. Gnome is moving to PulseAudio which is a much better sound server than either ARTS or ESD.
http://www.pulseaudio.org/
That’s just not true. KDE guarantees source and binary compatibility throughout major releases. This means that anything that worked with KDE 3.1, for example, continues to work throughout new releases in the KDE3 series.
Similarly there are guarantees for KDE4 in place, so that once KDE4 is released, it should be forwards compatible. Some parts are exempt for the 4.0 release though I think, plasma for instance. But in general Gnome and KDE both provide the same guarantees.
quote from http://www4.osnews.com/permalink?273661
The main problem with C++ ABI is the GNU folks. They keep changing their C++ ABI. Sun has had a very stable C++ ABI for years now. This creates the gross incompatibility situation discussed here.
This part isn’t true. Trolltech supports Sun Studio (SUNWPro C++ compiler) fully on Solaris.
Adopting gcc over Sun Studio isn’t an option as the gnu folks refuse to integrate certain patches that would fix many of the performance problems that some of their generated code has on SPARC, etc.
Yes, this is another reason.
This is one truth. GNOME has a far more mature accessibility environment and Sun, as a public US company, has a necessary commitment to section 508 compliance.
Regarding the C++ reliability – it’ll be interesting once LLVM’s CLANG front end is finished; with the backed support of Apple it’ll mean that ABI stability will be a must – hopefully then for C++ ABI stability, the competition another compiler brings, it’ll force GNU to get its act together.
At the end of the day, if harmony can be found between them, then hopefully LLVM will be mature enough for GCC to be yanked from Solaris and replaced with an opensource compiler which realises playing nice with each other (including proprietary compilers) is a good thing(tm).
Yes, Sun ships GNOME because they can then add proprietary closed source applications to an open source operating system without paying someone licensing fees. Isn’t that just wonderful?
Using GNOME means more closed source! So, if Sun (and you) truly believe in open source and are not just using it because they can be greedy off the backs of volunteer developers, they would include a purely open source desktop like KDE.
Personally, I want MORE open source, not less.
That’s a load of crap. If KDE were the only game in town they they would just pay the licensing fee. They are not broke after all, they have money. The reasoning though is why poay for something that you could already get fro free. Considering there are other incentives to use Gnome/GTK+ I don’t see your point. You also have to remember that issues with Qt licensing stem from whenever the KDE project was first created. Qt restricts a users freedom to sell his product if he can’t afford the licensing fees. Sun in this case is most likely looking out for themselves and trying to save some cash, but that is not to say that their developers don’t appreciate the fact that they don’t have to pay a licensing fee to write commercial apps.
I would also say that you could probably find more closed source qt apps than you would gtk, but that is a bold statement that might not actually be true. I have never seen a closed source gtk apps in the wild, but have seen seen a couple of qt applications here and there that are closed source.
Actually, look at Maemo for a load of closed source applications using Gtk in the wild. As well, look at the gtk based chinese mobile phones. Thats what happens when you use the LGPL – you get more closed source applications. If I wanted that, I would stay with Windows or Mac. I want more free software.
I think the assumption that they would produce Free Software if Gtk+ would be GPL is false.
They would still produce non-free software either by choosing a different toolkit, by building their own toolkit or by choosing a complete different (non-free) platform for their devices.
On the other hand by using Gtk+ they at least spread some Free Software and maybe lead some people trough it unintentionally to Free Software. They may also improve Free Software (Gtk+) by adding features they need, fixing bugs they found etc. Which will than benefits Free Software as a whole. Maybe they will employ Gtk+ hackers to work on Free Software and so on.
Also by making Free Software somehow the “standard” underlaying technology for such devices, even if the programs on top of it are (today) non-free, could help Free Software and increase Free Software usage and development in the long term.
Edited 2007-12-26 18:26
If there was no LGPL GTK, they would indeed do any of the things you mention. Or use GTK/Qt in the GPL version, but properly PAY for it – and FOSS would benefit.
So a toolkit should be GPL, not LGPL – both allow commercial development, yet the latter doesn’t allow FOSS to profit from it.
Stallman is that you?
Again. That’s a load of crap, because the qt toolkit doesn’t automatically make more free application just because it forces developers to pay for commercial licenses. Gnome is not less free than KDE. In-fact gnome is more free because it allows you to link to closed source applications and to commercialize your product without the restrictions that the Qt license has (I call it the “Pay-up bitch!” license), while the gtk license is more lax (I call it the “Please bend me over and screw me” license). As long as no changes are being made to the gtk toolkit itself I don’t see an issue. Trolltech gets to benefit from changes made to their toolkit from the community (like phonon) and make cash off of it. GTK+’s license forces developer to release code if changes are made at the toolkit level, so in the end it all really works out.
Edited 2007-12-26 18:30
>Stallman is that you?
Definitely not. RMS doesn’t use the Web and Afaik RMS has nothing against KDE or Qt but he definitely supports Gtk+ and GNOME which are both GNU packages.
Edited 2007-12-26 18:36
gnome is not ‘more free’. The consequence of using an LGPL toolkit like gtk, is that you will inevitably get closed source applications. The LGPL restricts your rights as a user.
Using GPL libraries like Qt insures that _all_ the applications developed with it are free, forever.
If you don’t like free software, go back to windows and leave us free software loving people to use linux.
The LGPL is not about free software, it’s about closed source.
You are confusing two different definitions of “free”. Your definition of “free” is ‘let other people do whatever they want without consequences’.
The other definition (the one you don’t get) is about the “freeness” of software.
Both definitions have their right to exist. The first definition sets will of individual humans above everything. The second one sees software as a form of knowledge that should be free to access for everyone in the society. (Eben Moglen has held a very fascinating speech about the latter. It’s available under http://video.google.de/videoplay?docid=6345039926759549406 – his speech starts at 5:20 minutes)
There are also lots of closed source Qt apps out there with Skye and Google Earth being among the most prominent Qt apps.
>Personally, I want MORE open source, not less.
I want more Free Software too.
But that’s not necessary the answer of the question “GPL or LGPL?”
It can be good to use the GPL for single libraries like described here: http://www.gnu.org/licenses/why-not-lgpl.html
But i think a GUI toolkit is a classic example for the LGPL case. A GUI toolkit is as essential as a simple c-library because today most programs will come with a GUI.
What would happen if the glibc would be GPL? All non-free software companies would have used a different c-library and all distribution would ship this library at least as addition to the glibc. So this wouldn’t increase the number of free software packages on you system but the number of non-free packages. At the same time the non-free software developer wouldn’t put their efforts into the glibc but in their own c-library. This would mean less developers for the free software c-library.
At the end you would only lose. Less developers for the glibc and one additional non-free package. So it was and still is good that glibc is LGPL and the same is IMHO true for a GUI toolkit because today it is as important as a c-library.
If Qt would have been LGPL from the beginning today we would probably have one supirior toolkit and less problems (all the Gtk+ <-> Qt compatibilty/usability issues). But so Qt has split the community and the development power. First because it was non-free and than because it devides between free software developers and non-free software developers.
Edited 2007-12-26 18:01
> If Qt would have been LGPL from the beginning today we > would probably have one supirior toolkit and less
> problems (all the Gtk+ <-> Qt compatibilty/usability
> issues). But so Qt has split the community and the
> development power. First because it was non-free and
> than because it devides between free software
> developers and non-free software developers.
If Qt were LGPL from the beginning, there would be no more professional development of Qt, and there would be no GNOME as well.
>If Qt were LGPL from the beginning, there would be no more professional development of Qt, and there would be no GNOME as well.
As long as you don’t explain your conclusion i can’t response to it. But i would say that it is (at least in this general statement) false.
