Trolltech has announced the release of Qt 3.3.4 which is a maintenance release including many bug fixes and optimizations. The final Qt 4.0 release is now anticipated for late second quarter of 2005 with two interim beta versions to be released in March and May.
Qt 3.3.4 Release and Qt 4 Schedule Adjustment
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
45 Comments
Looks like the guys at Trolltech are correct when they say that there isn’t an API to draw proper tabs in Panther. It seems to be a problem with the DrawThemeTab function that’s provided by the Appearance Manager. Apple engineers conveniently ‘forgot’ to update that function for drawing panther tabs. Instead they added another API called hitheme, which also conveniently didn’t include anything to draw updated tabs.
So I guess I’m really to hard on the Trolltech guys and girls as this isn’t their fault. Apple really needs to get their act together on this one.
Why OSNews covers more articles on Qt compared to wxWidgets?
I want to see a detailed comparision of the two technologies.
Simple Question: What for?
How about Qt support ObjC and help out the current debacle with GCC on GCC 4.0 having broken ObjC support?
What is in Apple’s interest to expose all of Cocoa’s ObjC’s interfaces to C++ across the AppKit?
Apple is slowing removing Carbon from OS X. Apple is working on providing ObjC++ once Tiger is released and will role that into GCC but how does Qt improve Apple’s business?
The KDE Project helps with regard to WebCore with its ObjC++ KHTML/KJS code. Apple gives back code that it utilizes in the Open Source Community.
Qt allows a lot of software to run on the Mac platform. It also makes the Mac platform more viable for development especially for those who value portability. Objective-C + Cocoa is not portable. No one apart from Apple uses it hence no one outside Apple is actually interested in it. But that’s beside the point.
Carbon isn’t going anywhere. Carbon is meant to be a transitional API. It’s transitional in the sense that it allows developers to port existing apps to the Mac platform. By not updating the API to reflect the changes made to the UI elements, Apple is actually making the Mac user experience worse for those who use Carbon apps. Apple needs to get its act together and ensure that the APIs they provide are comparable to one another and provide similar functionality.
Using dirty double licensing tricks Trolltech found a way to earn a fortune on GPL-ed product. This “smart” move is disastrous for the Linux as a whole. Qt is a category killer and simultaneously blocks the possibility of Linux being adopted as an industry standard. One cannot expect the software industry to willingly pay a “developer tax”
in favor of an obscure Norwegian company whose sole achievement was to convince a group of irresponsible Linux developers to base Linux desktop on their non-free toolkit.
corrections:
its called dual licensing and its practised by several people including Mozilla firefox, Openoffice.org, MySQL, Sleepycat etc. its perfectly legal and moral.
QT is licensed under GPL for X11 and Mac’s and hence it definitely isnt a “non-free” license by any means. you have to pay if you develop and distribute a proprietary product which isnt “developer tax”
dont be a silly troll. if you want a gratis toolkit for use in proprietary products use GTK but dont whine
Why OSNews covers more articles on Qt compared to wxWidgets?
I want to see a detailed comparision of the two technologies.
wxWidgets is quite unusable on Windows. Any resizing causes horrible flickering. To avoid that you have to subclass every widget you want to use and write your own erase handlers. It seems that for some widgets it is sufficient to just disable the erasing. It’s worrying that an old library as this still has fundamental, obvious flaws.
Using dirty double licensing tricks Trolltech found a way to earn a fortune on GPL-ed product. This “smart” move is disastrous for the Linux as a whole. Qt is a category killer and simultaneously blocks the possibility of Linux being adopted as an industry standard. One cannot expect the software industry to willingly pay a “developer tax”
in favor of an obscure Norwegian company whose sole achievement was to convince a group of irresponsible Linux developers to base Linux desktop on their non-free toolkit.
You are exactly right, and it just pains certain people to hear the truth, thus the “Already Viewed” on your comment and others. Too bad that KDE had to choose a toolkit that can never be a standard which would promote Linux development on the desktop.
