OSNews

I do agree!

You still have to compile the program to run on Haiku. It's not like some universal binary will be floating around. The OS/2 analogy is a bad example; save that argument when Wine is ported over. In the meantime, Haiku instantly gets the awesome Arora browser, which has been my browser of choice for the past month.

Long term, there could be a port of KDE programs to Haiku.

If anything, this will make adopting Haiku easier for developers.

Why?

As a developer I would not care about the BeOS frameworks if my application is QT based.

If all functionality that my application requires is provided by the framework is already there, why should I look elsewhere?

This is exactly the problem of multiplatform GUI development.

If you want to take advantage of what makes a certain platform unique, your application will be tied to that system. If on the other hand you want to make your application work everywhere, then you have the problem that your framework can only provide common features.

Porting of framworks such as QT will make it easier for developers to port QT applications.
I doubt many will write a dedicated Haiku app in QT.

My fear is that the platform will be completely reliant on ports rather than applications being natively developed. If no apps are native then the benefits of this platform won't be taken advantage of.

I hope not. Haiku has it's own UI and doesn't need KDE or another desktop environment. And Haiku has a nice coding team vith a good vision. I don't want Haiku to became like Linux with 10000 distribution, 10000 desktops, 10000 window managers, 10000 libs that does the same thing.

I don't think you are right. What does it mean "native"? Qt is a toolkit, and you can use it on BeOS pretty much the way you use MFC, Windows Forms or WPF on Windows. The vast majority of Windows apps is written using MFC not just straight, plain winapi. Does that became a problem for windows? No.

For me, "native" meanse more like natively compiled for BeOS as I not a BeOS purist.

I must add, Qt is a very good tooolkit and enables people to write very good qualityy software in a small amount of time.

Qt applications won't be lightweight and fast compared to Haiku native UI, won't have queries, attributes etc...

They won't be integrated with the rest of the system, sort of Linux. Sorry but I will use Arora for sure until Ryan will complete the native browser but there's no future relying only on ports, in the long run it will hurt Haiku.

Talking about OS/2, there is now a new port of QT4 to Warp. Check it at:
http://svn.netlabs.org/qt4

Thanks

OS/2 isn't dead (see eCS), but the reason it's less popular is because of IBM kicking out MS for refusing to abandon Win3.x APIs. Of course, then MS abandoned it anyways for Win32 for more control (although to be fair it's considered better anyways). OS/2 also lacked a lot of drivers, and IBM eventually stopped working on it. Plus, no offense to Serenity, but eCS is way too expensive (and buggy, sheesh, still no updated live CD .iso, old 1.2 won't boot for me).

So, no, compatibility is always good, that never killed any platform. It's when people start getting paranoid and stubborn, refusing to be compatible that things go wrong. Or do you really think nothing good ever came out of Cygwin, FreeDOS, DR-DOS, eCS, WINE, HX, MOSS, DOSBox, DOSEMU, EMX, RSX, DJGPP, OpenWatcom, etc.??

"We saw a similar thing happen with OS/2 Warp. Because it could happily support windows 3.1 apps, no one bothered making native ones."

here's how I see it. IBM wanted you to also buy their IBM PS/2. They didn't want drivers for non-IBM hardware.

Microsoft didn't care what PC you had: if their OS ran on many pc's it would be better for them. IBM's approach was doomed along with Amiga, Apple and all the others that didn't encourage open hardware.