Edited 2007-12-26 18:22
>>If Qt were LGPL from the beginning, there would be no more professional development of Qt, and there would be no GNOME as well.
>As long as you don’t explain your conclusion i can’t response to it. But i would say that it is (at least in this general statement) false.
The reason GNOME exist in the first place was because of the Qt license, so had Qt been LGPL from the start GNOME would never have been started.
The professional development of Qt are payed for by companies wishing to use it making close source applications. If they could do the same for free with the LGPL, this revenue would vanish. Without this income there would have been no money to pay the 50+ full time developers working directly on Qt.
>The reason GNOME exist in the first place was because of the Qt license, so had Qt been LGPL from the start GNOME would never have been started.
Than GNOME wouldn’t exist as a conseqeunce of a non-free toolkit but maybe as a consequence of different reasons. I think changes are high that we would nevertheless have more than one Desktop. Because not everyone want the same desktop. But the underlaying technology would probably be the same (Qt). Would we have the same GNOME like we have today? Maybe yes and maybe no. But i don’t bother that much about the existence of one special software. I bother more about Free Software in gernal.
>The professional development of Qt are payed for by companies wishing to use it making close source applications. If they could do the same for free with the LGPL, this revenue would vanish.
They would definitely need a different business modell. But i believe in commercial free software. And i think there could exist a software company around a LGPL toolkit which drives the development (within a community) and makes money with services, fast solutions, fast response times, etc for other commercial (free and non-free software) developers.
Edited 2007-12-26 19:06
Read the comment below, I believe this links to it: http://osnews.com/permalink.php?news_id=19086&comment_id=293332
I think it is very true…
The point of the Lesser GPL license is to provide a compromise for certain applications or libraries that don’t otherwise offer a compelling enough advantage over non-free/non-GPL alternatives. It is discouraged by the FSF unless otherwise necessary.
Gtk relies on LGPL for acceptance because the very implication is that basically nobody would pay a license fee to use it. Qt can utilize the GPL (as well as a variety of other recognized OSS licenses) because if offers enough value that Trolltech still earns millions in license fees for commercial usage. That simple fact speaks volumes.
I don’t think SUN initially went with GNOME instead of KDE for stabilitys sake, nor for technical superiority. I think it’s a GPL vs LGPL for the base libraries thing.
I generally think this is why GNOME has more corporate backing than KDE. They don’t want to say to their customers: “Want to use our supplied platform to write non GPL apps? Contact this third party and pay them for a license.”
(Yes, I know. You can run GNOME apps under KDE but it’s not the same thing. Corporations don’t want to support two stacks if they don’t have to.)
(Btw, I’m not saying GPL is bad or good for the platform libs. I’m just stating why I think the situation looks as it does.)
Seriously, you believe that a company like SUN that had kept a language like JAVA Closed-Source for years would even spend a thought on GPL/OSS vs. LGPL? I believe that all those reasons are a red herring.
Just look at it from a marketing/technical point of view and the reasons for SUN’s desktop choice become very, very clear.
Qt offers a cross plattform framework with a nice API that caters for almost exactly the same developers as JAVA does. It extends C++ with a lot of capabilities that make it a viable and sound competitor to JAVA. KDE is built 100% on Qt.
Now if you were SUN and you’d like to provide a desktop plattform that advertizes JAVA as the development plattform of choice: What would you choose: A desktop that is completely and successfully built on top of a library like Qt that offers many of the advantages JAVA does?
Or would you choose a desktop that is built on top of C and a toolkit that makes any modern developer cry for JAVA? Certainly choosing the latter would make you sleep much better at night.
Of course you’d never admit that openly, so instead you’d come up with a lot of reasons that get gladly accepted by license- and C- zealots.
Your Java/LGPL vs GPL statement is untrue. SUN isn’t the only company who has chosen Gnome over KDE based on Trolltech’s licensing scheme. RedHat, Novell, Ubuntu (though they had other reasons besides just licensing) have chosen GTK+ because of the ability for developers to write commercial applications without having to pay a third party. Just because Sun has had proprietary applications of their own doesn’t mean that they aren’t interested in how licensing affects their customers. At this point you are either naive or a fanboy. I tend to think the latter.
Qt is a nice toolkit. Ars is running a story on some of the great things coming in 4.4(http://arstechnica.com/news.ars/post/20071226-first-look-qt-4-4-0-w…).
However the fact remains that the dual license approach of the toolkit and prior issues with licensing, which kick started gnome in the first place, are affecting how corporations view KDE adoption. These companies aren’t actually thinking about customers and trying to avoid the heartache of paying a third party licensing fee for commercial development for their customers, they are trying to avoid paying the fees themselves. Companies like Novell, Sun, and Redhat, build services into their offerings that they may want to charge for, some like RedHat release all their code, but maybe Sun doesn’t want to do that and if that is their choice why pay for a license to commercialize your application, when you can get it for free.
I tend to think that all this corporate mentality has actually hurt GTK+. The toolkit has changed very little and only recently has started implementing a more current feature set. The corporate mentality makes changing things very hard because one thing a business minded company wants is for things to progress slowly, no change would probably work better for them. Qt has the luxury of being able to make drastic changes to their toolkit because they are competing with other toolkits. Trolltech is trying to sell a product and features are what attract customers.
The idea that Qt is unattractive to corporations is suspect. From Google to Skype to Adobe to NASA, there are tons of people who are building very serious software on Qt.
http://trolltech.com/customers
Why do they pay for something when they could get gtk+ for free? Think about it.
>The idea that Qt is unattractive to corporations is suspect.
You have to distinguish. You list software companys who decide to use Qt. That’s not the “problem”. Sure, every company is free to choose their toolkits and Qt isn’t that expansive if you compare it to the costs of large software projects.
The “problem” arises for companys which offers an operating system like RedHat, Sun, Novel,…
If they would choose KDE they would “force” their customers to become customers of a second company if they want to write non-free software in/for the supported and recommand environment of the OS.
By contrast if they choose GNOME they can deliver their customers everything they need to do whatever they want. They than can of course still decide to use Qt.
Edited 2007-12-26 23:45
All true, except that the fact a distribution ships KDE by default doesn’t mean its customers have to develop with Qt… If that would be true, WxWidgets would never be used, not having a major DE build upon it
If they would choose KDE they would “force” their customers to become customers of a second company if they want to write non-free software in/for the supported and recommand environment of the OS.
Well, you have two options. Either you put all your time and resources (and that’s not insignificant) into developing something that is good enough for developers out there to want to use, or you basically enter into a partnership with a project and a company to offer a path for non-free software development. Software companies do these agreements all the time, and they won’t bat an eyelid.
Again, this is reasoning that means nothing to a developer in an ISV out there, or to an end user. As a distributor, you want developers writing software for your platform to increase your user and customer base. All the distributors out there are currently doing an exceptionally poor job at that, because the first option has failed. The sole concern here is to get developers developing for your platform in order to increase your customer base, not that you are going to be charging them money to do so!
You also have a semi third option, which is to integrate GTK into KDE, and it can be done ;-).
Right.
I’d like to ask you to explain why an ISV can choose to use Qt to develop an application for a GNOME desktop, but cannot choose, e.g. Gtk+, for developing an application for a KDE desktop.
Sound like a contradiction to me.
>>They than can of course still decide to use Qt.
>Right.
>I’d like to ask you to explain why an ISV can choose to use Qt to develop an application for a GNOME desktop, but cannot choose, e.g. Gtk+, for developing an application for a KDE desktop.
An ISV can of course also choose to use Gtk+ and/or GNOME if his OS come with KDE.
