Trolltech has released the first Beta version of the upcoming Qt 4.0. The final version is expected to be released in late first quarter of 2005. An updated online Qt Reference Documentation is also available.
Trolltech has released the first Beta version of the upcoming Qt 4.0. The final version is expected to be released in late first quarter of 2005. An updated online Qt Reference Documentation is also available.
No windows version on ftp.
There is no Windows version of the Open Source edition.
This means that we can’t really expect KDE 4.0 until mid-late 2006, which is a shame. I guess we will just have to wait.
KDE 3.4 promises plenty of great improvements. I just can’t wait for the improvements to the overall codebase once the porting to 3.4 is finished.
and what will those improvements be?
http://developer.kde.org/development-versions/kde-3.4-features.html
There is no Windows version of the Open Source edition.
It looks like Trolltech doesn’t care about crossplatfrom open source anymore. Luckily, Gtk+ looks great on windows these days.
Merry Christmas
Is it true that Qt 4 uses double buffering?
Yay…everyone going slower and slower and slower…:P
This version of QT looks extremely promising, the painting engine looks brilliant and although it is buggy the new TextEdit widget with Html import/export is a big improvement. There are several new container classes, and double-buffering is enabled by default.
What I currently don’t like is the new ToolBar layout on the mainwindow, the icons are huge on the Apple Mac and it doesn’t seem possible to reduce the size of the toolbars, have a look at Qt Assistant or try making your own toolbars and you will see what I mean. ( I Hope they change or fix this). They look fine on Linux and Windows.
QT 4 is also now completely multi-threaded which means you can use Signals and Slots across threads and QTcpSocket and QServerSocket can be used across threads. This is going to make life so much easier for developers.
Luckily, they have left a few QT3 compatiblity classes in so that you can gradually port QT3 applications into QT4, although this looks like it is going to be a monumental task for KDE.
There is a Beta of QT 4 for Windows, but unfortunatenly only customers can download it.
No windows version on ftp.
—-
who said there would be?
OpenGL backend?
Luckily, Gtk+ looks great on windows these days.
Yeah, everyone’s using it =)
Oh wait..
Why would it take to 2006? They have been working with Trolltech very closely to get QT 4.0 out the door. In other words, they have been writing code for CVS QT 4.0 for some time now and have access to an alpha now. I might agree with you if you said a “stable” KDE 4.0 for 2006. That seems reasonable.
3.4 is going to be groundbreaking, I think, in its sheer polish. Hopefully I’m not wrong there. 3.3.2 is already more than good enough for anything I do.
Yeah, everyone’s using it =)
Oh wait..
—
gaim, gimp, inkscape, abiword?
I meant commercial companies who do cross-platform development.
Some aplications that are using the GTK api on linux are using the win32 api on windows… e.g Firefox
I geet the ‘feeling’ that Firefox is faster using the win32 api on win32 than using GTK on win32…. dunno why
Yes, Firefox is faster on Windows than it is on Linux, but that’s not nessecarily GTK’s fault. I’m sure someone has a more scientific explanation for this issue.
Lumbergh, your kidding right, GTK+ on Windows is much buggier and always out of date. Where is GTK+ 2.6 on Windows. Trolltech does not provide an open source version on Windows due to piracy. but they do have evaluation editions and the free edition.
BTW: Qt 4 is looking great.
I don’t think that KDE 4 will be that late. Note that KDE team tries to keep “in sync” as much as possible with Qt. E.g, the latest Qt release is 3.3.something and the latest KDE is 3.3.2. When KDE team releases KDE 3.4, there will probably be a QT 3.4.something too. So, my guess is that KDE 4 release date won’t be too different from QT 4.
Why does every little bit of news about KDE/Qt always have to be full of people who have to try to dig up some FUD? If you don’t like/use KDE or Qt, don’t whine about them, be happy with whatever you happen to use and let people enjoy good KDE/Qt news.
