The Qt 4 Resource Center has a look at the demonstrations and examples of Qt 4 RC 1 in a new article sprinkled with screenshots. Showing not only the different applications, but also different modes for filling, aliasing, drawing and transforming, the shots provide a good view of the next generation of Qt.
The Look of Qt 4 RC 1
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Can they get their fonts out of 1993?
Maybe I’ll be interested at that point.
Qt supports AA fonts just fine, it’s just that the author of the article chose not to the use them.
Looks great. I can’t wait to see what this will bring to the KDE desktop.
Best of all, as of QT 4.0 kde apps (and perhaps … the full integrated environment) will no longer be limited to posix compliant os’es. Yes. All kde apps will run on windows, with intact integration, virtual filesystems, scripting options, universal viewer (also known as web browser), printing support and hopefully we’ll convert a few people to linux/freebsd/<insert free OS here> while we’re at it.
Next up : KDevelop for windows !!!!!!!!!!!!!!! An open source integrated development environment with integrated GUI builder, access to and integration with all sorts of open source libraries, CVS and SVN support, …
Having KDE on Windows would be interesting. It would save me a lot of hassle having to run Linux in VMware. Though I fail to see how WinKDE would “convert” anyone to Linux. Most people who run Windows *have* to run Windows due to apps, tools, etc.
Moving more of Linux to Windows is a convenience for developers, nothing more. And quite possibly, will help derail the Linux movement as more Windows developers will see how good their tools are compared to Linux.
Windows is posix complaint, to a degree.
Too bad they don’t have .Net bindings to make KDE/QT apps.
When I look at KDE/QT, I see great technology. Much better quality and superior to GNOME/GTK. KDE/QT is underrated.
There are .NET bindings for Qt and in fact Qt supports .NET integration quite well.
I’m really looking forward to Qt 4.
It would save me a lot of hassle having to run Linux in VMware.
Oh, poor you.
Though I fail to see how WinKDE would “convert” anyone to Linux.
No, I don’t either but at best it’s a strategic move to get people familiar with Qt programming a bit like Firefox on Windows gives other browsers on other platforms a strategic foot-hold. In terms of porting KDE and large-scale applications to Windows, nope, I think it’s absolutely pointless. Effort for nothing.
Most people who run Windows *have* to run Windows due to apps, tools, etc.
I’d agree with you in general (that doesn’t mean it can’t change rapidly though), but it really sounds as though you’re trying to reassure yourself more than anything else with that sentence. I know *a lot* of people who instantly become defensive and do that the very second you mention Linux, open source software, or actually, any bit of software that isn’t pre-fixed by ‘Microsoft’. It’s quite amusing to watch happen if you’ve ever done it.
And quite possibly, will help derail the Linux movement as more Windows developers will see how good their tools are compared to Linux.
Says the person who compared existing Windows development and .Net to a proper C++ development tool done the right way! Just because you’re hanging on to Microsoft’s coat-tails, shaking at the very mention of any alternatives (don’t worry, I know a lot of people like you), it doesn’t make it right you know.
There are .NET bindings for Qt and in fact Qt supports .NET integration quite well.
They have some .Net integration, but it’s hardly full support for writing full .Net/Qt applications. Qt is very much a natively compiled, C++ development tool.
Why they should have .Net bindings and full .Net support, I don’t know. .Net simply isn’t that great (well, maybe it looks that way if you’ve been doing MFC programming for all these years).
@Alex
Actually, has I read it, the QT# binding for .NET are incomplete and need a lot of work.
.NET is the future so might as well get compatible with it early on like Ximian did.
My company sells products to companies, mainly very large corporations, running Windows.
These companies are unlikely to:
1. Switch off Windows anytime soon.
2. Give a shit about open source.
To develop *Windows* software to sell to these big companies, there is nothing on the market today better than Visual Studio and .NET.
If Qt doesn’t go .NET, they will end up missing the boat. C++ is going the way of the dodo for development of end-user apps and when a few things change, operating systems will be written in managed code as well.
Qt’s .NET bindings (like their COM support) are left choking in the dust compared to native .NET/Windows tools. And Qt’s “Designer” is primitive and gross compared to Visual Studio.
For the future of C++, I think Qt is a good solution. It doesn’t do anything in particular really well, but it does do something of everything. Just like C++.
Next up : KDevelop for windows
There are quite a lot of better IDEs available on Windows. Eclipse, Visual Studio, Delphi,…
Though I fail to see how WinKDE would “convert” anyone to Linux.