That’s not the point. The point is that “i” as a operating system vendor have two options:
1. Sell “my” customers a OS with GNOME and Gtk+ as the main platform for best integration, best support etc. Than “i” can sell my customers everything they need for the main platform.
2. Sell “my” customers a OS with KDE and Qt as the main platform for best integration, best support etc. Than “i” can’t sell my customers everything they need for the main platform but have to tell them “if you want to do non-free software on the main platform you have to become customer of company X too”.
In scenario 1 “my” customers can leave the main, best integrated, best supported platform if they think that the KDE/Qt platform is superior.
In scenario 2 “my” customers have to leave the main, best integrated, best supported platform if they don’t want to become a customer of an additional company.
Edited 2007-12-27 18:22
A misunderstanding then. The way you phrased triggered this interpretation and it seemed weird.
The error in the assumption is that one always has to use the “native” API directly.
For example the success of Qt/Win among Trolltech customers shows that it is totally sufficient to have your API provider/vendor deal with the native API and let your developers experience a nice API given a specific target language.
Any ISV concerned about best possible integration on more than one platform, e.g. delivering Windows and Linux versions, will basically have three options:
– old style separate development/porting
– using a platform wrapper, e.g. wxWidgets, Java/SWT
– using an abstract platform with integration “plugins”, e.g. Qt, Java/Swing
I agree that the first option will require a non-free software project to acquire appropriate licences from Trolltech, but how likely is it that a company wants to implement just a KDE and GNOME version or both of them separately?
Another assumption is that the “developping” entities intend to sell the product.
There is a large amount of development that is done and stay in house. In this scenario, a perfect integrity between the gui toolkit and the interface (in large sense Gnome+gtk) is more important than supporting more than one platform.
I must admit that I don’t know if a trolltech license is required in this case.
No, the license, whether GPL or proprietary, covers distribution. If an organization utilizes Qt to develop inhouse applications that are not intended to be distributed outside of that organization, there are no licensing requirements. It would be no different than modifying GPL software for internal use without publishing source modifications, there is no such requirement.
That’s a point that seems to be overlooked by everybody saying “my customers don’t want to pay a license for developing applications.”, when really it’s only ISV’s that should be concerned, and they have a completely separate set of metrics for determining value. It’s also worth noting that many of Tt’s corporate clients using Qt internally pay the fee anyways, with annual renewals, simply for access to the level of support Tt can provide.
>For example the success of Qt/Win among Trolltech customers shows that it is totally sufficient to have your API provider/vendor deal with the native API and let your developers experience a nice API given a specific target language.
Yes, but Qt integrates well into Windows while neither Qt nor Gtk+ integrates well in GNOME respectively in KDE.
So basically you as a OS vendor has the option to deliver your customers everything (GNOME/Gtk+) or deliver only the free world (KDE/Qt) and refer to a second vendor for the non-free world.
I have nothing against KDE or Qt. I use Qt by myself regularly and for me it is one of the best if not the best GUI Framework. But i also have to admit that for an OS vendor like RedHat, Novell, Sun, etc who wants to offer their customers everything out of one hand GNOME and Gtk+ is the logical step. Because with KDE/Qt they can’t offering everyting by their own.
Edited 2007-12-27 20:17
True. Probably the only way currently to get good integration is to use a wrapper, e.g. wxWidgets or Java/SWT
My guess is that Unix desktop integration of Qt and Gtk+ will improve as the Unix desktop platforms converge in their service layer.
No. As I wrote before, this would only be true if using “native” APIs was actually required, but since it isn’t, ISVs won’t restrict themselves to one desktop platform by using such “native” APIs but will almost always use an abstracted approach, or not care at all (currently even more popular than using an abstraction platform)
This is a good point.
With Qt any ISV customer would buy their support from Trolltech since it is the source of the framework, while with Gtk+ there is no such thing as a canonical source and the OS vendors get an additional revenue stream.
If you really wanted your customers to only deal with you, perhaps you could buy Qt licenses for your customers and then resell them to your customers?
None of the organizations you’ve mentioned have ever pointed to Qt licensing as an issue. Even Red Hat acknowledged Qt licensing was satisfactory and started supporting KDE long ago.
Unless you were privileged enough to be seated in the boardrooms at the time these decisions were made, how are you able to state this so authoritatively?
The Java argument for Sun is perfectly valid speculation. Qt competes on many levels with Java. Nobody would rationally expect Sun to invest money in supporting Qt.
Novell is pursuing mono as a cross-platform development framework. Seems to me that would also be at odds with supporting Qt-oriented development, although they were forced to do an about face regardless.
Red Hat adopted Gnome from the beginning because of their stance against the original Qt licensing, long since addressed. But they invested in it, employ a number of developers, and one can’t overlook their investment in classpath, which much like Sun, is somewhat conflicting with the Qt approach to cross-platform development. But even cynical minds have to notice that Fedora has slowly started to embrace KDE, and where Fedora goes, Red Hat often follows. Ironic, that.
It’s pedantic to claim that Gnome has been adopted by these organizations because of licensing, particularly without any clear points to back that up. I’d speculate rather, that these organizations selected Gnome due to the success of Trolltech’s model, since it conflicts with their own agendas.
But hey, speculation, I won’t claim it as a fact.
If anything’s a red herring it’s your statement above. Open source and free software are concepts that took a lot of time to be even partially accepted by the IT industry. Saying that “Sun kept Java closed for years” is grossly misleading at best. They didn’t “keep” it closed, it was the way things were done and are still done all over the industry. It takes time and a change of vision to embrace them.
Eventually, Sun opened Java, but for a very specific reason: to give it a fighting chance against .NET. Microsoft has had it in for Java for a long time. They tried to hinder it by releasing their own slightly incompatible JVM. When that didn’t work out so well, they cloned the concepts and came up with .NET, hoping to push it on all the Windows machines.
As previous experiences have shown, a niche vendor like Sun that relies first and foremost on hardware sales (see Apple’s adventures) will have a hard time when disputing a piece of software with Microsoft. Such a vendor has to maintain a quite delicate balance between their hardware and software offers, and that makes them move vulnerable than Microsoft, whose hardware support is independent and diversified in the form of the many PC and PC-component vendors and manufacturers.
Opening Java meant saving it by placing it in the hands of the one and only fenomenon that has proved so far successful against Microsoft’s tactics: FOSS. It was a last resort, and it took Sun, one of the last surviving UNIX titans (UNIX being the original closed platform), a lot of time to decide to do this. They came to recognize FOSS as their friend and savior in this particular case.
I see, you are implying that Java is the best toolkit there is, that Qt comes close and GTK+ doesn’t. Which is rubbish.
If you’ll recall, they first tried a Java desktop, and it didn’t work out so well. Not for lack of quality in the desktop itself, but because it was an isolated desktop. You need a certain set of apps to make a full desktop environment. Sun had to choose between developing their own from scratch and thus reinventing wheels already implemented fully by FOSS projects; and just picking an existing DE.
It’s obvious they eventually went with the second choice, just like Apple did in a similar situation. But when it was time to pick they had to compare them based on a developer’s perspective, since they would need to combine their own pieces of software with it. So the licensing and ABI factors are very important and not a red herring at all. It’s a no brainer that a commercial vendor building on top of a DE would want to be free of licensing contraints even when making commercial apps, and they wanted a stable ABI. GTK+ was the obvious choice for them.
> Seriously, you believe that a company like SUN that
> had kept a language like JAVA Closed-Source for years
> would even spend a thought on GPL/OSS vs. LGPL?
Errr… yes. If you are going to invest in a platform, one of the things you care about is the license. And correct me if I’m wrong, the license situation of QT was much worse back then (late 90s).
<Lots of other trollish comments snipped>
What?
