Qt 4.0
release is on the wire. Among other things, that means the long awaited
GPL-licensed Windows version is now available for download, and Qt4 is officially dual-licensed across the board. Also of note in Trolltech‘s announcement: integration with Visual Studio .NET; great strides in graphics, threading, footprint, and performance; and separation into Desktop, Desktop Light, and (non-GUI) Console versions.
It’s been a while since I last used QT. It seems that they have improved a lot. I will give it a try to see how it integrates with MS VS .NET
It’s been a while since I last used QT. It seems that they have improved a lot. I will give it a try to see how it integrates with MS VS .NET
It will never beat MS at integrating with MS. However, if you need cross platform development, you can’t beat Qt. Also, if you’re just plain used to C++, it’s a GREAT C++ toolkit.
I really don’t like the new Designer. I much prefer the one from QT3. I know why they changed it, but I think it stinks now.
Can you reason what you don’t like?
QT overall is very cool and fun to use. Can’t please everyone.
I would like to see them on windows…
I’m interested in doing some developement in QT but I don’t have any experience with the IDE. Why do you think 4 stinks?
whadaya reckon?
in time for the SUSE spring 06 release?
He didn’t say Qt stinks, but Designer, which is only a tool for designing the GUI. It got an overhaul in Qt4 so it looks a bit different, I’m guessing that’s what he thinks stinks.
video is hilarious!
So, since when KDE devs have been following distro releasers’ timelines? But well, take a look at the http://developer.kde.org page and you may find something iirc.
Where is it?
> in time for the SUSE spring 06 release?
No. Earliest autumn’06.
This is truely great news. Qt is just a pleasure to work with, and it’s design and architecture is awesome. Singals and slots rule. The docs are unparalleled with any other library/toolkit/whatever I’ve had to work with. I am looking forward to playing around with this.
I know there are a few different and strong opinions on topic of the Windows version being GPL’ed now. Overall, I think it’s a good thing, and should help bring many cool OS apps to windows. Although a few people have complained in #qt “Great, now we’re gonna get all those Win32 mIRC’ers in here”.
On topic of the new Qt Designer, I like it. I think it may take some time getting used to, especially for windows programmers who may not see a distinction between a designer and an IDE. UI design should be separate from implementation. IMO it was a mistake when they introduced the whole “edit in designer” and .ui.h concept. That’s not how Qt is suppose to be programmed. I am glad that’s out.
Just my 2 cents/euroes/whatever.
Bojan
I’m interested in doing some developement in QT but I don’t have any experience with the IDE. Why do you think 4 stinks?
Last time I checked they had change it to a set of floating Windows ala GIMP, which I hate. I much prefer a style similar to the Windows version of Photoshop. Additionally, it used to be a semi-IDE so I did much of my code editing and GUI design from one place. Now you end up having to hope for integration with a seperate IDE and the integrations I’ve seen so far are nowhere near as nice to use the original Designer interface.
Last time I checked they had change it to a set of floating Windows ala GIMP, which I hate. I much prefer a style similar to the Windows version of Photoshop.
I think it’s time to check again
http://qt4.digitalfanatics.org/articles/des2.html
An open source license handicaping an open source project? Dude, we’re writing code for fun and out of passion, not to make some company happy. I don’t want my code to be reverted because Sun or Ximian thinks it doesn’t fit their business model.
Where the f***** is it? The reason why im not programming on linux is because I refuse to waste my time with f****** makefiles and bad IDE’s.
Integration with vs.net would be adding a QT Application option in new project.
You enable the vs.net integration by using the d****** *****. Works perfectly for me.
IS AMAROK for WINDOWS!!!!
real helpful.
I only wish there was an XFCE style lightweight desktop but QT based. It would be a great open source project.
“Also, if you’re just plain used to C++, it’s a GREAT C++ toolkit.”
moc anyone?
[quote]Where the f***** is it? [quote]
Naturally, integration with VS is available for the commercial Qt. The open source Qt works with MinGW under Windows. This isn’t really weird at all, and makes perfect sense.
[QUOTE]The reason why im not programming on linux is because I refuse to waste my time with f****** makefiles and bad IDE’s.
