Qt ported to Mac OS 9/X, BeOS

Trolltech have announced that they have successfully ported Qt to the Macintosh platform. A demo can be downloaded here, and screenshots can be viewed here, here, and here.
At the same time, Qt Free Edition has been ported to BeOS (screenshots: 1, 2, 3, 4, 5, 6, 7, 8, 9) by Zenja Solaja, who is also considering a port of KOffice, Konqueror and KDE to BeOS. Along with Qt/Embedded and Cygwin/KDE, KDE/Qt might soon be the most ported toolkit and desktop environment.

Dot Categories: 


by Waldo Bastian (not verified)

We ship our own, slightly enhanced, version of ICE in 2.2.


by Matt Newell (not verified)

>>To what extent does the KDE code rely on X11? I think it'd be wise to start taking out the X11 specific code
>>from the core (Maybe this is already done? I haven't checked the code all that carefully; but if it was, I figure
>>we'd have KDE for fbdev now) and put a layer between KDE and the display (manager).

>>We have a good start with QT. Even if we're not interested in porting KDE to other operating systems,
>>we are, with any luck, going to get a better system to replace X sometime in the future. Better get ready for it!

QT already runs on the framebuffer, so does konqueror embedded.

Unfortunatly it stops there. Having KDE on the framebuffer in its current state would absolutly suck. Just imagine, no HW accel(except blitting), no 3d, no remote applications, no xlib/gtk/motif/x support.

You people like to blame X for everything, but you really don't have the slightest idea what you're talking about.

Matt Newell

PS. Sorry for being a dick, but I am sick and tired of people spreading FUD about X.

PPS. You guys really need to get the HTML security stuff fixed, plain text sucks.

PPPS. What would be really cool is if Konqueror had text formatting for line edits by using a subset of html tags. This could be fairly easily done by using the QT3 richtext control. Maybe if I get time this summer:)

by Johnny Andersson (not verified)

>>QT already runs on the framebuffer, so does >>konqueror embedded.

Yes, that's why I used that as an example. KDE on the framebuffer wouldn't be very useful (yet), but in this crazy open-source world, someone's bound to try it. ;)

>>You people like to blame X for everything, but >>you really don't have the slightest idea what >>you're talking about.

X is fine, I'm not blaming it for anything; we all know that X has its shortcomings (need I enumerate them here? probably not), but if it wasn't for X, I'd be staring at a tty right now. I'm just saying that if anything new (and good, and network-transparent, like X!) comes up, we should be open to it. We're already replacing old X programs with (better?) KDE-ized ones, so the legacy X apps we need aren't that many.

>>PS. Sorry for being a dick, but I am sick and >>tired of people spreading FUD about X.

Yeah, well I'm sick and tired of the "word" FUD. ;-)

by CY (not verified)

Berlin is basically pre-alpha at this point, but between it and ggi there is some COOL stuff happening. X is good, but even the best software sometimes ages after several decades. Even the great TeX, possible the single best piece of end user work ever, lacks some useful abilities, like handling jpg and png natively, which would require restructuring. X was built for dumb terminals - most networked computers aren't dumb terminals anymore. (Although at my school sometimes it feels like they've had a lobotomy... WHY did we have to pay for NT just to remotely display X applications?)

Ahem. Anyway. If X gets updated, great. If Berlin storms in, great. If both happen, best of all. Then we get to decide on merits, and since they are both open they can ultimately talk to each other if someone wants that to happen. You gotta love open source.

by Macka (not verified)

To be able to run KOffice whether I'm sitting in front of a system running MS Windows, Linux, *nix, *BSD, BeOS or OSX ... now that has the potential to really make someone in Redmond sit up and take notice.

It would be a very liberating experience too. I'd be effectively free to float between OS's as I chose. It would also do wonders in attracting new developers from all kinds of places.

Just imagine if KOffice ran on Windows and rivaled MS Office to the point where ordinary MS users were happy to use it as their first choice. It would really open the door for new people to try out Linux (or any non-MS platform) with minimal risk to their data.

This could create an incredible amount of leverage. Yum :)

by Martina (not verified)

Microsoft Office will not be easily detronized.
People are used to it. There are some free office suites (in much better shape than KOffice) for windows and not so many people use them. They don't need them. Microsoft Office is everywhere. And most PC sellers ship MS Word and MS Works with their PC's. So, most people see MS Word as a free application. I know many people and none of them bought MS Word. But Microsoft is its own enemy. If it starts changing licensing policy (charging per year or something) then Linux & KOffice have chance to expand. But then...multimedia :( There is a long way to catch up. More and more internet sites offers streaming media only if you use MS Media Player or Apple's Quick Time. Real player simply disappears. Guess why... Multimedia IS a future.

The future Office suite will be nothing but the multimedia player. You will be able to edit your documents in it, you will be able to listen to mp3 (or whatever music), you will be able to watch streaming video, you will be able to create multimedia presentations, browse internet, chat etc. And if KDE goes right way, and I am sure it does, you will be (very soon) able to do all above mentioned things in Konqueror. Integrated. Konqi's toolbar will not look like it looks today. It will have 5-6 buttons when you launch it: Office, Multimedia, Internet, Control, Development. Click on Internet button and whauu, you have a browser, click on Control button - you get Configuration tools (now Kcontrol), click on Development button and you get Kdevelop embedded in Konqueror etc. KDE already has some of these things but partially: in Kcontrol, in Konqueror, KOShell.