I for one think too, that KDE on Windows will be positive for Windows – at least at first. On the other hand it may attract more developers. And when “Desktop-Linux” (defined as seemless integration of the underlying system and the desktop software) becomes reality, we may see a huge wave of users switching their os.
Actually, has I read it, the QT# binding for .NET are incomplete and need a lot of work.
Ah, you mean the bindings for Mono/Pnet etc. not .Net. No, it would appear those bindings are defunct simply because I think the developers realised that there was absolutely no benefit at all to them. Higher level bindings like Ruby etc. maybe, but considering what Qt and C++ gives them in terms of applications development I think they’ve found there’s simply no point in using Mono or C#.
In terms of .Net support Qt on Windows gives you it.
These companies are unlikely to:
1. Switch off Windows anytime soon.
2. Give a shit about open source.
Response to 1:
There’s no problem with using Qt for a single platform, even if the single platform is MS Windows. All you gain is the ability to write GUI code with a simple and powerful toolkit that will allow you to be cross-platform compatible if you need that in the future.
Response to 2:
Your customers have no reason to care about open source; however, _you_ do. Open source libraries might fill a gap in your company’s ability or codebase. They can prevent you from spending time reinventing the wheel.
For the rest of the comment:
.NET can stand for a number of things, I’m assuming you’re talking about C# .NET since you voice dislike for C++. Qt Designer is getting a major facelift come Qt 4.0. I agree that Designer for Qt 3.0 is a bit primitive, but it’s not supposed to be an IDE – come Qt 4.0, Trolltech’s taken all project management out of Designer and have instead made Designer work alongside the developer’s IDE of choice.
I have extremely limited experience with C#, but I think C++ will be around for awhile in the desktop application world. I also think that Python, Ruby, and whatever new languages come along will give C# a run for its money.
Oh, Python and Ruby both have Qt bindings too.
Get a grip on reality people, .net hasn’t killed C++ yet and it won’t. Assembly language is still around today, C++, Pascal, Java, Python and several other languages have come around that were easier to use, and those haven’t killed it, its just moved into a niche now and if you ask me its a very healthy niche.
C++ is a great programming language, it may not be as simple as Java or .net based languages but its good none the less. Sure .net and Java will take away some users, but most certainly not all of them, C++ still compiles into native binaries that run faster and it can still produce binaries for far more operating systems than Java and .net can run on.
I started learning how to program with Java, then I had to use Visual Basic and J# in computer science and I still chose C++ as the programming language I want to use, why? Because its still farther developed than either Java or .net, it has more APIs and it has Qt which makes it as simple to use as any corporately developed programming language. C++ isn’t old, its mature; people preaching the death of C++ and the like are no better than people preaching the death of Apple, Linux or Microsoft every time a minor event in IT takes place.
Obviously some people here have proved that they never mature beyond the age old genital measureing contests in which the outspoken members of the Assembly/C/C++ camps say Java/VB/.net programmers are novices who don’t really know or understand programming, and the outspoken members of the Java/VB/.net camps say that Assembly/C/C++ are old and dying languages.
what you forgot it that Kdevelop for windows would mean an alternative to Visual Studio. When windows programmers use the GNU compiler tools conversion to KDE and Linux platforms becomes easier. It will also hurt Microsoft’s business and weaken their development monopoly. while the latter is no end in itself it will weaken the plattform dependencies.
These companies are unlikely to:
1. Switch off Windows anytime soon.
2. Give a shit about open source.
Just as they didn’t give a shit about Excel when they were using Lotus 1-2-3 or Visual Basic when they were using Turbo Pascal. You’re not doing yourself any favours.
To develop *Windows* software to sell to these big companies, there is nothing on the market today better than Visual Studio and .NET.
Considering they’re taking all of the best features from tools like Eclipse and transplanting them in to VS, I’d beg to differ. Oh yer – you can easily create Windows applications with Eclipse and Java and still have Windows support. Many large companies do.
And if you’re privileged enough to *sell* software to these companies, then quite frankly, they don’t care what it’s written in. If they get a nice installer, click next, next and then finish and they can run it then they don’t care.
If Qt doesn’t go .NET, they will end up missing the boat.
Miss what boat? HMS Hype? Companies want applications that run.
C++ is going the way of the dodo for development of end-user apps and when a few things change, operating systems will be written in managed code as well.
Would you care to tell me how many Microsoft applications are being re-written with .Net? Nope, they’re all C++ and native. I’d quit reading the MSDN blogs and those silly Microsoft Certified Partner magazines that do the rounds to work out what’s really happening.
Qt’s .NET bindings (like their COM support) are left choking in the dust compared to native .NET/Windows tools.