First of all, how do you justify investing in a platform that you think is inferior, in turn making it better, just because you don’t want to invest in the better one.
Second, GTKMM et al also gives you an attractive platform for C++ development. Which is better? Who knows. I don’t think it matters much.
Third, I wasn’t talking about SUN exclusively.
Fourth, I didn’t say it was the only reason SUN went with GNOME. But i bet it weighted quite heavily.
Fifth, and yes, I admit it. I think Java and C# is an evolutionary step to C++ and think they are going to be used more in the future. (Though I’ve written tons of more C and C++ code myself over the last 12 years I’ve been coding.)
Edited 2007-12-26 17:11
I say try both and stick with the one that fits your needs better. There’s no point arguing saying one’s better than the other – that’s simply not true! They both provide a great way for people to work in an environment they prefer.
I used to love KDE but now I’m pretty partial to Gnome. I’ll give KDE 4 a fair chance but Gnome has become more appealing to me than KDE 3.5 with it’s restricted and intuitive feature set.
KDE 3.5 just has too many options scattered everywhere.
I mean… I’m now fully capable of working soley in a command-line environment, without any desktop. So why would I need a desktop with the whole kitchen sink?
I for one prefer the looser pant.
Ok, Whos with me?
…
Boxers!
I prefer roomy “apartments”
… I hate those tight thingies that kills your samosas 
I prefer boxer briefs. Best of both worlds.
cOMMANDO BABY!
Is there any impartial benchmark about GTK+ vs QT4?
To me, the main question is “what DE is quicker”?
I use Gnome all the time, but i’m the first to say it feels sluggish compared to KDE or even Windows’s EXPLORER…
Does anyone agree?
>Is there any impartial benchmark about GTK+ vs QT4?
To me, the main question is “what DE is quicker”?
I use Gnome all the time, but i’m the first to say it feels sluggish compared to KDE or even Windows’s EXPLORER…
Does anyone agree?<
I am not aware of any impartial benchmark that could give a definitive answer as to which is faster, GTK+ vs QT4.
I will say that on modern hardware I personally don’t notice much of a performance difference (core 2 duo, 1.5gig etc..).
However on a slightly older machine of mine (athlon64 2800+ 1gig etc..) I will say that It really does seem that KDE is more snappy and is less likely to bog down under load.
Ultimately I do prefer KDE over Gnome simply because I have used it more and am used to it.
The simplicity of Gnome really seems to get in my way more then it helps me. But I am also not your average office user either.
Edited 2007-12-26 14:27
Hmm, I expected some kind of juicy usability article, and I’m stuck with a polite but nearly useless introduction to GNOME. Seriously, what’s the point of such an article?
I dont see all the big fuss with Kde vs. Gnome.Maybe because I use both and like them with no preference of one over the other.
I agree with you. The desktop used by a distribution is less important than HOW they use it. be it Gnome, KDE or any other. As long as the desktop gets out of my way by behaving like I expect it to, lets me easily set the wallpaper, and set the taskbar to the top of the screen (A personal quirk), I could care less what it is as long as My apps work.
I find KDE unusable these days; it just feels cluttered and counter-intuitive in terms of the menu structure. I like GNOME more and more with each release.
But that’s like me saying I like strawberry ice cream instead of chocolate (or vanilla, or rocky road, etc.)
There’s so many choices with Linux that you don’t have to settle for something that’s not right for you.
“I like GNOME more and more with each release.”
More and more users around me seem to feel the same way. Some of them, having been KDE users for years, like to try Gnome more and more, and some of them have completely changed their daily use desktops from KDE to Gnome.
In my opinion there’s an interesting approach in Gnome, as far as I know: Simple changes of desktop look and feel can be done easily, more complex, complicated or basal settings need a more advanced way to be changed. On the other hand, in KDE nearly everything is changable very easy (too easy?).
But I have to admit that I do not use KDE or Gnome on a daily or “only” basis, because I simply don’t have a need for a desktop environment, I prefer a fast and usable window manager, i. e. WindowMaker for me. But I usually use different KDE and Gnome based applications (Qt and Gtk+ ones), so “keeping the look consistent” does not matter to me anyway, so maybe I’m the wrong person to ask which one of Gnome or KDE is “better”.
“But that’s like me saying I like strawberry ice cream instead of chocolate (or vanilla, or rocky road, etc.)”
A valid comparison.
“There’s so many choices with Linux that you don’t have to settle for something that’s not right for you.”
I think this kind of functional diversity is one of the biggest strengths of UNIX / Linux: You are not forced to use something you don’t like. Good point.
There’s more to life than KDE and Gnome.
There is the wonderful world of window managers and console applications that offer unique user environments. Alas, the open source world is in such a rush to the bottom (the Windows-like KDE and Mac-like Gnome, both placing a huge emphasis on eye-candy), that little innovation in user interfaces will be seen and there is so little to differentiate the open source Unix environments from their competition.
CDE and Window Maker are superior to either.
Care to elaborate? A comment such as “they both suck, these are better” with no qualification renders the opinion rather meaningless, yes?
This is a typical response to any article such as this. $window_manager is superior to $window_manager or $desktop_environment. It boils down to individual tastes, really. What may suck for some works great for others.
“CDE and Window Maker are superior to either.”
I think you cannot compare them to Gnome or KDE.
CDE, for the first point, is “stuck in past”, allthough it runs on many different UNIX OSes and is still in use. On x86 UNIX or Linux systems, XFCE 3 is a nice approach to the classical CDE look and feel, while XFCE 4 is a good example for a lightweight and highly functional desktop system. CDE definitely has some (!) advantages over KDE or Gnome, this is possible and is very dependant on individual experiences and feelings.
WindowMaker, for the second point, is not a desktop environment, it is a window manager. From my individual point of view, it is the best window manager I’ve ever used. Desktop organisation, mouse and keyboard support, look and feel and of course speed (!) are most impressive. That’s why I’m using it for more than 5 years now on a daily basis. And you can use both KDE or Gnome based applications with it, along with any other application that uses “non mainstream” toolkits (such as Tck/Tk or Xaw).
I’d like to know what your arguments are.
Sorry, I do not know WindowMaker, but I had to work on a CDE Desktop for 2 years.
That have been 2 years of desktop pain for me.
CDE:
– Hit the maximize button and you instantly loose access to the panel, or
– configure the panel so that it stays always on top, then hit the maximize button, and you instanly loose access to the bottom of your application.
It’s like a choice between HIV and cancer, both will kill you slowly and painfully.
If I had to decide which is worse, Windows or CDE, I would not know. Windows lacks virtual desktops, which is absolutely necessary for me, and CDE lacks sane panel/Window interaction and wastes loads of vertical screen space.
The author forgets that both, Gnome and KDE, run on many platforms, not just on Linux. KDE is even supposed to run on Windows. I am comfortable with IceWM, but I always admired the idea that computer user can log to any machine, running any OS, and still have the same, familiar working environment.
I think that platform independence is important. Let’s face it, people use applications, not an OS. OS differences are the consequence of their imperfections. Users should not think about OS, but focus on their work, which is the reason why they have computer in the first place.
KDE and Gnome do good job hiding OS from the user.
Just KDE apps are now porting to Windows, not KDE desktop itself.
“The author forgets that both, Gnome and KDE, run on many platforms, not just on Linux.”
One of their strengths.
“KDE is even supposed to run on Windows.”
As it has been mentioned before, “running on ‘Windows'” refers to KDE applications (Qt based ones), but not to the K desktop environment itself. Just imagine. Getting rid of the “Windows” way to do things, having your favourite KDE and your essential “Windows” applications and games…
“I am comfortable with IceWM, […]
Yes, I remember having used IceWM with pleasure, too.