I can't emphasize enought what a good idea this is.

Anything that's good for the Mac is good for Linux, and a toolkit that makes Windows/Mac/Linux porting easy is the best thing that could happen for Linux.


Because every ISV that doesn't have a Mac version of their product wants to have one. Linux is a distant third for most of them (though that could change).

Still, even though they'd like it, they don't have those Mac ports.


Because it's hard!! And (in my case at least), they told themselves back in '95 that they'd be able to use MS's WIN32 cross-development tool to get there easily and then found out that it didn't work completely and was withdrawn from the market (actually, I've heard that it works now, but only MS has access to it). And the current Mac market isn't big enough to justify the cost of maintaining 2 code bases.

No matter how good KDE and GNOME are, they're not going to get mainstream support on those merits alone. But Windows/Mac/Linux from one code base is too good to ignore.

So, TT. You should advertise this and advertise it hard. One suggestion, though. The price has got to come down to the same ballpark as VC++ or Metrowerks. Not because Intuit can't afford it, but because you need to get that snowball rolling. At least come up with an exception for GPL'd apps. Anything to get the buzz going.
Once you succeed in becoming the de-facto standard multi-platform tool, you'll make plenty.

by stew (not verified)

The BeOS port ain't that exciting yet as it requres X11 for BeOS. The GTK port is closer to finishing, as it already runs without the need for an X server.

by Helgi Hrafn Gun... (not verified)

I'm one of those weirdos who know a bunch about Windows, MacOS *and* Linux.

I use Linux at home, I try (miserably, since we have this horrid Lotus Notes solution) to use Linux at work, but I've always liked the Mac and stood up for it when them close-minded Window/Linux-fanatics have been whining about how little you can do with the MacOS (which I must add, is complete and utter nonsense and always has been, but that's entirely beside the point).

I seriously dislike Java Swing apps, because they try (often miserably, may I add) to emulate what the operating system is supposed to do for you. A good example is scroll-mousing, which I haven't figured out how I'm supposed to get up and running in a Swing app. This is due to this (in my opinion childish) solution of emulating widgets that don't really talk to the underlying operating system at all... this whole virtual machine thing just gives my the creeps when it comes to GUI apps. You are supposed to be able to choose an operating system by the features it has. If I want to use MacOS, it's because I want the Aqua thing, the menubar in one place, the scroll-mouse thing working and so forth. If I want to use Windows, It's because I want the cold steel-look and manubars all over the place. I *don't* want applications to look and feel different than the ones of my neighbor, and I *do* want an operating system's user to decide for himself what he/she prefers in a GUI.

First of all, I must congratulate you on the dual-license for X11. Furthermore, I must congratulate you on *not* also using the dual-license in the Windows/Mac versions. This is not because I'm a Linux-fanatic (on the contrary I'm perfectly satisfied with the marketshare of Windows/Mac/Linux as it is), but because I think commercial apps belong to commercial operating systems, and open-source apps belong to open-source operating systems. Two satisfied consumer-groups which I believe must be kept apart up to a considerable extent, and are not meant to be mangled too much together. The open-source people should have their playground, and the closed-source group should have theirs. The groups have incredibly different goals & needs, and that's how I believe it's supposed to be. I wouldn't legislate anything, but rather use a form of encouragement... exactly the form you're using. *Allowing* Windows/Mac developers to create open-source software, but encouraging them not to do so, by making them pay for their GUI library. Furthermore, due to the dual-licensing, *allowing* Linux developers to create closed-source and commercial software for UNIX, but encouraging them to do open-source by letting them have the code for free as long as they only do open-source applications. Quite obviously, the Windows/Mac communities are *not* trustworthy enough to be given the free version, simply on their word for that they'll only develop open-source software with it. The open-source community (mostly Linux/BSD, I assume) however, is.

Whoof. I just can't stop babbling about this. ;)

Now, using the native components of every operating system, writing in a programming language that is supported by every single home computer operating system I'm aware of, is just beyond belief.

I write open-source software in my spare time, but I'm tellin' ya... the moment I'm capable of writing something useful in Qt on X11 with the GPLed version... I'm going to buy myself a copy of Qt for Windows and Mac, compile my applications on those platforms, and release them. Then I'm gonna sit back with a cigar and a beer, and think "Long live TrollTech". Sounding like a fanatic, I know, but this is just so great. :) Cross-platform C++ programming without even having to buy myself the operating systems I'm developing on. So much money and time is saved, and you deserve a piece of the pie for your work.

Furthermore, maybe this will open people's eyes towards the Mac and other exotic platforms. :) Maybe people will actually try them out before judging, thus, strengthening the all-so healthy and badly needed commercial competition in the operating system market. You guys might really, REALLY make a difference for operating systems like MacOS and not to mention BeOS. Just promise us you won't sell out to Microsoft. ;)

Just... great work, ladies and gentlemen. ;) I can't say it often enough. Great, f***ing work.

And for the bottom, I agree to what someone above mentioned... this is the best learning tool I've found for C++. I've tried Borland C++Builder and VC++, but they always either do *everything* or *nothing* for you. I'm writing a package manager for UNIX in Qt, doing fine so far, and I plan on writing some Internet applications as well. I've never been able to learn so much C++ in such a short period of time... well, with the exception of Gtk--, but still, you're still a little bit ahead when it comes to an easy API.