Integration with vs.net would be adding a QT Application option in new project.[/QUOTE]
I really dont see whats so hard about qmake, make. If you absolutly need an IDE, on Linux, KDevelop is a great IDE with many more features than VS. The only area it may lack is in the visual debugger, but this affects different programmers differently, and i a matter of opinion, and personal preference.
Also isn’t your post kind of a contradiction. You basically say that you would be programming on Linux if only Qt integrated nicely into Visual Studio. Doesn’t make much sense to me.
Personally, I don’t think the Qt build process could be any more painless and simpler, even if it is from command line. What kind of programmer isn’t comfortable with command line and makefiles anyway? Just give it a try.
moc anyone?
Yes, if using moc is too much automatism for your, you can work without it.
Actually, as some people really seem to dislike automated parts of their tool chains, I recommend they don’t use the preprozesser either, manually inserting declarations is a lot more direct than those pesky #include markups
>What kind of programmer isn’t comfortable with command line >and makefiles anyway?
15 years ago I was, but now I can’t be bothered with it anymore. I’d rather let the computer to the work for me so I can concentrate on other things. Same goes for GUI work. We have computers to help us, so why not let them?
…some buffoon has not complained about the license yet. Oh that’s right, there’s NOTHING to complain about now.
>>What kind of programmer isn’t comfortable with command line >>and makefiles anyway?
>15 years ago I was, but now I can’t be bothered with it
>anymore. I’d rather let the computer to the work for me so I
>can concentrate on other things. Same goes for GUI work. We
>have computers to help us, so why not let them?
Sometimes, I really wish they would.
I do ASP.NET in VS.NET day in and day out, and we use SourceSafe. I have a number of complaints about the situation: If you use VS.NET to put projects and solutions in SourceSafe, those projects “know” their own SourceSafe information. I don’t mind the redundancy, except for when I actually need to move some code around, rearrance some projects or solutions, etc., and then VS.NET and SourceSafe act like a 5-year-old, running around and screaming, “NOOOOOOO!” This happens a lot, and there are several variations on the exact problem.
And believe it or not, often the problem can only be gracefully solved by restarting.
I’m not kidding about that. It happens when you are trying to share a .NET DLL using VS.NET/SourceSafe, because you need a CERTAIN VERSION of a library, and there is a different version installed on the computer. You can look right at the project file, and the project file says, “THIS IS WHERE THIS LIBRARY IS LOCATED! IT IS HERE! HERE IS WHERE IT IS!” But it will not run, because it is trying to link to the installed version, for reasons which are beyond my understanding. And you restart, and then it will work.
All of that to say, before you say that an IDE or a GUI tool is always better, think carefully about whether you are really getting more work done, or if you are just more comfortable because have to remember hjkl, or Makefile syntax, or whatever.
And I happen to really like Smalltalk, which is IMHO the crowned king of blazingly integrated environments, so don’t claim I’m against IDEs altogether. I’m not.
I wonder when the first comparisons will show up.
Trolltech claims multiple times how Qt4 is more lightweight than Qt3, which is “Contrary to industry tradition”.
I quote:
1. Qt 4 incorporates a wide range of new and updated features and capabilities, increasing developer productivity while improving runtime performance.
2. Qt 4 applications will become faster and use less system resources.
3. … the new version of Qt offers more functionality while reducing memory footprint and resource usage.
How much of this will be true? I can’t wait to see the numbers!
First thing I’ll do is build a minimal Qt3 and a minimal Qt4 application, and compare the memory footprint. Damn I’m so anxious to see how much of an improvement it will prove to be.
>All of that to say, before you say that an IDE or a GUI tool >is always better, think carefully about whether you are >really getting more work done, or if you are just more >comfortable because have to remember hjkl, or Makefile >syntax, or whatever.
I’ve never had any problems like you describe, but then I don’t use versioning as extensively as you do. The use of an IDE with a good GUI designer allowed me to create a viable business, which meant turning out GUI based software on a 6 month time frame (including documentation). If I had to do the work manually it would have taken much longer.