“[…] but I always admired the idea that computer user can log to any machine, running any OS, and still have the same, familiar working environment.”
That’s what we do in UNIX world since the 70s, man!
“I think that platform independence is important.”
In fact, it is, but it’s complicated. I’ll comment to this later on.
“Let’s face it, people use applications, not an OS.”
This is correct, but still, the OS is the part of software that makes it able to run applications at all.
“OS differences are the consequence of their imperfections.”
No, OS differences are an expression of having different solutions to a problems that are still of the same worth. Furthermore, different tasks require different tools. For example, you cannot expect an OS for distributed processing in a mainframe environment to be a good desktop OS. But still you can use KDE or Gnome on both of them to have a familiar interface.
“Users should not think about OS, but focus on their work, which is the reason why they have computer in the first place.”
I agree to that statement, but today’s computers aren’t that advanced to implement this goad completely, so their users need to be advanced.
Just imagine what usually happens to “Windows” PCs if the user just focuses on his work, for example, downloading porn, playing games and treating the PC as a better typewriter – the PC turns into a threat to the user and to others who are connected to the Internet, see viruses, trojans, and spam. An advanced user who knows about the weaknesses of his OS will be able to fix these problems very easily, but the OS itself usually does not provide sufficient means to do this.
“KDE and Gnome do good job hiding OS from the user.”
Hmmm… I agree with you, but I’d like to point out that this implies new problems. Let me explain:
In former times, there was a difference between the user and the administrator. The administrator’s job was to keep the OS running correctly, to install and maintain services and to maintenance and system repair tasks if needed; therefore, the administrator had the proper knowledge and experience. The user did not have to know about all this stuff, because the administrator provided a running system to him.
Today, user and administrator are one and the same person, at least in the case of home computing. Because the user feels as a user, he does not recognize his tasks and responsibilities (!) as administrator. Instead, he expects the OS to maintain itself, or at least a GUI to take care of everything.
This is where KDE or Gnome enter the game: They provide a kind of “mapping” for OS functionalities and configuration tasks onto dialog boxes and applications. Therefore it’s neccessary for the Gnome and KDE developers to know the OSes to provide interfaces to, or at least expect the OSes to have common interfaces for these tasks. The problem here: Many Linux distributoins handle the same things in different ways (e. g. system organisation, package management, install procedures), and the BSD UNIXes are different, too.
Usually, the user thinks very concrete, beginning from the goal of an operation: “I want to have my pictures shared so I can access them from the office.” One valid solution for this task could be to have a FTP server running, another one could be to upload the pictures to an accessible server. So, what can Gnome or KDE do too assist the user? Provide a menu with “Share pictures: (1) Start FTP server, (2) Upload pictures”? Still to complicated for the user (termini technici: “ftp”, “server”, “upload”. How far can you abstract? How much control should a desktop environment be able to take?
Conclusion: You can abstract interaction results just to a certain point. Computers aren’t easy to use, even today. A minimum of knowledge and experience is still needed, in most cases, combined with a bit of common sense and educated thinking and the ability of abstraction. These are requirements to the user, not to any GUI. KDE and Gnome, as well as other GUI solutions, can help and assist, but they cannot read your mind.
That’s why we don’t have a semantic compiler yet.
int a, b, sum;
sum = a – b;
semcc: Error in 2: You did mean +, didn’t you?
In other words, if an OS is installed correctly with everything needed, if package management works correctly and if the user only buys additional hardware that is supported by the OS and KDE’s / Gnome’s set of applications and interfaces, no major problems should occur. But just imagine how complicated it can get if an update fails, a new graphics card is not supported correctly by X or a careless sudo call eliminated an essential GUI library…
I don’t use Gnome or KDE, I use particular programs that do what I want how I like it. As the majority of these are KDE programs, I also use KDE as my DE. But the choice is that way round. If I hadn’t been so let-down and disappointed by some Gnome programs over the past six or seven years – Evolution, mainly, which is the spawn of Satan in my book – I’d probably be using Gnome instead. In my experience either Gnome or KDE make mighty fine desktops for anyone, newcomer or old hand alike. There’s no real evidence I can see that one is somehow “easier” than the other or aimed at the less sophisticated user, whoever that is. Many folks often like to claim this in order to distinguish between the two but show us the goods, mate, instead of the hearsay. There’s nothing difficult about KDE.
In the longer view, with iteration 4 KDE is now putting into place the framework that will carry it through to new and better things for a long time to come. KDE also has the much underrated KDE Office, of which great things may soon come. Gnome has neither, so the amount of attention paid to it may be a mistake – I’m thinking of a lack of attention paid to Kubuntu, for example, which is getting what looks a little like second-class citizenship in St Shuttleworth’s rather resistible Magic Kingdom. No new LTS version for them, only for the Gnome-based edition! Just my 2 cents.
This is the receipt for your 2 cents, thank you, and this is my receipt for your receipt.
“I don’t use Gnome or KDE, I use particular programs that do what I want how I like it.”
I do the same, allthough I do belong to the “poorminded ones” who do not use a desktop environment and insist on a fast and versatile window manager instead.
“In my experience either Gnome or KDE make mighty fine desktops for anyone, newcomer or old hand alike. There’s no real evidence I can see that one is somehow “easier” than the other or aimed at the less sophisticated user, whoever that is. Many folks often like to claim this in order to distinguish between the two but show us the goods, mate, instead of the hearsay. There’s nothing difficult about KDE.”
I agree, I don’t think KDE or Gnome are aimed at such differnt user groups as, let’s say, special window managers (e. g. wmii) do.
One point of distinction between KDE and Gnome, in my opinion, is the way of organizing the installation amount. While KDE’s philosophy seems to be: “The hard disk is big enough, just put everything on it by default, so if the user needs something, it will be there”, while Gnome seems to be more fine-grained, in a way that “Let’s start with a minimal, but working amount of application and let the user install afterwards what he’s going to need” would be a valid statement.
“KDE also has the much underrated KDE Office, of which great things may soon come.”
Personally, I like the good integration of KDE parts, such as KOffice, which I hope will be a viable replacement for OpenOffice so you can create a “KDE-only” system which still has a powerful office suite included.
“Gnome has neither, so the amount of attention paid to it may be a mistake […]”
Doesn’t Gnome have GOffice? By the way, I think Gnome users don’t mind installing OpenOffice, allthough it does not look natively like Gnome, in the same way KDE users would refuse to install it because of the same reason.
The level of attention for a specific desktop environment should not be a function of how it includes an office suite. Integration of its own parts and parts of different environments should be a better point to start from.
One experience I had with KDE: Internationalization. My uncle set up a PC-BSD system with KDE 3. Everything was in German, but the error messages were still in English, this did confuse him. Then, he decided to install Kmplayer, but it wasn’t in German, too. Explaination: Kmplayer, allthough prefixed with K, was not part of the KDE suite so it wasn’t affected by the language setting.
For me, internationalization doesn’t matter because I do use the native english language output which I prefer to the poor and sloppy german translation which you can find in most places. But if a mainstream desktop environment is going to be in a good leading position to promote Linux or UNIX to average home users, they should be able to make it usable outside the US and the UK, too.
I toyes aound with some live system CDs (which did not require an install) which are based on KDE and Gnome. From my not very comprehensive point of view, the Gnome based ones did a better job regarding language issues (e. g. geubuntu, openSUSE Gnome, Fedora 8). But I was told it would be better if they got installed instead as just used as a live CD, but what’s the sense of offering a language selection when there’s no effect?
http://osnews.com/permalink.php?news_id=19047&comment_id=291898
Edited 2007-12-26 20:29
Your analogies about KDE and Gnome are pretty accurate.
