New Python/C# Bindings Expand KDE Languages

Dre writes: "Today marks a special coincidence. First, Adam Treat released the initial version of Qt bindings for C#, which consists of 476 Qt classes converted to C#. The bindings work with the Mono compiler, runtime environment and class libraries, enabling a fully Open Source implementation of C# for Qt. While not yet ready for a real application, Adam has managed to write and execute a Hello World! program (screenshot). KDE bindings are on the drawing board. Shortly thereafter, Phil Thompson, Jim Bublitz and theKompany.com released KDE 2 and KDE 3 bindings for Python. Together with the Java, Objective C and C bindings in the kdebindings module, as well as the Ruby bindings, KDE is providing developers a broad gamut of application development languages."

Perl

[redcliffe]

It needs a perl binding.

Re:Perl

[sleepingduke]

The 'kalyptus' bindings generation tool has a '-fperl' option to generate .pig (Perl Interface Generator) files. The .pig files can be then be fed straight into Ashley Winters' Pig.pm generator.

kalyptus is in the KDE cvs under kdebindings/kalyptus.

The main technical obstacle to producing Perl bindings for Qt 3/KDE 3 is that they assume that the moc generated code works like Qt 2. Qt used to call slots via function pointers, but in Qt 3 it uses an integer offset into the slot table.

So the slot/signal handling needs to be redone to get the Perl bindings to work.

And Common Lisp for Qt/KDE should certainly be possible.

-- Richard

3rd post!

[hkhanna]

NT.....waiting 20 seconds....... Mod this down please.

Why?

[Garthnak]

Alright. At the risk of sounding like a troll or a flamer, I'm going to ask what I do think many other people just asked when seeing this article:

Why?

Okay, okay. This has probably been covered in other Slashdot articles covering C#, .NET, etc. But for one thing, I do not see these things gaining a lot of acceptance in the open source community. There's the Mono project, but in my personal interaction with other OSS programmers, they agree this is a Bad Idea.

Alright, I'll spare you the rants and reasons on the subject - you've heard them all before - but my point is here is not necessarily my own beliefs and opinions on the C#/.NET situation. My point is that this is something that _most_ of us (in my experience) are rejecting outright. I think Miguel de Icaza is crazy for spouting pro-.NET rhetoric in the GNOME project when there's such a clear lack of support of most of his community, and this Qt thing just doubles that across both desktop environments.

Am I totally out of line here? Will anyone _actually_ use this? Is there really a population of hackers out there who will be writing X apps with C#? Personally, I'm highly dubious about the whole thing. If you agree with me, reply. If you don't, reply. But flames can be filed in /dev/null.

--Garthnak

Re:Why?

[WasterDave]

I agree with you. It's all completely barking mad. We really need one, maybe two languages that developers can rally behind. One is C/C++, depending on whether you're in the Gnome or KDE camps (-1, Flamebait). The other should maybe be Java (-2, Pisses off Python jockies), although the open source-iness of Java is currently under some debate. Perhaps Perl? Given it's solid backing in the scripting / admin community, it would make a sensible choice.

So, choose our language, and get some support behind it. Witness the wonder that is KDevelop :) And get some good documentation, which was part of Miguel's reasoning to use .net.

But, yes, bollocks to C#

Dave

Re:Why?

[Chmarr]

The more languages supported, the better.

If I have to rewrite my 20,000 line application just so I can have a KDE-native version, then I'm probably not going to do it. However, if all I need to do is link in with a particular binding, and change some of the interface stuff, then things seem much more likely.

One should be looking at things from a "The Right Tool for the Right Job", and you can't answer that question with support for just one or two languages, or even 5. There are going to be applications that call for a particular set of languages and libraries. Its not your choice to decide what the 'right' language is, but instead the application developer.

Re:Why?

[miked1001]

It DOES make sense to implement Qt with C#!

Since Visual Studio .NET was only just released, it may not be very apparent how widespread the usage of C# will turn out to be. I'd wager the language will be very widely used to develop new Windows applications. Obviously, particularly those based on .NET.

In my opinion, it could only help the development of Qt based apps once C# developers realize they can easily develop in a syntax they're used to.

Work on this project needs to start now so that when that day comes, the C# developers will have the tools they need. It's a matter of anticipating the needs of future Microsoft-flavored programmers when they make the switch to a superior OS.

Let the flames begin,
Mike D

Re:Why?

[Hard_Code]

Why do geeks do *anything*?

1) because they can
2) because they felt like it

I'm so sick of these "why are we doing this! whine!" questions. Nobody is forcing you to use C#.

Languages are like hammers

[torkar]

You're missing the point.

You pick the language you are familiar with or most importantly of all - the language that makes it easier for you to solve a problem (doesn't matter if it's scientifical app or GUI or ..)

Languages are just like a regular tool, you pick the right one to do the job.
If you want to do some heavy crunching you pick a heavy hammer (fortran, C) and vice versa.

In this case they want people to use *their* favourite language to make GUI applications for KDE.

Uh... no.

[__past__]

...released KDE 2 and KDE 3 bindings for Python.

Well, they didn't. Actually, a first beta was released, and even that is not yet mentioned on the linked page.

Which is a pity, because, having tried the alpha versions, programming with PyKDE is really, really nice. Unfortunatly it's quite a huge package, so unless Distributors ship it by default, you shouldn't expect users to be too happy it you tell them to compile it themselves.

Yea!

[RevAaron]

Man, this Mono stuff is taking off pretty well... it may get me to switch back to Linux from OS X just yet! :) Does Mono, at this point, rely on any ASM, or is it all C? Has anyone compiled it for PPC?

Re:Yea!

[high]

Some parts of it compiles under PPC. But mono is a big project so its hard to say how useful it is for the average developer =)

They are working on an OS X/Darwin port too!

To back up my statement:

[Mono-list] (MacOSX) Requesting porting guide clarifications
http://lists.ximian.com/archives/p ublic/mono-list/ 2002-March/003528.html

Re:Yea!

[RevAaron]

Awesome, thanks!

It's a trap!

[setag]

Duck and cover! Implementing .NET and c# on Linux
is just asking for trouble from M$!

Fight the power!