Beware of a character called Goldstein. Mark my words, he will be posting soon. He continually trolls the Qt boards, pretending to be a Qt user and making the following assertions:
1) Qt Designer sucks and isn’t as good as WinForms in .NET.
2) Qt lacks any decent widgets.
3) Qt4 is no good, wait until Qt5.
4) Qt should be written in C# because C++ is dead.
He will offer no evidence to back any of these allegations up, but will continue to re-iterate them, through a number of tedious posts.
As a Qt/VS .NET user, I can testify that 2, 3 and 4 are simply not true. I’ve never used WinForms, so I can’t give a balanced opinion (although I’m sure that the IDE integration is better – obviously). Has anyone used MS WinForms and Qt Designer and can they attest as to which is the better? I am particularly interested in how WinForms widgets copes with dynamic layouts/resizing, varying font sizes, platforms and screen resolutions and i18n of text. Qt has never let me down on these criteria.
The designer in vs.net is fantastic. I did toy around with the 3.x designer a little, just to see how it works and its pretty good. No real complaints at first glance.
When it comes to solving the dynamic resizing problem, Winforms has the best solution anywhere. It has no layout managers in any shape or form. All you do is place all the controls where and how you want them to be, and then gue them to the edge of the dialog with anchors (or not if you dont want the control to change shape). When the dialog gets resized, the controls will resize along with the dialog. You can glue them to any or all 4 sides of the dialog. Simple solution for a pretty small problem (i mean realy, how often do you resize windows?).
>1) Qt Designer sucks and isn’t as good as WinForms in .NET.
I probably have to agree with Mr. Goldstein here, I might not use the word ‘suck’ but I think VS.NET and Delphi are easier to use than the designer that comes with QT. I am sure that if I spent more time with designer I would figure it out but in terms of intuitive use, VS.NET and Delphi are difficult to beat.
How much of this will be true? I can’t wait to see the numbers!
Why don’t you test it yourself? You can use of the example programs which comes with Qt. Some are identical only using different Qt versions, you only have to take the time compiling and running the tests.
Disclaimer: I work for Trolltech.
Re your comment: Empty app, just linking in the Qt library: approx 200KB with Qt 4, approx 900 Kb Qt 3 (internal numbers based on 4.0.0, SUSE 9.2 on AMD 64).
The separate libraries allow more fine-grained control over what you link in.
YMMV of course but our assertions are supported by the facts.
does this mean KDE4 will be blazingly fast
> If I had to do the work manually it would have taken much longer.
How do you know you couldn’t achieve the same thing faster manually, or with a custom code generator? I’m just saying, think carefully about this. But my experience is with the web forms designer and System.Web, not with WinForms.
>I’ve never had any problems like you describe, but then I >don’t use versioning as extensively as you do. The use of an >IDE with a good GUI designer allowed me to create a viable >business, which meant turning out GUI based software on a 6 >month time frame (including documentation).
It could be for a number of reasons. I would say that a very well-organized project does not usually run into most of the problems I mention. But I still believe the IDE should have the potential to do that well, especially when the product costs real, paper money and requires a server platform that costs a lot MORE cold hard currency. I mean, Jesus! $20 billion every year, and they just couldn’t whip up a more forgiving source control scheme.
My thoughts mimic someone else’s thoughts about Java (as compared to Smalltalk): You have to remember a bazillion rules, and if you do something wrong the environment makes you sit in the corner until you fix it. But, with VS.NET, you cannot actually find out what the rules are; you just trip over one now and then.
And then there are some things that just have no explanation in the whole reational ideaspace, such as that the Javascript debugger will not debug into the first line of a function. You can’t do that. No, really, you CANT.
I would certainly rather work in something else, but good luck to you if it works for you.
When it comes to solving the dynamic resizing problem, Winforms has the best solution anywhere. It has no layout managers in any shape or form. All you do is place all the controls where and how you want them to be, and then gue them to the edge of the dialog with anchors (or not if you dont want the control to change shape). When the dialog gets resized, the controls will resize along with the dialog. You can glue them to any or all 4 sides of the dialog. Simple solution for a pretty small problem
Now see I don’t necessarily agree here, what you describe here is what I have observed using Delphi (which I do have a lot of experience with), It was never really good enough. You see the solution and the problem are neither simple nor small. Anchoring to sides of a form falls down where you have a number of widgets that are not really aligned with any side of a form but are instead aligned with other widgets within the form. The solution you have described seems to work for you because you disparage resizing (and the abchor technique works fine for fixed size dialogs).