When Gnome went 2.0 they consciously made an effort not to try to reinvent the wheel for everything. They wanted gnome to be exactly what it is today, a dekstop environment. They didn’t want to make a desktop environment/distro. In other words, they avoided writing apps that a third party could write themselves, but decided to focus their attention to apps that are essential to the desktop experience, one where tight integration is a must, anything else would be handled by the distromakers. I think that is the right choice. Think of Gnome as plain old windows without all the OEM crap installed, and the distro as the OEM who adds functionality to the desktop. KDE has a lot of projects under its wing and while you don’t have to install all of them, its usually encouraged for a KDE distro.
In other words, they avoided writing apps that a third party could write themselves, but decided to focus their attention to apps that are essential to the desktop experience, one where tight integration is a must, anything else would be handled by the distromakers.
While KDE definitely bundles more software than GNOME, some of this seems to simply just be a matter of time. Both bundle web browsers, full featured email suites, video players, music players, CD burning software. They’re both trying to deliver the same thing. A complete desktop environment. They simply differ in their approach.
Personally I love and hate this, as while I love choice, I have too much positive and negative to say about both making it difficult to choose!
KDE doesn’t bundle software. The packager does. There’s lot’s of software written within the KDE community, but most of it is optional. I don’t have to install KPDF, I don’t have to install Kopete, I don’t have to install Amarok, etc. These are all separate packages at least on (K)Ubuntu and openSUSE.
Usually distributors define meta-packages like “kubuntu-desktop” on (K)Ubuntu that install many apps in one go, but kubuntu-desktop != KDE just like the ububtu-desktop package is not identical to GNOME.
It’s true that bundling is up to the distro, but many of these projects are tied to their respective desktops. KPDF and Kopete are both official parts of KDE. Amarok is as well to a lesser degree, being part of KDE extragear as unlike the prior two it’s not tied to KDE’s release schedule.
Likewise though of course, Evince is a part of Gnome. Empathy is on it’s way to being a part of the main desktop, and Rhythmbox is practically already there.
Though the most important point you hit on is that you’re not required to have any of these with most modern distributions.
You’re blurring the line between KDE and distro here. KDE itself consists of two packages, kdebase and kdelibs. That’s the core framework for KDE with enough bits to give you a workable desktop. All of the other apps are packaged in functional groups (kdeutils, kdegraphics etc.) and then there are optional add-on packages. And then there are the KDE Extra apps, which aren’t part of the official KDE project. The devs have gone to great lengths to make package selection modular without interdependencies, which is admirable. The fact that distro packagers often just throw everything together as a static meta-package shouldn’t detract from the work they’ve done.
Oh geez where do these guys come from? KDE4 at this point has shown little and at this point its nothing more than a lot of promise. Who knows it might end up exactly where KDE3 is at right now, meaning a whole bunch of featuritis. It may actually do well. Who knows, but gnome isn’t going anywhere. Gnome doesn’t need huge releases to update their desktop they do it slowly (sometimes ploddingly) but they get there eventually. GTK is what needs to change not Gnome. GTK needs to gain more developers, in-fact I rather that gnome take a backseat for a cycle and that all devs go work on gtk for those 6 months and add some much needed features. At this point KDE4 is nothing more than hype, even the release candidate, is not a release candidate,its more of a release candidate for a beta release, if you ask me. There are a couple of things I find interesting such as Plasma and Phonon, but I have yet to see where things are going and I will wait. Besides, most distros chose Gnome for reasons already stated here, and among them the 6 month release cycles being the key component for the most popular one(ubuntu). If the KDE devs and trolltech can work on the license and fix their release cycle, then yeah, maybe distros might give KDE4 a second look, otherwise, I don’t see it gaining any momentum past second-class citizen with certain distros.
huh? Since when is a publicly developed software a ‘promise’? Anyone can download it and compile it and use it.
oh my goodness, heaven forbid new features in software!
The real reason why gnome/gtk development is slow, is becase of the toolkit and API are difficult to work with.
The reason why kde had such a big jump from version three to four is that Qt changed drastically. You have to do that sometimes to weed out the flotsam, and produce a better API that developers will have an easier time producing better apps.
You seem to think that changing the gtk API will not effect or change Gnome.
Then again, you mentioned ‘featuritis’ as a detractor for KDE and then say there are some much needed features missing from gtk.
huh? How is releasing a release candidate be considered ‘nothing more than hype’?
No need to work on the license, as the GPL is just fine and promotes more free software, unlike the LGPL which promotes closed source.
Edited 2007-12-26 21:52
First of all there is a difference between features that make sense and features for the sake oif features. Yes I believe that too many features isn’t a good thing, especially when they are not presented to the user ina coherent manner.
No working on gtk+ doesn’t necessarily mean that it will break compatibility with gnome, I’m suggesting they add things not take anything away. If you already know how gnome works you know already know that changes in gnome are spread throughout versions so anything that needs deprecation will eventually get to that point as the toolkit infiltrates itself into gnome. By your definition then KDE must have broken compatibility just by switching to Qt4 (here is a hint, they didn’t).
There is a difference between lots of features in a toolkit and too many features being displayed to users in a Desktop. One is strictly for developers, thus makes their lives easier makes development go faster, lets them do things that could enhance the user experience. Having too many features in a desktop application or throwing the kitchen sink at the user is a sign of a developer who has no idea what he wants his app to be, or isn’t confident in who he made the app for. A more focused approach is always better when it comes to applications that you would expect a user to use. Throwing everything at them and seeing what sticks is the wrong approach, imo. A toolkit should be featureful, a desktop should have whatever is necessary for the user to get from point A to point B, he doesn’t need to go through A-, B- whatever you want to put in between in order to get to B. Its just bad design. Some people like this, I moved to a Mac and Gnome because I don’t.
I still think you are wrong about the LGPL vs. GPL. One does not have any bearing on the other. There are many companies willing and able to pay the cost of licensing for commercial development, there are many that do already. Your Stallmanesque view on the whole situation seems to be very close minded, imo. In another post I might even agree with you, but for some reason, I find your reasoning flawed, and kind of unfair. in this instance.
Does anybody know if Qt can be forked? If so why hasn’t this been done and the licensed changed to allow it to be LGPL’ed. I don’t think so though but I’m not sure.
Does anybody know if Qt can be forked? If so why hasn’t this been done and the licensed changed to allow it to be LGPL’ed. I don’t think so though but I’m not sure.
KDE won’t dare to fork Qt because being honest ain’t nothing but TrollTech’s bitch, as bad as it sound it is.
And you can’t turn a GPL license into an LGPL license, KDE would need to make its own toolkit and who knows if they have the resources or talent to do it.
Edited 2007-12-27 01:05
And you can’t turn a GPL license into an LGPL license, KDE would need to make its own toolkit and who knows if they have the resources or talent to do it.
Translation: I want KDE to run around jumping through unnecessary hoops because I don’t like the fact that they’re going to be a huge, and probably insurmountable, distance ahead in terms of features and quality, which other free desktops are going to have real difficulty keeping up with.
When you’re talking about KDE and the features that matter, ‘and it can’t even beat Gnome’ (in your words), that’s a pretty damn sad thing to have to come out with, isn’t it?
Only ejits on forums and on development lists talk about forking something simply because they don’t like the license. Sadly, this has a negative effect on end users, and the funny thing is, no one talks about how the fork is going to be superior to the version developers already have, and how that will translate into a better desktop and applications for end users. That’s all that matters to everyone.
riiight… I know enough about Qt 4 to know it not not backwards compatible, so I doubt that KDE 4 is.
Closed minded? Maybe, but probably more realistic than someone who thinks LGPL is somehow “more free”, because it restricts the users rights to source code.