If i recall correctly… the windows versions appear after the others are released… i wouldn’t expect that there would be one before that happens…
But one never knows…
There will be no Qt3.4 – yes it is breaking of ‘version compatibility’ between Qt and KDE but developers of KDE decided pause will be too long between KDE 3.3 and KDE4 (which is planned on fall of 2005).
If i recall correctly… the windows versions appear after the others are released… i wouldn’t expect that there would be one before that happens…
The Windows version is non-free. Although there’s a Cygwin port at kde-cygwin.sf.net, it misses the native features Lumbergh and others want. But the KDE people don’t care Windows or compatibility with Windows on this layer, because its not in their interests. Trolltech’s interests in Windows are solely commercial.
Compare that with the more liberal LPGLed GTK from e.g. the point of a Windows developer or a Windows/FOSS developer and you got Lumbergh’s point.
@Alex
Lumbergh, your kidding right, GTK+ on Windows is much buggier and always out of date.
As of the newest version of GTK, the WIMP engine and other things are now part of the Official GTK distribution. This means that the Windows versions will soon be more frequently released. Last, “much buggier” is subjective.
Yes, Firefox is faster on Windows than it is on Linux, but that’s not nessecarily GTK’s fault. I’m sure someone has a more scientific explanation for this issue.
yeah i’ve been pondering this point too, what is the reason its faster on windows? is it something to do with linux kernel, kde, GTK or just linux firefox? any one have a good reason for this ?
I also don’t think that it would take that long… kde 3.4 will have some really nice features and as I see it kde 4 will be a port of the kde 3.4 code base to qt4, maybe with some more feautures… and I’m sure that everybody (wanting a stable and good in shape kde4) will help with the porting. I really can’t wait to see kwin ported to cairo (which will be supported by qt4) w/ the composite extension (hopefully more stable)…
kudos Trolltech and KDE team!
<<your kidding right, GTK+ on Windows is much buggier and always out of date.>>
Bah this is FUD. Ever since GTK+ 2.4, Windows support has been superb. I’d rather use GTK+ for crossplatform work, than shell out $1500+ to some dinky company in Norway.
Gnome has already overtaken KDE, their most recent release has native support for composite, and because of it, it has true transparency.
Gnome also has HAL and D-BUS support. KDE if it wants to compete should not wait until Qt 4 is released to integrate these features.
I am a long time KDE user, and I enjoy its non spartan gui.
>Gnome also has HAL and D-BUS support. KDE if it wants to compete
>should not wait until Qt 4 is released to integrate these features.
Kde already have dcop, just do the job for now! The current integration of dbus in gnome is ridiculous in comparaison of what is doing dcop in kde
For HAL, i have read somewhere that someone is working on a kio plugin for removable devices.
Kde cvs support composite.
Moving from dbus to dcop is a hard work for kde devel, not the same probleme for gnome because there is no ipc system for now in gnome. But for now, gnome applications aren’t using dbus.
I just right click a window and select the transparany level. I also don’t think you could justify any statement lixe that by listing one aesthetical feature.
——–
Gnome has already overtaken KDE, their most recent release has native support for composite, and because of it, it has true transparency.
Gnome also has HAL and D-BUS support. KDE if it wants to compete should not wait until Qt 4 is released to integrate these features.
———
I don’t think so. Composite lacks driver support and is still considered experimental in X.org 6.8, if I am not wrong, it’s not even enabled by default.
HAL and D-Bus are also inmmature technologies, they need polish and I’m confident KDE will do adopt them when they’re ready.
I don’t know how you think Windows support has been so suberb, it has always been one release behind for months. Also in Gaim’s sys tray for example, menu options are not even highlighted and it seems even slower than on Linux.
<<I don’t know how you think Windows support has been so suberb, it has always been one release behind for months.>>
More FUD.