(i mean realy, how often do you resize windows?).
I resize windows a lot, it’s preferable to extra scrolling. It’s a sad legacy of Win32 that most dialogs do not allow resizing (The MS Outlook ‘Set Names’ dialog is a dire example of this). You will see this behaviour much more on UNIX/Linux platforms though. Usability wise, it’s easier and faster to make the content larger than to have all that extra scrolling on widgets like lists, tables and trees. This is why when people start using Qt GPL on Win32, we will start to see more apps with resizable dialogs.
How does winforms handle different font sizes? Will anchoring help at repositioning widgets/controls?
In the dutch .Net magazine, there was a story about .NET on Linux with mono and gtk#. They were praising the layout system of glade, and even went as far as saying that microsoft would start using a ‘table-like layout system’ ala glade in the future (maybe with xforms? I’m not sure).
I wish I could find the same article on the internet, but I haven’t found it yet. So you’ll just have to believe me
Anyways, I’ve used four methods of managing layouts: winforms, Qt, GTK+ (glade) and swing. It’s hard to say which is best, they all have their own advantages and disadvantages. They all get the job done.
However, if you like quick and dirty, then winforms is the best “solution”. Most Windows developers don’t seem to like resizable windows anyway (they still like to believe that everyone runs a 96dpi screen at 1024×768).
Ehm so you anchor the controls the same way you anchored the controls with which they have to be aligned? I realy cant think of a design for which the anchor system would not work properly. It would have to be a pretty clumzy UI.
If a window has to be resized so that all the controls fit in properly, then the guy who designed the interface is an idiot. By default a window should be exactly the size that it needs to be.
By default a window should be exactly the size that it needs to be.
So which language do you base your fixed size on?
Hopefully no someone with short words and phrases like English.
For every control you can set a font. Anchors will not reposition controls. They only resize controls and maintain the same dialog structure. Why would you need to reposition them anyway?
So here’s the memory usage for there calculator example
reported under /proc (linux kernel 2.6.10):
VmSize: 16440 kB
VmLck: 0 kB
VmRSS: 9092 kB
VmData: 2036 kB
VmStk: 68 kB
VmExe: 44 kB
VmLib: 12444 kB
VmPTE: 28 kB
I don’t have a comparison to qt3, but this still doesn’t seem very “light”. The application requires 16 MB of total memory (physical ram plus swap) and it using 9MB of physical ram.
A simple Xlib program uses:
VmSize: 4380 kB
VmLck: 0 kB
VmRSS: 1636 kB
VmData: 176 kB
VmStk: 56 kB
VmExe: 56 kB
VmLib: 3932 kB
VmPTE: 20 kB.
Ehm so you anchor the controls the same way you anchored the controls with which they have to be aligned? I realy cant think of a design for which the anchor system would not work properly. It would have to be a pretty clumzy UI.
Or maybe you’re failing to see the big picture. You’ve already admitted to not seeing a need for resizable dialogs. I’ll reserve my judgement on anchors until I’ve used them again.
If a window has to be resized so that all the controls fit in properly, then the guy who designed the interface is an idiot. By default a window should be exactly the size that it needs to be.
Please, take a look at the ‘Select Names’ dialog in Microsoft Outlook. Now tell me exactly what size does it need to be? Exactly, it depends on the number of email adresses a person has and maybe even the length of the longest address. There is no way to design a dialog ‘the size that it needs to be’ in this case. Obviously the limiting size is the size of the screen, but to popup a dialog at that size would be too obtrusive to the user. The answer is to display the dialog at a ‘reasonable’ size and allow the user to resize the dialog as he sees fit.
Dynamic content requires a dynamic solution.
So you’re just hard-coding the fonts and strings of every control, ignoring any user preferences regarding language and font size, which you can configure in the display properties window (in the Appearance tab)? Damn you’re a good UI designer. Why didn’t I think of that?
outlook is a big program, I have no idea what dialog you have in mind (and I use outlook express).