Yawn. You can quite easily communicate with and use .Net/COM components with Qt on Windows, fully supported by Trolltech.
And Qt’s “Designer” is primitive and gross compared to Visual Studio.
Yaawwwnnn. So I take it you’ve used Qt Designer then? Care to tell me when Visual Studio is going to get half-decent layout manager support so no one needs to spend all week re-arranging widgets?
For the future of C++, I think Qt is a good solution. It doesn’t do anything in particular really well, but it does do something of everything. Just like C++.
Well, it’s a nice try, but really, the desperation is self-evident. If you like Visual Studio and .Net, think it’s the greatest thing since sliced bread (finger down throat) and you feel you *have* to run Windows (and hence, everything produced by Microsoft) then just stick to it. There’s simply no reason to go off on one about a development tool that you’ve obviously never used but seem to be quite worried about.
.NET is not the future of anything. It’s a good language, and a good platform, but when the hype dies (which it basically has by now), it’s basically just a slightly improved Java. Java is used a lot, but it isn’t replacing the traditional native end user applications at all.
No matter how you spin it, virtual machine languages like C# and Java will always be slower and more memory hungry than native apps. You might say that machines will get faster to compensate, but we’re still not there yet! How fast do they have to be?
Think of the large scale Java or .NET apps that you have used.. They’re all annoyingly slow, even on modern hardware.
There’s a reason why Microsoft backed off from .NET for Longhorn. It’s just too damn slow.
Those type of languages have a place, but they’re definitely not the be all, end all of computing.
how great technology QT is. And new QT 4 is a big step forward for creating a completely different and improved UI.
All one has to do is nest a few templates and you will be in love with C++ — forever. All the screaming joy of C++ will be echoing around the room.
I will laugh at the C++ jockers who think the language is not already fading away. Oh, it will never fade away 100%. Kinda like Fortran, Lisp, Cobol, Pascal, etc. It will still be around for the jockers that really need their fix of “object construction kit”.
But most developers will have moved on to higher productivity tools.
As I have said elsewhere, Qt is a good C++ framework. But it takes much time and many rewrites to build a good C++ framework. And the pace of development of Qt is far too slow for most markets.
In the most optimistic of dream worlds, Trolltech will allocate some money from their recent funding to build a modern widget library for Qt. If the widgets were better and the API were cleaned up to be all Java-clone or all C++-native, then Qt would stay around longer. In any other dream world, Qt follows C++’s gentle fade to black.
>Yaawwwnnn. So I take it you’ve used Qt Designer then? Care to >tell me when Visual Studio is going to get half-decent layout >manager support so no one needs to spend all week >re-arranging widgets?
Delphi and it’s copy VS.NET is far far more productive at producing good looking GUIs compared to anything else. I hope neither will ever get layout managers. Anchors and aligners plus panels is all one needs. If you don’t believe me check out all the badly layed out GUI Java apps. If you really spend that much time layout out controls then you’re doing something badly wrong.
Damn, someone with real world experience using tools that are widely recognized as the best there is (Delphi, VS.NET). Someone who is not high on the ivory tower UI opium you find in Java and to some extent in Qt.
At least with respect to Qt 3.X, there was no way to make a good UI. “From code” worked better than the retarded designer that could not even generate a proper UI class. The developer had to derive another class to be the real UI class. That the “designer” (aka “the time waster”) could ship with the basic design so flawed amazed me. Really amazed me.
Maybe Qt4’s designer is better. But if you are still putting those damn springs all over the place and dealing with a “layout manager”… skip it.
I say again, WAIT FOR QT5!
heh, when I implement a widget in Qt, I have to set two properties and my widget is behaving nicely in layouts. In .net, I have to implement the layouting logic in each and every widget I implement. Qt even handles height-for-width correctly (for example a toolbar that can either be wide or heigh to display its contents).
I agree that .NET gives you the better first impression, but when you actually program stuff, you’ll find out that a lot of things are solved more elegantly elsewhere and that your productivity gain will drop fairly soon.
And who in his right mindset invests money in a cross-platform .NET application? Mono is lagging and will always be lagging behind the newest .NET features, it doesn’t give you all the powerful APIs (and it’s legally questionable whether they ever will) and the major Linux distributor doesn’t even ship Mono due to licensing fears. Show me the manager who takes that risk, please.
I always see people posting about Eclipse and Delphi being such fantastic IDEs for programming. I agree for Java and Pascal, but did anyone ever seriously use the C++ part? When we hire C++ programmers and they claim they use Eclipse for their C++ work, they usually failed the idiot test.