GTK+ 2.4.14 released
http://mail.gnome.org/archives/gnome-announce-list/2004-December/ms…
Dec 03, 2004
GTK+ 2.4.14 Windows Installer
http://sourceforge.net/project/showfiles.php?group_id=121075&packag…
Dec 18, 2004
<<Also in Gaim’s sys tray for example, menu options are not even highlighted and it seems even slower than on Linux.>>
Works in Gaim 1.1.0 // WinXP Pro. Also your gauging the quality of a toolkit by the quality of one application? Wow! *claps*
plus wimp from gladewin32.sourceforge.net works great, looks great, and is all around excellent crossplatform development environment.. 2.6 just came out a few days ago. the 2.4 packages were excellent.
HAL and D-Bus are also inmmature technologies, they need polish and I’m confident KDE will do adopt them when they’re ready.
I’ve hear that before, if GNOME implements a tecnology before KDE devs say “It is to inmature to implemented”, translation (We dind’t have that idea, lets copy it in our next release)
Examples:
SVG icons, GSTREAMER and now D-BUS.
I thing their K-Ego is to much.
Since dbus is heavily inspired by KDE’s DCOP system, your comment
>Gnome also has HAL and D-BUS support. KDE if it wants to compete
is quite ridiculous. And of the two, currently DCOP is more mature.
currently DCOP is more mature.
Well, sounds good on theory, but when plug something on GNOME it detects it, KDE don’t, simple.
Users want actions, no words.
>> currently DCOP is more mature.
> Well, sounds good on theory, but when plug something on GNOME it detects it, KDE don’t, simple.
> Users want actions, no words.
What are you talking about? DCOP is an IPC mechanism and doesn’t necessarily have anything to do with plugging in devices. Users (like myself) appreciate the scripting that can be done with it. Developers appreciate the communication between apps that it makes possible.
38 posts in a thread about QT, and the majority of them stemming from a couple of gnome trolls who try to compensate for something by attacking KDE.. Welcome to osnews..
Unfortunately QT Designer for QT 4 is still miles behind the competition which means that GUI development is still more tedious on QT than it should be. Hopefully they’ll improve the designer in future releases. QT 4 however looks like a major improvement.
Look in the future KDE will be using D-BUS, HAL and composite. I am stating gnome has this already in their latest release, KDE does not. They obviously thought it was mature enough to be integrated, the gnome folks that is.
One poster said composite needed its own custom drivers, which it does not.
I’ve hear that before, if GNOME implements a tecnology before KDE devs say “It is to inmature to implemented”, translation (We dind’t have that idea, lets copy it in our next release)
—–
D-BUS is mostly a copy of DCOP, because DCOP works so well.
DCOP was in KDE four years ago.
enough said.
“GTK+ 2.4.14 released
http://mail.gnome.org/archives/gnome-announce-list/2004-December/ms…..
Dec 03, 2004
GTK+ 2.4.14 Windows Installer
http://sourceforge.net/project/showfiles.php?group_id=121075&pa…
Dec 18, 2004 ”
Check the dates, the Windows version of 2.4.14 was released a ew weeks after and GTK+ 2.6 is still nowhere to be found.
“Works in Gaim 1.1.0 // WinXP Pro. Also your gauging the quality of a toolkit by the quality of one application? Wow! *claps*”
No, I am not, it was simply an example. Also, we were never judging the quality of the toolkit, rather the support for Windows vs Linux for GTK+.
Can’t you just be glad we have a choice on the Linux desktop?
<<Check the dates, the Windows version of 2.4.14 was released a ew weeks after>>
And you said:
<QUOTE>
I don’t know how you think Windows support has been so suberb, it has always been one release behind for months.
</QUOTE>
So now your contradicting your own argument.
<<GTK+ 2.6 is still nowhere to be found. >>
GTK+ 2.6 was released 6 days ago. As I already pointed out, GTK+ win32 builds are not released for a good 2 weeks after the offical announcement. Don’t KDE users care about stability and quality? I guess not, if your demanding such an unreasonable timeframe.