I never said that resizing is completely useless and unneeded. Most dialogs do not need to be resized and adding that much complexity to designing a user interface just because of something you more often then not will not have to worry about is dumb.
Nothing keeps you from changing the font at any time while the program is running.
outlook is a big program, I have no idea what dialog you have in mind (and I use outlook express).
It’s exactly the same in Outlook Express. It’s invoked when you choose names from your address book when you edit the ‘To:’ field of a new message (very commonly used really), but never mind.
I never said that resizing is completely useless and unneeded.
And I never claimed that you said that.
Most dialogs do not need to be resized and adding that much complexity to designing a user interface just because of something you more often then not will not have to worry about is dumb.
‘more often then not will not have to worry about’ is just plain shortsighted. Considering this issue is not dumb. It’s complete. It’s the difference between good interface design and playing at it.
Think about this, even when the dialog is not resizable by the user, it still needs have a resizable layout. Other people have tried to hint to you here about the issues of font size, internationalization, screen resolution, etc. But you clearly aren’t coding with that in mind so you just don’t get it. I can show you apps that I have written that will display nicely on three different platforms, regardless of the font size or the language being used. If you can do the same, I’ll happily eat humble pie.
And as regards complexity, that’s my whole point – with Qt it is not complex to do. I wanted to know how easy it was with WinForms – but clearly, you don’t know.
Hi!
Is there a windows binary of GPL Qt anywhere that integrates with VS.NET? I only got some sourcecode, and i don’t have a clue what to do with it. I didn’t find anything in the documentation (I’m blaming my high fever for this though)
Since Qt4 is now intergrated with the .Net platform, maybe it will play well with Mono…
The news posted doesn’t mention it, but you HAVE to see this video: http://www.trolltech.com/video/qt4dance.html
it’s so ridiculous that it’s quite funny
I hope you are aware that .Net != C#. C# are only one programming language usable with the .Net framework, other options are Visual Basic .Net and C++. A Python implementation are also in the works if I’m not mistaken. Since Qt are C++ one should assume it intergrates with .Net from the C++ angle. Making any relation to Mono, a C# implementation, rather farfetched.
> does this mean KDE4 will be blazingly fast
No, it doesn’t because KDE’s slowness and bloat don’t come from Qt, but from kdelibs and kdebase (qt: 9,7 MB; kdelibs: 41,8 MB; kdebase: 54,8 MB).
By the way, why do people always think of KDE when someone talks about Qt? Qt is not designed for KDE (but it’s nice that it is some sort of free advertisement for Qt).
I can’t wait until somebody ports kdevelop over to windows. The free (as in freedom and beer) IDEs on windows just aren’t that good. Mingw + Msys + QT + KDevelop sounds good, and yes I know that by that configuration I could just as well use Linux.
You sir, are a troll.
Since when size = bloat?
So if you have rich documentation with lots of screenshots, increasing the size of kdelibs, then your programm is going to be slow?
If you increase the number of programms, the unmodified ones will run more slowly?
KDE is lean, deal with it. KDE technologies are great. KHTML is the fastest browsing engine available, and lean too.
KDE is memory efficient: see konsole x xterm for one example.
People don’t know what bloat is, that’s why they’ll point to “examples” of it, which really are only making themselves look ignorant.
Oh noes, KDELIBS is teh fat!!!!1111oneone111!!!
Get a clue, all this means is that they’re centralizing and therefore reducing the amount of code. If want HTML rendering in an application, then you use KHTML, it’s in one library, in one place and not rolled out on a per app basis. If you want to utilize the kgethotnew stuff framework same thing. KIO? Well guess what that seamlessly is integrated everywhere thanks to the fact that it’s in the core framework. There are loads of frameworks and it just works, lets not even get into the power of kparts and how much code reuse that leads to.
The KDE project has built up incredible frameworks and centralized them, they may seem like bloat, but they’re actually removing redundancy and adding value.
good news indeed. I will have to wait, though, until a gpl-win pyqt4 is released to actually do any porting with it. Life goes on in kde/qt3 in the meantime. Hope it’s not too long.