Actually I think GTK with a C++ wrapper like GTKmm or VDK for those who don’t want to do streight C would be a better standard for the Linux desktop than QT for the following reasons.
GTK is free for both closed source commercial and FOSS developers under the LGPL. (I am of the Open Source school that believes there is room for both the FOSS and “proprietary” development model.)
GTK was a standard on the Linux Desktop even before KDE and thus QT came along (as a part of the GIMP image authoring and manipulation kit.) and is pretty much like the win32 API equivalant on linux now.
The two competitors to KDE, GNOME and xFce use GTK as their widget set.
Mono uses GTK as its widget set.
GTK is a GNU product and thuss fits the GNU/Linux spirit better than QT with its “proprietary” origins.
As secondary toolkits for spacifically C++ developers that don’t want to be involved with “wrappers” around a C library I think Fox and FLTK are good bets. Both of these come under a modified LGPL that has even more liberal terms for commercial developers than the regular LGPL. The only thing wrong with FLTK is that doesn’t have a printing canvas yet. (That will brobably be a part of the 2.0 version when it is complete though.) However one of the good aspects of FLTK is that it can be statically linked for both closed and FOSS development allowing the development of distribution independent GUI software.)
Sad to read that people are keeping bashing on TT about the licensing issues. Get a life. This was fixed several years ago. And no, they don’t have a GPL’ed version on windows. They are trying to make a living on the Qt, and are doing this by selling commercial versions on windows.
But beside this, al those people that are calling ‘GTK++’, ‘Fox’ and on: have you ever really take a good look at Qt? I’ve tried the QT 4 Technology Preview, and the latest QT 4 Beta1, and it is simple: if you want a C++ cross-platform toolkit there isn’t a better one than Qt. It is simple as that. You may hate the company, you may hate the KDE community, but is a very, very good toolkit. Download the beta version, have a good look around at some of the very nice and elegant solutions they have (the new graphical engine Arthur is very, very impressive) and then form an opinion.
‘ve tried the QT 4 Technology Preview, and the latest QT 4 Beta1, and it is simple: if you want a C++ cross-platform toolkit there isn’t a better one than Qt
And the most expensive, propietary too, QT aint GPL just because TrollTech guys are good persons, all is part of a strategy to sell it.
An get all GTK developers get pay full time for develope just like QT developers and you’ll get tool as good or even better than QT.
trolltech might not at all be the perfect choice for spreading the gnu/linux spirit to the world, but they make a quality product, on which another really nice windowmanager, namely KDE, is based upon.
now, KDE has significantly changed many peoples views upon linux, and I have to say it’s for the good.
If there ever comes a day when QT will be unacceptable or unethical, some other toolkit will surely come and outperform it. until that day QT raises the standard and gives us a better use of our computers.
(yupp, tried gnome. even tried xfce, fluxbox and enlightenment. I ended up with my lovely KDE anyway.
and in my opinion, before we start making a great fuss about our toolkits, help y-windows out to make a consistent API and get the horrible X11 out of the game.
y-windows, as per the white paper, will end all these toolkit wars.)
“Using dirty double licensing tricks Trolltech found a way to earn a fortune on GPL-ed product. This “smart” move is disastrous for the Linux as a whole. Qt is a category killer and simultaneously blocks the possibility of Linux being adopted as an industry standard.”
Okay. Here is a shocking fact for you. TrollTech is a company! Okay, got that? Companies exist to make money, they don’t exist to fullfill your ideologies.
Qt is a professional commercial toolkit that does the job. The cost of Qt licensing is insignificant in any commercial application budget.
Often people here think using GPL software reduses the expenses to $0. You forgot that TIME costs money, and it’s the balance between time and the cost of software that matters, not just the cost of software itself.
KDE has significantly changed many peoples views upon linux,
This topic is about QT, not about KDE.