<<Also, we were never judging the quality of the toolkit, rather the support for Windows vs Linux for GTK+.>>
And again, you judge the quality of support by the performance of a single product, which BTW I mentioned was working properly on my WinXP system. Your going to have to do a better job to prove to me that GTK+ win32 is not up to par. It’s solid.
Looks like they’ve solved the text alignment problems in controls! Text is now properly centered in QPushButtons on OS X. They’ve also updated the toolbars to be more OS X like. Nice!
Now all they need to do is to update the way tabs are drawn and you’ll not be able to tell a Qt app from a native Cocoa app.
Why should KDE developers need in hurry to implement DBUS, when DCOP already serve nicely for years? Think as KDE devs, DBUS is not light years ahead than DCOP, is it? Implementation of DBUS is for compatibility with freedesktop. The need of DBUS is not that great. How could Gnome is better than KDE in this matter? I mean about the applied tech not its name.
“Gnome has already overtaken KDE, their most recent release has native support for composite, and because of it, it has true transparency.”
FUD. This is already in KDE CVS
“Gnome also has HAL and D-BUS support. KDE if it wants to compete should not wait until Qt 4 is released to integrate these features.”
FUD. This is already in KDE CVS
KDE CVS will obviously become KDE 3.4
And in what way is GUI development ‘tedious’ with Qt designer?
The way you have to lay out widgets? Qt is much better that any competing product in that sense.
Connecting GUI to code? Well, subclassing might not be the ideal solution, but it’s not tedious. I do like the libglade approach more though.
I don’t know any competitors that do this better. Java is horrible because it mangles GUI code inside a method and has horrible layouts (it took me a long time to figure out how to work with those). .NET now has partial classes (works kind of (but not really) similar to subclassing), but doesn’t have layout managers so it’s pretty much worthless.
Glade on the other hand is really nice, although the learning curve is more steep than Qt.
XAML is supposed to be like Glade afaik, but I haven’t seen it yet so I can’t comment on it.
Anyway, I’d say Qt does a REALLY good job, and leaves its competitors far behind. Saying otherwise is spreading the word of FUD
gtk is very well designed, but from a technology point of view it is behind qt. qt is not totally propriatary, but it is created for the purpose of making money. NVu was the first qt app i tried in windows (wanted to try it here at work) and i was extremely impressed. i use gaim in windows cause of the familiarity, and you still have alot of the problems with gtk, just more evident (like slow redraws) when running beside something more snappy. i love gnome, but because of the solid design, not cause of gtk.
DCOM is nice, but it is tied to qt. D-BUS is a little nicer, and is designed to have multiple interfaces. we are talking about everyone playing nice together in the linux desktop world. as for not having HAL cause it isnt mature is more then a bit of a cop out. kde doesnt have hal cause they havnt gotten around to it yet, no other reason. when it comes to the linux desktop, we are playing catch-up still with the competition, bleeding edge stuff is included asap.
kde needs work on the design side, they are extremely strong from the technology side. gnome needs work on the tech side, they are very strong already on the design side. one or the other being “better” is a really relative thing.
one last thing. how about just for once, for a change of pace, we dont let the dm/os/distro trolls hijack the conversation. just once, thats all i ask 😉
> NVu was the first qt app i tried in windows
NVu is no Qt based application.
> DCOM is nice, but it is tied to qt
DCOM?!
Personally I think GIMP on windows is terrible. Crashes all the time for me.
Abiword is not a GTK+ app on windows, unless something has changes recently. Abiword is also unstable on Windows.
Personally I hate GTK+ apps on Windows they feel ugly and unprofessional.
My experience tells me QT apps on Windows are fast and beautiful.
I have a amd 2800 + with ATI radeon 9200. I got Xorg 6.8.1 running on Gentoo. glxgears does around 3000 fps.
Now I went through a gentoo-wiki and got composite with shadow and transparency running. I ran it first in gnome 2.8.2 and then in kde 3.3.1.