Normally when a market expands, prices go down.
The C++ market must be falling apart for Trolltech to jack up prices by $1000/developer/year.
Maintenance is outrageously expensive, pushing $2000/developer/year for a mid-range “2 platform” license when you include the “solutions”.
It puzzles me why a company would get a bunch of VC dollars presumably to expand their market share and then raise prices which will shrink their market.
Companies will charge what the market will bear.
=> More demand = Higher Prices
That is the so called Market in the system we call Capitalism.
New Qt designer sucks
Ain’t that bad, you can do more than the older one, when you get used to it you can even like it more.
[i]You can’t integrated with Visual Studio at least you buy the commercial version[i]
I have no problems with it, Im glad they at least made it GPL for Windows, maybe because they main goal is to make it easier for KDE applications to be ported, they still have to make money, so give the Visual Studio integration for free wouln’t be a smart move from them.
The move of making Qt 4 GPL on Windows comes mainly because some GPL applications that needed to work on Windows too were restricted by the licence so they used GTK instead (inkscape and bittorrent come to my mind), so now there is no excuse for a GPL developer to write its software with Qt anymore.
What TrollTech gets with this?
More developers, and more GPL applications running on Windows hence more “use cases· for Qt and all get translated in interest in Qt of companies that don’t even know it, those companies can pay Qt licese w/o trouble and they will apreciate the framework.
In other words, everybody wins.
Wins KDE, because now its application will be able to run in Windows (Kontact, Konqueror, etc). and users using these applications on Windows wont have problems to migrate to Linux and KDE, because now they will be familiar with the applications.
Wins the GPL developers because now they will get the benefits to port its application to Windows w/o lisences issues w/o the need of GTK of wxWidgets, knowing that Qt offers a better and more integrated framework.
Wins TrollTech because now Qt will be know as a viable multiplatform toolkit by people who don’t even know it and you can translate that in more licenses hence more money.
Wins the user because now they will have more options.
So, we all are winners.
The only think I wonder is what Aaron Aseigo (a KDE developers) think about this, he was the autor of the controvertial article of “How to kill open Source Software” or something like it, wheere he was againts porting oss from Windows to Linux.
New Qt designer sucks
Ain’t that bad, you can do more than the older one, when you get used to it you can even like it more.
You can’t integrated with Visual Studio at least you buy the commercial version
I have no problems with it, Im glad they at least made it GPL for Windows, maybe because they main goal is to make it easier for KDE applications to be ported, they still have to make money, so give the Visual Studio integration for free wouln’t be a smart move from them.
The move of making Qt 4 GPL on Windows comes mainly because some GPL applications that needed to work on Windows too were restricted by the licence so they used GTK instead (inkscape and bittorrent come to my mind), so now there is no excuse for a GPL developer to write its software with Qt anymore.
What TrollTech gets with this?
More developers, and more GPL applications running on Windows hence more “use cases· for Qt and all get translated in interest in Qt of companies that don’t even know it, those companies can pay Qt licese w/o trouble and they will apreciate the framework.
In other words, everybody wins.
Wins KDE, because now its application will be able to run in Windows (Kontact, Konqueror, etc). and users using these applications on Windows wont have problems to migrate to Linux and KDE, because now they will be familiar with the applications.
Wins the GPL developers because now they will get the benefits to port its application to Windows w/o lisences issues w/o the need of GTK of wxWidgets, knowing that Qt offers a better and more integrated framework.
Wins TrollTech because now Qt will be know as a viable multiplatform toolkit by people who don’t even know it and you can translate that in more licenses hence more money.
Wins the user because now they will have more options.
So, we all are winners.
The only think I wonder is what Aaron Aseigo (a KDE developers) think about this, he was the autor of the controvertial article of “How to kill open Source Software” or something like it, wheere he was againts porting oss from Windows to Linux.
Don’t argue with that nutjob, he’s off in some irrational alternative reality where the SUN shines and Linux will die tomorrow.
It’ll lead nowhere.
> You sir, are a troll.
Sorry, no. I’m not. It’s just too easy to call other people trolls if they say something you don’t like!