Qt can be forked, sure.. after all it is GPL. But you cannot change the license to something you do not own the copyright to. GPL is always going to be GPL and always going to produce GPL software that links to it. Unlike LGPL, the GPL is a more pure form of open source. It gives everybody the same rights.
Again. LessorGPL IS about closed source, there is no debate about it, you cannot successfully argue that it isn’t. If you don’t believe it, you need to do some reading. There is a reason why it is called the _lessor_ GPL.
http://www.gnu.org/licenses/why-not-lgpl.html
As far as kde4 is concerned you are seeing what you want to see so there’s probably nothing to be gained pointing out to you that its NOT FINISHED YET!
Gnome is fine as a beginners os and if it suits your needs fine, just use it shut up. KDE does not have to dumb down to gnomes level just to make people like you feel better about your choice of desktop. Kde is looking to the future and gnome, as you correctly state, isn’t going anywhere. Deal with it.
Right here, you have eloquently summarized one of the biggest advantages KDE/Qt has. KDE devs don’t need to waste cycles developing the toolkit, there’s a very competent commercial organization doing it for them, and the community reaps the benefit.
By arguing that Gtk lacks proper developmental support, you’re pretty much validating the pay-for-commercial model Qt employs. The community benefits from free use of a powerful framework, and commercial developers benefit from the availability of a powerful and supported framework. It’s all good.
I never argued against it. In-fact I wish that GTK had the same model or at least that the project would detach itself from Gnome. What I was arguing against was the whole GPL vs LGPL promotes more free software argument that was being spread around.
I’d argue against this point. KDE and Gnome exist at a layer above the underlying platform, both are to one extent or another designed to be OS agnostic. In the specific case of KDE and Gnome, it’s the responsibility of the distros utilizing them as a component to provide that type of configurability. KDE/Gnome have no business wasting time on dealing with the myriad of issues related to the use of the platform, they simply exist to provide a rich environment for application use.
KDE and Gnome should be able to handle ftp as a protocol for uploading pictures, for instance, but setting up an ftp server and dealing with the distro-specific issues is far out of their realm.
Granted, and to your point, KDE4 is taking an interesting approach with Solid and building platform-specific requirements into a single interface, to alleviate application developers from having to care about whether they’re running on lin/win/OSX. But I still think that’s a far cry from making the DE’s responsible for management of the underlying system. That’s one of the particular values that the distro should be bringing. Some like openSUSE or Mandriva, provide a powerful management interface for handling that, others rely on CLI magic, but that’s separate from the DE projects.
I’ve made a mistake in my post. There could be a distinction between OS and platform. KDE could be a platform and OS could be whatewer it is. Platform which is OS independent means that person has the same experience no matter which OS is running.
Yes, people using UNIX have similar experience on Solaris, FreeBSD and Linux. But there is a small number of UNIX users, I mean persons sitting in front of UNIX machine. It is a shame, but that’s the fact.
KDE and Gnome could go beyond UNIX.
I agree that computers are not easy to use. I am not trying to address that. I just say that it is easier if one has consistent interface no matter what kind of OS is running.
The problems you described (ftp, upload) could be partially solved if platform and applications support scripting, including GUI scripting. I’ve set up Linux with Windowmaker and XDialog scripts for my mother who had lot of problems with Windows before that. Now, all she needs to do is to activate an icon. If I needed I would have written macros for OpenOffice, too.
So, I think there should be exist a scripting language that works with both, platform GUI and applications. Microsoft made an effort with Vbscript, but I didn’t like the result.
…I prefer Gnome, but that might be because I’ve used it since 1.0 (part of RH6).
Gnome is very tweakable, but unfortunately many of these tweaks are unintuitive to apply (GConf is not userfriendly, nor intuitive). KDE is better at presenting tweaks in a userfriendly way. Gnome however has the cleaner layout so IMHO it doesn’t matter what you choose, as long as you choose the one that works for you.
This may be even more true when KDE4 is released.
> Open source and free software are concepts
> that took a lot of time to be even partially
> accepted by the IT industry. Saying that “Sun
> kept Java closed for years” is grossly misleading
> at best.
So you’re trying to tell me that
a) they decided to use Gnome due to the “more visionary” license concept in 2000
b) on the other hand they have resisted for 8 more years to use that same “more visionary” license concept for Java (despite vast complains from the community).
And on one hand you try to tell me that it took them so long “to see the light” and on the other hand you admit that they got basically just pressured to open JAVA to stay competitive. I rather see a lack of choice here than a true process of “insight”. Therefore I question that a) was an important issue.
> I see, you are implying that Java is the best
> toolkit there is, that Qt comes close and
> GTK+ doesn’t. Which is rubbish.
Java, .NET and Qt provide a modern development framework (plus they are including their own language or extending one). All of them are much more than a toolkit.
Gtk on the other hand is still mostly a toolkit and leaves it up to the developer to pick from the wealth of other available (free) framework components.
So they are very different in nature: the former provide a full consistent and comprehensive solution for the whole development of an application.
The latter mostly provides a solution for a single aspect of application development.
> So the licensing and ABI factors are very
> important and not a red herring at all.
While I don’t question the importance of a stable ABI it has been pointed out by others already that this doesn’t/didn’t apply in this case, so the reasoning is flawed.
SUN had the choice to put one of their most important software products either on top of the show car of a product that competes strongly with their most important product. Or they would put it on top of an environment that would provide development tools that don’t fall into the same category and don’t compete as close.
It’s pretty clear that this was one of the major driving factors behind their decision.
For most other companies the decision which toolkit to pick relies on the availability of developer/knowledge resources inside the company or close to their company.
Given that I promote Free Software myself I’d love it if companies picked their tools just for philantrophic-ethical reasons, good karma and the deep insight that Free Software is a superior concept.
However market realities unfortunately are still different these days. Much of the time the true realities and motives are disguised the way people just want to see them.
>Java, .NET and Qt provide a modern development framework (plus they are including their own language or extending one). All of them are much more than a toolkit.
Gtk on the other hand is still mostly a toolkit
yes, Gtk+ is “only” a toolkit like Swing or Window.Forms are only toolkits. You can use the toolkit Gtk+ with plain programing languages (C, C++,…) or with complete frameworks (java-gnome, Gtk#).
It is up to you. Nobody forces you to use Gtk+ only with a plain programing language.
>SUN had the choice to put one of their most important software products either on top of the show car of a product that competes strongly with their most important product.
Qt is a great GUI toolkit and a great framework for GUI apps but it doesn’t really compete with Java. The strength of Java is the client/server market (and maybe the in-house desktop market) but not the traditional desktop market.
Edited 2007-12-26 17:46
In terms of API vs API, well, as an end-user it’s difficult to evaluate. However,for my personal needs and preferences of behavior I’d say that KDE-based projects dominates my desktop, which is curiously Gnome-based:
Konqueror, KTorrent, K3B, Amarok and the list go on. Big exception is Gimp, which is probably still the best Gtk-based end user app.
Then, for me, my balance is given by very good desktop consolidation (byGnome) combined with a lot of KDE-based apps.
Then, an two interesting question arises:
1. Why KDE (Qt) is producing a much more effective apps for my needs? (I can say that honestly I have tried many of the Gnome propositions)
2. and why the KDE desktop does not feel as-good-as the Gnome-based?
For question 1, we can just say it is a matter of preference and that my brain is more in “synchrony” of how the KDE-apps are designed.