I must say Composite works much better in KDE. Firstable before start using any of the flashy stuff gnome is going 10 times slower when composite is enabled in the xorg.conf.
Secondable the real transparency is still immature because it makes the window really slow.
In KDE things go as smooth as always with shadowing no problem but if you turn on the transparency the window that is transparent will be slow. also if you do any window animation they will be really slow too.
So, as far as I can tell QT is doing a better job than GTK for now.
KDE CVS is not stable yet. Gnome is.
yeah i’ve been pondering this point too, what is the reason its faster on windows? is it something to do with linux kernel, kde, GTK or just linux firefox? any one have a good reason for this ?
It could simply be the fact that the Windows version gets a lot more testing and thus maybe more performance fixes. It’s hard to tell.
Oh dear. “Let’s pan Qt with some non-existant and non-sensical remarks about it and GTK (even though Qt is much more than GTK)!” Please – GTK on Windows, a Mac (or even Unix) is just a joke from a commercial (especially died-in-the-wool Windows) developer’s perspective. Yes, I’m sure the screenshots look lovely, but developers out in the world have to be able to use it. That’s the problem with “oh, look at the screenshots!” and almost working – it’s close, but there’s no cigar.
There is nothing wrong with GTK in an open source sense. Gnome is based on it and it mostly works for what people want it to do along with the programming environments on top in Mono/Java/Python et al (GTK itself isn’t enough). However, in terms of getting developers outside of the usual open source forums like this one to use it, all of this technology fails absolutely miserably. The real-world Windows developer view of Mono when they’ve tried it is “It doesn’t work”. Go on a .Net training course – trust me.
In view of what’s on offer people will pay for good development tools – and even if everything is free people will still pay several thousands for IDEs. That’s the way these people work:
I’d rather use GTK+ for crossplatform work, than shell out $1500+ to some dinky company in Norway.
Then you’ll spend two, three, four times as much getting it to work, and at the end anybody who is foolish enough to have hired you will have an application based on something that just plain doesn’t work or perform acceptably. In the real world people pay for development tools – OK? In the real world people also pay vast sums for office suites and other things, but that’s another topic. You may want to screw your career, but those out there actually doing it don’t.
As Microsoft has proved, having a good desktop environment all stems from programming and development. Gnome and GTK along with any touted additional stuff like Mono or Java, whatever their features, are simply not up to snuff to sell to the wider world at the desktop end. Qt isn’t perfect, and even with 4.0 it will still need further development, but it is so far ahead of GTK, Mono, Java or any other Gnome-oriented development technologies in terms of completeness and things actually just plain working it isn’t believable.
Qt 4 will have solid commercial grade support for DBUS and AT-SPI, and Trolltech have put their money where their mouth is on that subject. DCOP can’t possibly disappear because it works and KDE has such extensive support for it, but DBUS is interesting because it provides the possibility to have system-wide communication outside of KDE and Gnome that we haven’t had before. With full DBUS support in Qt 4 KDE developers will be able to work fully with DBUS on a solid foundation. For all those excited by DBUS and HAL, yes they look promising but at the moment they are just not mature enough because they are so new. KDE and Qt will support them fully at a time when they are ready.
Gnome can’t ditch Bonobo and all the other stuff fully for DBUS overnight either because so much depends on that infrastructure. So no, Gnome does not have full support for DBUS. Only KDE is talking about moving from DCOP and possibly moving to DBUS fully, interestingly enough, and they will have a solid development platform on which to do so.
I’ve hear that before, if GNOME implements a tecnology before KDE devs say “It is to inmature to implemented”, translation (We dind’t have that idea, lets copy it in our next release)
If GStreamer was called KStreamer, in all honesty, would the Gnome people be talking about using it as much as the KDE people have looked at and developed with GStreamer? Errr, no
.
you’d rather spend your effort on cooperating – i don’t think gnome and kde devs have time to fool around like what i’ve read in (some of) the past 57 posts. LSB is a start but before that you’d better grow up and stop FUDing around
my 20c, absolutely off-topic (from point of view of the original topic)
If GStreamer was called KStreamer, in all honesty, would the Gnome people be talking about using it as much as the KDE people have looked at and developed with GStreamer? Errr, no
.