> So if you have rich documentation with
> lots of screenshots, increasing the size
> of kdelibs, then your programm is going
> to be slow?
I’m not talking about source code size, but about the size of the binaries on disk!
$ rpm -qi kdelibs3 | grep Size
Size : 41822632
$ rpm -qi kdebase3 | grep Size
Size : 54751327
> If you increase the number of programms,
> the unmodified ones will run more slowly?
I don’t increase anything, this is a minimum installation.
> KDE is lean, deal with it.
No, it isn’t. Have you ever started YaST on a SuSE system from outside of KDE? It uses kdesu for that and it is painfully slow and a memory hog because this single mini application named kdesu brings the whole kdelibs up. “kbuildsycoca running…” – for half a minute.
Pure Qt applications can be lightning fast.
> KDE technologies are great.
Maybe they are great, but neither fast nor lean.
> KDE is memory efficient: see konsole x xterm for one example
Yes, but from within KDE only. It’s a pain if kdelibs is not already loaded – “kbuildsycoca running…” – for half a minute.
This is the bloat I mean. I never said that KDE is slow or bad, but it is absolutely not lean.
> Oh noes, KDELIBS is teh fat!!!!1111oneone111!!!
You can talk like this to users, but you won’t get friendly feedback!
> The KDE project has built up incredible
> frameworks and centralized them, they may
> seem like bloat, but they’re actually
> removing redundancy and adding value.
Yes, and the effect of “incredible frameworks” and “removing redundancy” is that a mini-mini applet like kdesu loads Megabytes of stuff into the memory.
Hey, it’s an application that takes my password and passes it to su and it needs Megabytes of libraries for that? Do you really call that “removing redundancy”?
G. W. (IP: —.dip.t-dialin.net):
>I’m not talking about source code size, but about the size of >the binaries on disk!
>$ rpm -qi kdelibs3 | grep Size
>Size : 41822632
>$ rpm -qi kdebase3 | grep Size
>Size : 54751327
Heh, you learn nothing. Get in the clue line.
I’ve just finished dowloading it a while ago and its very nice, but since there are significant changes I don’t know how to go from a GUI made in designer to a working program.
Back with version 3 I would start off by having Qt Designer create a project, then I would use it to make the GUI, and finally I would have Qt Designer generate a main.cpp file. Now I can only make ui files and I’m afraid I don’t know where to go from there, I know how to use qmake to generate project files, but when I try to create a main.cpp file I’m not sure what to put in there. I tried referencing my Qt 3 books, and I’ve made main.cpp files by hand a few times just so I could say I knew how, but it’s different and I’m not sure how to make one that works.
If there is a quick switchover guide for people comming in from using Qt 3 that would be very nice. I don’t write commercially so I don’t spend much time developing programs, and its been a long time since I’ve used Qt because I’ve been busy with adjusting to a new job.
Btw. When is a Qt 4 book expected? I could really use one 🙂
Yes, and the effect of “incredible frameworks” and “removing redundancy” is that a mini-mini applet like kdesu loads Megabytes of stuff into the memory.
Hey, it’s an application that takes my password and passes it to su and it needs Megabytes of libraries for that? Do you really call that “removing redundancy”?
Okay, this is plain silly, you’re loading up a set of base libraries, of course it’s going to take time, usually this happens at the begining after the x server starts and then you don’t have to worry about it. All desktop environments have this problem, gtk or gnome apps in KDE? They take time too.
Besides, nothing in your example says anythng about removing redundancy. In anycase, chances are in KDE4, things will be broken down further so libraries are smaller chunks, your complaints are still ridiculous.
> When is a Qt 4 book expected? I could really use one 🙂
I’m sure that it is being worked on an update of the excellent Qt 3 book.
Why do I have to start up my whole OS when all I want is a web browser?
Have you seen the size of the binaries I have to run? BIOS, OS, graphics, network stuff, device drivers etc.
All that bloat…just for a browser.
That’s comedy gold!
Seriously, it’s so true, people don’t know what bloat is, but it’s fashionable to name call.
how about the customary “when will we get pyqt4” post?
KDE 4 anyone?
Not before fall 2006.