For question 2, that’s more complicated, because I can say that OpenSUSE is almost there, and the main reason that I prefer Ubuntu is because debian-based distribution simmly rocks over rpm (before any troll talking, please consider that I use CentOS & Fedora at work). But the debian-based aspect is totally independent of the desktop choice. And Kubuntu, as others indicated, is by far not so polished as Ubuntu, much less anyways than OpenSUSE. I have also to admit that I prefer the approach of Gnome for the desktop design, I prefer its mechanisms for the management of windows and virtual desktops.
At end, and I can not repeat it enough, I really want that both window managers (and all of the others simplistic/experimental managers) continue to exist and, more importantly, to CO-EXIST until the end of time. The user experience of having so diverse approaches makes that at end you can find the best solution for your needs and liking.
It’s so MS/Apple fanboyism to read comments of people wanting to rule out the FOSS-propositions of their disliking, when the whole point is to have more freedom. I really believe that we can be much better than the products of disgusting marketing trends.
I have never learned to like KDE even though I tried, just out of curiosity and for exercise..I found it too odd and..well, cumbersome/cluttered in comparison to GNOME. GNOME is sleek and just suits me better. And since there are so many GTK+ apps I like and I absolutely hate mixing the toolkits I just see no reason to switch. And well, I just don’t think KDE or any KDE app has anything more to offer than what I already get and need from GTK+ apps.
I have never learned to like KDE even though I tried, just out of curiosity and for exercise..I found it too odd and..well, cumbersome/cluttered in comparison to GNOME. GNOME is sleek and just suits me better. And since there are so many GTK+ apps I like and I absolutely hate mixing the toolkits I just see no reason to switch. And well, I just don’t think KDE or any KDE app has anything more to offer than what I already get and need from GTK+ apps.
I have never learned to like GNOME even though I tried, just out of curiosity and for exercise..I found it too odd and..well, dumbed down in comparison to KDE. KDE is sleek and just suits me better. And since there are so many Qt apps I like and I absolutely hate mixing the toolkits I just see no reason to switch. And well, I just don’t think GNOME or any GTK+ app has anything more to offer than what I already get and need from KDE/Qt apps.
oh wait…
What an idiotic article that was that tells us absolutely nothing. LGPL, licensing, ‘uncluttered’, don’t bother with Qt and KDE applications because they look ‘funny’, Gnome is the ‘default’ in many distributions so you’re stuck with it…… Who on Earth thought that was interesting, I have no idea.
The fact of the matter is that free desktops will take lots of time (the best part of the next ten years probably), development and effort to achieve wider usage, and basically, the one that becomes clearly the most popular will be the one that is able to support developers the best – open source developers first and then others, in that order.
In view of this, and the fact that Gnome lacks features that Windows and OS X have (yes, the bastion of simplicity and usability) it’s going to be even more difficult than it is now to justify it to anyone on the basis of simplicity, being uncluttered and it being the default ‘so you’re stuck with it’.
I’ll leave it as an exercise to the average reader around here to work out what that means.
In view of this, and the fact that Gnome lacks features that Windows and OS X have
Pray what ?
OK I haven’t tried Vista Ugh! and I have only briefly played with an OSX system.
However every day at work I use XP and at home I use Gnome on Ubuntu Gutsy. Compared to the current Gnome, XP is kludgy, slow and old fashioned (which of course is true) while no one I know who has Vista at home likes it.
Gnome is simply the best desktop environment around at the moment. It’s some years since I last used KDE as my primary desktop and I might give KDE 4 a try when it comes out, but for the moment and for me Gnome is the best all round DE and it beats Windows flat out.
Edited 2007-12-27 02:37
“In view of this, and the fact that Gnome lacks features that Windows and OS X have”
Pray what ?
Yep, you heard that right. Gnome lacks the easy ability to run applications as another user, lacks the ability to configure screensavers as on other desktops as well as other shortcomings that basically stem from an architecture that would have a circular set of dependencies – the fileviewer view and other limitations being one of them (this was filed circa 2002).
1995 called, and they want their software back.
Compared to the current Gnome, XP is kludgy
What does kludgy mean. Specifics are everything here, because people tend to disappear into generalisations.
slow and old fashioned
If you think Gnome is fast compared with Windows XP then I’m sorry, I don’t know where you’re living (there was a demonstrable slowdown when Cairo became a dependency), and in terms of being old fashioned, have a look at the general features that you are lacking.
Of course you’re free to say that most users don’t use this stuff (so that’s OK isn’t it?), but you’ll never sell that to users already sitting on Windows and OS X.
If you think Gnome is fast compared with Windows XP then I’m sorry, I don’t know where you’re living
My system at work and my system at home are broadly comparable. The system at work has a faster CPU than my home system (3.2 GHz P4 compared to a 2.88 Celeron) both have 512 MB of memory and use an Intel chipset motherboard with Intel integrated graphics. OK it is a subjective feeling but the system fells slower and gets worse when memory consumption increases with more applications open. I haven’t tried any objective tests by comparing load times etc for applications that I have running on both (mostly Java together with some Python/tkinter and QT and GTK2 apps) I will.
Of course this is not a totally fair comparison, one needs data from a dual boot system. Since I have no intention in contaminating my system with XP, anyone out their have objective comparative data.
Also XP seems to be less stable. OK not the BSOD’s of NT4 but after a certain amount use the system starts behaving erratically and needs at least logging in an out again if not a total reboot to restore to normal operation. I do not find this with Gnome on Ubuntu.
I don’t know where you’re living (there was a demonstrable slowdown when Cairo became a dependency)
There certainly was a definite slowdown I noticed going from Gnome 1.4 to 2.2. However nowadays each point release of Gnome seems to be getting faster.
Gnome lacks the easy ability to run applications as another user
Try typing man gksu on a *nix to learn how, it is no more difficult than runas from the commandline on XP.
lacks the ability to configure screensavers as on other desktops
By this do you mean other virtual desktops on the same system, in which case the comment is irrelevant as XP does not come with virtual desktops by default (big minus) or on the desktop of another machine over the network. Not having tried this second option I don’t know if it is true or not – though it seems to me it shouldn’t be that difficult.
Finally on my Gnome desktop Compiz fusion is set up to make all the 3D bling work as a practical aid to productivity and it makes even OSX look old fashioned.
Gnome is currently the all round best desktop. I can work productively faster and more effectively in Gnome than on XP for sure, end story.
Edited 2007-12-28 03:24
Edited your post for you, no need to thank me…
[Original parent]
lacks the ability to configure screensavers as on other desktops
[Reply]
By this do you mean other virtual desktops on the same system, in which case the comment is irrelevant as XP does not come with virtual desktops by default (big minus)
I suppose he just meant the ability to configure the settings of a particular screensaver…You know, gnome-screensaver does not allow one to adjust the settings for any specific screensaver. This is off-topic but I’ll just mention that IMHO gnome-screensaver sucks anyway..It has like 3 screensavers: black screen, flying feet and some pop squares changing colors.. “Woohoo”. And well, the screensavers from xscreensaver don’t play well along with gnome-screensaver. Atleast on my machine none of the 3D savers work properly and quite a bit of the others are somehow messed up too.
Maybe true for a beginner, but please…
Gnome is such a poor thing.
Btw., virtual desktops is nothing else than a boss key.
You can have a Windows implementatin with the NVIDIA driver.
To be honest, there’s no need for such a feature in XP.
It maybe more sensefull in Gnome because of Gnome’s very simple start menu. But I’m sure, nearly every user uses it as a boss key replacement.
Compiz Fusion at the moment is nothing else than a desktop game with no practical use.
Even though the article wasn’t very good or helpful, the comments and analysis from the readers were a more interesting read. I learnt quite a bit, thanks guys.
My only problem with Gnome is that it feels sluggish. Other than that it is perfect for me. I love the simplicity and that everything is polished and integrated. Definitely my DE of choice!