I you are dumb enought to judge by a simple letter, then please, keep your opinions.
I you are dumb enought to judge by a simple letter, then please, keep your opinions.
Nope, but I know a group of people who would be – which incidentally was the whole point of the comment.
Anyone who wants to know why Qt is the only viable option in terms of alternative desktop development, look at the technology at the top of this page:
http://dot.kde.org/1103717662/
It isn’t perfect, and they have some way to go, but the quality and thought in these features is pushing things in the right direction.
Tulip:
Java style iterators are what developers would want to see with any development alternatives, it is what they will be used to if they have developed for Windows and they just plain make sense – it’s much more readable in plain English. Certainly not ground-breaking, but a step forward for a C++ toolkit.
Interview:
Being able to really program core components and create different views trivially from it is a very powerful feature. It is something that Microsoft has realised fairly well, especially with things like ASP.NET, although .Net really depends on Visual Studio doing things for you rather than fully integrated .Net support to make it all work. This is probably one of the best things to be added to Qt, because it isn’t clear what benefit it can bring unless you start doing things this way.
Arthur:
Again, nothing ground-breaking, but there is solid improvements here in being able to take advantage of local system components transparently. In terms of writing less code on different platforms and getting it to work natively and faster, that can only be a good thing.
Scribe:
Better and improved text handling is obviously a plus for any developer. The most interesting part of this is the further work at the bottom of the page for multiple views on to documents. This obviously continues the theme from Interview. They have improved rich text support described, but I think they need to start thinking about support for other things. If you’re going to promote open standards then put your money where your mouth is and provide full support for the Open Office XML document format. This is something that Trolltech needs to do a bit better in terms of thinking about what they are promoting and how they back that up with actual support. They’re doing it with DBUS and the accessibility framework.
SQL Module:
Obviously Interview has had an impact here, as you want to trivially create different views from the same data. Having integral model and view creation in the toolkit as they have with Interview is very intuitive, and I’m quite impressed. Not having to go off and do this yourself, or depend exclusively on an IDE to do it for you is a very good thing. The XML and SOAP module looks good, as support for this and web services is something they needed to do.
All in all it looks like a complete and very powerful package, and they’ve obviously put some serious thought into the quality of the implementation of these new features.
I have a potentially major project next year, and as we will be looking seriously into none Windows-based terminals for our client implementation, Qt 4 is strongly in the running.
Nope, but I know a group of people who would be – which incidentally was the whole point of the comment.
Those people are known as “TROLLS”, they don’t help anybody, not even KDE.
…but I won’t let that derail my response
Most of what I’m waiting to see will be in their first Non-Commercial release for Windows. I’m already developing two applications which make extensive use of QT4 features, namely signals/slots across threads.
I am certainly on the edge of my seat to see Picasso on Windows. I see people here are already complaining that it’s “double buffered”, which somewhat misconstrues its design and varies from platform to platform. Regardless, from what I’ve seen Picasso certainly isn’t slower than the rendering engine in Qt3/X11, and seems to perhaps be substantially faster for certain widgets, namely those backed by their new abstract data type classes… lists/tree views/directory views have improved dramatically.
I could respond to some of the trolls, but that’d be rather pointless. I think the only befitting response I can make is to simply say GTK sucks, Gnome sucks, get a life and stop trolling the Qt threads, some of us would actually like to discuss Qt4 without having to deal with horrible flamewars started by zealots with nothing better to do…
Isnt this Xmas ??
not even on xmas ?
when will there be peace ?
QT4 looks very promising .
i would say that linux desktop looks very promising for next year
also … even with qt3 and kde 3.4 … linux has a least a very nice desktop.