I still don’t know khow some trolls keep associating QT topics with KDE, nothing to do, KDE is a free desktop that uses QT, and QT is double licenced toolkit that charges more than $1,500 per developer, in other words, dont try to disctract that pointing to KDE desktop.
If you like KDE fine, that doesn’t change the fact QT licences are to expensive and uses KDE as a tool to promoto it.
When you grow up and get out in the real world you will se that the commercial licence for Qt are not expensive. No serius comercial software company can, or have claimed it is. So please stop talking nonsens and get your facts straight.
When you grow up and get out in the real world you will se that the commercial licence for Qt are not expensive.
Of course is not expensive, is super expensive, I must give thx more cheaper and with more quality tools like .NET/Mono and Java that allow me not to give a cent to Troll Tech.
And maybe when you see the world outside your litle QT/KDE world, you will see the damage QT has made and the real reason of tool of GTK to exist, of course, I don’t spec that from some one like you.
Since Qt is so super expensive as you say, please give ONE real world example of a serious commercial software company claiming so.
If the cost of Qt is a problem in any commercial setting you have to seriously reconsider the business model of the company or project. It’s fatally flawed.
And if you get yourself out of that fog of an anti Qt zealot out in the real world you’ll find even RMS prefers Qt’s license over GTK’s LGPL. And any damage only exist inside your head.
People should stop bashing Trolltech for releasing Qt under a dual-licensing scheme. After all, they are entitled to it. If you write GPL software anyway, the commercial license doesn’t affect you.
That said, I’m still quite surprised that people claim Qt is the best C++ toolkit. Admittedly, my experience with Qt is somewhat limited to just a few toy applications that I wrote to get a feel for the toolkit. What I found was that Qt wasn’t all that amazing, when compared to stuff like wxWidgets. Maybe there are some features that make it worth the $1500 commercial license, but I personally didn’t see them. This is why a wxWidgets vs Qt article will be very welcome :-).
If the cost of Qt is a problem in any commercial setting you have to seriously reconsider the business model of the company or project. It’s fatally flawed.
The assumption here is that everyone is able to afford the $1500 for a toolkit. Sure, big companies may be able to afford it, but definitely not the average shareware programmer.
Hmm. Perhaps that’s why in lieu of shareware apps, we’ve seen a move towards open sourced projects that ask for donations (not that it’s a bad thing; developers do need to eat after all).
The assumption here is that everyone is able to afford the $1500 for a toolkit. Sure, big companies may be able to afford it, but definitely not the average shareware programmer.
I second that.
Even if you are what you call average shareware programmer you will need some sort of plan if you intend to make any money. Else you should actually reconsider and do it like you say, open source it and perhaps ask for donations.
As a side note, you know that a big part of the shareware market actually consist of small and medium sized companies. Like WinZip Computing or as Jasc used to be.
And if you look at the cost of Qt, it’s something like what a decent developer will cost for like 1 to 2 weeks.
And if you look at the cost of Qt, it’s something like what a decent developer will cost for like 1 to 2 weeks.
Depends on what country you’re talking about :-). In my home country, $1500 is about 3 months pay for an average software engineer. It’s quite a bit and makes Qt a rather tough sell to management. The question that get’s asked a lot is whether Qt is worth that much.
You must also bear in mind that $1500 is for *one* platform only. Contrast that with other toolkits that are free for commercial use on all platforms. Sure, a lot of them don’t stand a chance against Qt, but wxWidgets can give Qt a good run for it’s money.
You must also bear in mind that $1500 is for *one* platform only. Contrast that with other toolkits that are free for commercial use on all platforms. Sure, a lot of them don’t stand a chance against Qt, but wxWidgets can give Qt a good run for it’s money.
I second that too.