Delphi is wonderful for Windows only apps in Pascal, but the C++-builder part is unusable due to the compiler not being able to parse standard-C++. I don’t want to pray everytime I write a template in hope that the compiler will get it this time. Besides, where’s C++BuilderX? Where’s the compiler they promised us 5 years ago?
Delphi and it’s copy VS.NET is far far more productive at producing good looking GUIs compared to anything else. I hope neither will ever get layout managers. Anchors and aligners plus panels is all one needs. If you don’t believe me check out all the badly layed out GUI Java apps.
When you actually know what a layout manager is and what it’s supposed to do give us a call. Even better, actually try using Qt’s, because that’s what the article is about!? Did you read the bit about .Net (and others) people spending five workdays plus on re-arranging user interfaces? I’ve seen it happen, and that’s the alternative.
Anchors, aligners, panels… Bugger. Do you like spending a week plus trying to work how to position all you widgets? Ah, but of course, that’s the Visual Basic upbringing.
Someone who is not high on the ivory tower UI opium you find in Java and to some extent in Qt.
.Net ivory tower alert!!
At least with respect to Qt 3.X, there was no way to make a good UI. “From code” worked better than the retarded designer that could not even generate a proper UI class. The developer had to derive another class to be the real UI class. That the “designer” (aka “the time waster”) could ship with the basic design so flawed amazed me. Really amazed me.
Ah, that must be that real-world experience talking again. It should be no coincidence that anyone who’s ever used Qt is wondering what on Earth you’re talking about. No shock there then.
I say again, WAIT FOR QT5!
You want people to wait for Qt 5.
I have to ask the question – are you getting paid by Microsoft for writing this drivel, because that “Amazed me. Really amazed me” bit really smacks of the way these people generally talk?
.NET isn’t likely to kill any kind of language. It will simply provide a whole new paradigm on how you build your applications.
Regarding matters of performance, I truly believe that a platform like .NET may not achieve the same level of performance, but that it will become good enough. Add to that the fact that no more “Memory leak fixed, now the application should not segfault and it will consume about 20% less memory” type of bugs won’t exist, and they may actually make non-optimized applications perform much better.
A managed platform, which will simply make the “bindings” work completely obsolete, allowing people to actually program on their favourite language is key to increase developers productivity and is bound to take away a huge developer share. We’re in the XXI century, and I think it’s about time we develop our applications accordingly.
You state that you went for C++ because of, and I quote, “it has more APIs and it has Qt which makes it as simple to use”. Well, if more and more development goes into .NET, the language you actually use rapidly becomes a non-issue, because whichever you choose, it will always have the same APIs and any graphical toolkits you might be interested in. With .NET, you don’t pick a language because of anything else besides the way you feel about it.
To finalize, it’s not the languages that will die. It’s the way you develop your applications. Let’s use a metaphor to make it easier to understand my point: the Internet didn’t kill french, german, english, or any language. If anything, it made them even more widespread. But the Internet changed the way people communicate with one another, the same way .NET will change the way application development is done.
Go Mono!
All one has to do is nest a few templates and you will be in love with C++ — forever. All the screaming joy of C++ will be echoing around the room.
You probably meant this to be some sort of joke, but you are actually right on the spot. Nesting templates in C++ allows one to write flexible and elegant code. I would certainly miss this feature if I had to use a language without a similar concept.
Btw., as you seem to be a fervent supporter of .NET, and at the same time agitating against C++, do you actually know that even Herb Sutter (whom you should know if you admire .NET) says that C++ is still the mightiest language for .NET development (and not C# for example)?
I will laugh at the C++ jockers who think the language is not already fading away. Oh, it will never fade away 100%. Kinda like Fortran, […]
Well, go to the people who do climatic calculations on multi-million-dollar expensive super computers, and tell them they should abandon Fortran development, instead using some of your “higher productivity” languages… I leave it to the astute reader to figure who is laughing at whom then.
(I could add that C++ can under special circumstances perform as well as Fortran for number-crunching)
“2. Give a shit about open source.”
If it can save them money they will do so. You can’t beat free (as in beer).
Why do you care about Qt if .NET and Visual Studio satisfies your requirements?
Why must you continually toot praise about MS based solutions?
Why does anyone give a shit of what you say?
We use what we use, you use what you use.
Stay on your side of the f*cking fence and we’ll stay on ours.
Agreed, Goldstein needs to real his head in instead of continually patrolling any and every Qt-related board with his tired inane crap about C++ being a dying language, Qt being no-good until Qt5 (what?) and lacking any decent widgets (still no examples cited, after being asked many times over).
It’s really quite tiresome, not to mention unsubstantiated bullsh$t.