>Depends on what country you’re talking about 🙂
True, and that’s a factor you have to consider when starting a project. Also remember a developer cost more than her/his salary, in some contries much more:-)
To sell to management it depends on two factors beside cost, one is higher productivity(less important in some countries). And professional support, and that’s usually a winner with management types:-)
That said, wxWidge is a good toolkit. Depending on circumstances it can be a better choice than Qt. And you can get professional support for it, but this will increase the cost.
With all this talk about the price of QT no one mentioned its second not so hot feature. That is that its signals and slots system requires TWO COMPILERS (TrollTech’s moc and the regular gcc/g++ compiler) to build apps with it. This can be one big MESS on any of the Linux IDEs particularly when QT Designer is involved in GUI building. I perfer toolkits that are compiled into apps by the C/C++ compiler, the Java or Mono compiler or the Python byte code compiler ALONE. GTK alone or with a C++ wrapper like wxWidgets, VDK or GTKmm or streight C++ toolkits like Fox and FLTK fit this bill better than Q (two compiler) T.
I perfer toolkits that are compiled into apps by the C/C++ compiler, the Java or Mono compiler or the Python byte code compiler ALONE. GTK alone or with a C++ wrapper like wxWidgets, VDK or GTKmm or streight C++ toolkits like Fox and FLTK fit this bill better than Q (two compiler)
None of the language bindings – for Qt require you to use a moc, so bringing up a bunch of gtk bindings as examples is a bit of a straw man.
If you don’t want to compile a Qt Designer .ui file to C++, you can read it in an runtime via the QWidgetFactory class just like you can in Glade. Or you can compile it to C++ code with the uic tool if you prefer. And you can do this in various other languages too via bindings – python, java, ruby, C# and perl all have their own versions of the uic compiler.
Qt Designer is now integrated into the next release of KDevelop for C++ and Ruby, so you can edit .ui files from within the IDE just be clicking on them in the File Selector tab. Then the C++ or ruby code gets generated with a simple makefile rule once you’ve finished changing the .ui file.
The assumption here is that everyone is able to afford the $1500 for a toolkit. Sure, big companies may be able to afford it, but definitely not the average shareware programmer.
Nobody ever claimed that the average shareware programmer is the target audience for Trolltech.
Maybe they aren’t?
Maybe it is companies who employ C++ developers, want to create a product and share as much code as possible between different platforms’s builds?
Maybe it is companies that would also consider buying other professional development tools like Rational Rose?
Nobody ever claims that Qt’s commerical licences are cheaper than gratis alternatives, but obviously Trolltech’s licencees think it is worth it.
Maybe it is companies who employ C++ developers, want to create a product and share as much code as possible between different platforms’s builds?
Maybe it is companies that would also consider buying other professional development tools like Rational Rose?
I don’t think do, no when there are cheaper with more quality options, Troll Tech is overpricing QT, simple strategy, make users fall in love of KDE then give then a trial GPL licence after that sell them a $1500 licence per deveper, good think QT ain’t popular among comercial developers and GTK is getting better.
I don’t think do, no when there are cheaper with more quality options, Troll Tech is overpricing QT
Ok, your opinion. Trolltech’s customers seem to disagree, otherwise they wouldn’t have chosen Qt.
good think QT ain’t popular among comercial developers
On the contrary, it is! More astonishing mostly for Windows development, also there is an increase in Unix/Linux usage.
GTK is getting better
Good, especially for C developers.
Before you start saying these high proces are good for a company and its licensees I would suggest you take a look at Borland. Ever since they dumped the shareware developer “standard version” of their profucts and started charging QT like pricing for their products Borland has been a MESS and its Yahoo finance readout looks like a classic SCO pump ‘n’ dump. No one is saying that QT should be gratis for shareware developers but a price somewhare on the low level between gratis (probably in the 100 to 300 dollar range) and FIFTEEN HUNDRED DOLLARS for what another double licensing company Garage Games, calls an “Indie” License would
probably be good for both the shareware developers who want to take their wares cross platform, for TrollTech and for Linux too.
I hope this does’nt affect KDE 4’s release, can’t wait for it.
QMacStyle
Get the correct interval and number of tickmarks for QSlider.
Fixed text position on buttons.
Fixed drawing of drop menus.
Fixed bug where text in a scroll view would be displayed as
disabled.
Fixed drawing of QTabWidget when an icon is too big for a tab.
Improved look of buttons with several lines of text.
Fixed background for toolbox tabs.
Fixed text truncation in QToolBox.
Finally!!!!!! Some much needed fixes on OS X and more importantly, the text alignment on buttons on Panther has been fixed! Took them, only 1 year and 4 months to do it, but hey, as long as it’s fixed.
Trolltech is not advancing the market share of their main framework very well, mostly because of the insane pricing.
Nor are they advancing the technology at any reasonable pace.
They think mobile phones are the future, but do not realize they are being totally outflanked by Microsoft.
It is sad. But it is life.
Trolltech thinks the mobile market is the future, but doesn’t really care about windows anymore as demonstrated by them not releasing the open source windows sdk anymore.
Novell just hired the windows gtk guy to work on it full time, so it doesn’t really matter anyway.
There’s still plenty market for them to make a great profit. It’s kinda like CPU’s; a lot of chip makers get flak from people here and on /. for trying to sell CPU’s when there’s no way they can compare to Intel or even AMD, but what these people are missing out on is the fact that the CPU market is MUCH wider than the x86 desktop market. Even as a small player, in any market, it’s very possible to make a good living.
The post I’m replying to might sound like a troll, but I think it’s just a common misconception. And about the other point, that they’re not advancing things fast: compare their progress to other toolkit makers’ progress. Trolltech really isn’t doing that bad!! How many other toolkits can you name that are 100% X11, Quartz and GDI compatible?
Trolltech really isn’t doing that bad!! How many other toolkits can you name that are 100% X11, Quartz and GDI compatible?
wxWidgets is 100% X11, Quartz and GDI compatible. And they actually honour the look and feel on OS X, since they use native components instead of java Swing style emulation of the various system look and feels. Plus it’s free for commercial and non-commercial use.
Only downside is it isn’t as well documented as Qt and doesn’t have a company standing behind them.
if q2 3005 is the new date anyone remember what the orginal was?
Trolltech thinks the mobile market is the future, but doesn’t really care about windows anymore as demonstrated by them not releasing the open source windows sdk anymore.
While Qt doesn’t provide an open source SDK for Windows, ones exists. http://kde-cygwin.sourceforge.net/qt3-win32/
It’s basically a fork of the X11 GPL version of Qt with all the relevant X11 bits being substituted for Win32 equivalents.
“if q2 3005 is the new date anyone remember what the orginal was?”
Probably a date *this* millenium anyhow ;-P
Finished compiling this new release. There are still some things that can be done to make the Mac style better. Panther doesn’t have tabs. They are replaced by things that look like lozenges. Trolltech should update Qt to reflect this change. Toolbar buttons should not have visible boxes around them. This is alright on other platforms but it isn’t native-like on OS X. A thick black border around buttons on a *white* Aqua toolbar sticks out like a sore thumb.
Qt4 doesn’t seem to fix this either and the situation can/will only get worse once Tiger is released. If Trolltech wants to make a presence on the OS X desktop market, it needs to work on ironing out the kinks in Qt.
q1 2005
How come they aren’t implementing it using native widgets on OS X. Is this not technically feasible? (How is it done on Win XP?)
> Qt4 doesn’t seem to fix this either and the situation can/will only get worse once Tiger is released.
Do you like to spread FUD? Tiger will finally provide an API for “correct” tabs, see http://lists.trolltech.com/qt4-preview-feedback/2004-12/msg00266.ht…
Do you like to spread FUD? Tiger will finally provide an API for “correct” tabs, see