[Kde-bindings] Future and status of qtruby (on Windows ?)

[Richard Dale]

On Thursday 30 September 2004 20:58, Han Holl wrote:
> Hello,
>
> I've been playing with qtruby (on Linux, I'm a Linux only person), and I've
> been surprised by the completeness of it all. I't far from robust (feed the
> wrong kind of object to a method, and you're sure to get a segfault), but
> as long as you don't make mistakes, everything seens to work. This is of
> course the best of two worlds.
Maybe report any seg faults on this list, and I'll see if they can be fixed. I 
don't personally get them, but perhaps I don't make programming mistakes :) 
There should be type checking on the arguments to a method and so on, and you 
should get a ruby exception rather than a seg fault if you pass the wrong 
type.

> At work I'm programming servers in ruby (fun, and works great), but
> colleagues of mine are creating GUI's fow Windows (to communicate with my
> servers), and they are working in Delphi, which really is a very poor
> environment when compared with Qt and ruby.
> I wonder if in future they might switch to Qt/Ruby, which for us would have
> the additional advantage of having the same development language, apart
> from the enormous productivity gain.
> For this to happen I must convince my management, that Qt/Ruby on Windows
> has a future: as long as it's a one man project I don't think they are
> going to go for it.
Well there are several people other than myself who are familiar with it - 
Alex Kellett was the co-author, and it was derived from the PerlQt project by 
Germain Garand and Ashley Winters. As it's part of KDE, it isn't really a 
stand alone project, and I would hope there will be a family of bindings 
based on the Smoke library.

> The question really is: is there any chance that TrollTech is going to
> adopt these bindings (probably Python and Perl as well) on Windows? It
> would probably been seen as a direct competitor of their own QSA product,
> which I cannot believe has a strong market position.
You can already buy a commercial version of PyQt. If there was sufficient 
demand for QtRuby I would love to be able to make a living out of it. I met 
Matthias Ettricht and Eirik Chambe-Eng at the aKademy KDE conference and had 
discussions with them and generally evangelised ruby as much as possible. I 
think they believe that dynamic languages like ruby will become more 
important in the future. Normally if you write some software with the Free Qt 
license, you can't change to a commercial version. But I'm pretty sure they 
wouldn't object.

Harri Porten has a small company called 'Frog Logic' which sells a 
testing/scripting environment, and he has no problem with being seen as too 
small, and commercial customers are happy to deal with him.

> Paying for Qt licenses is of course not a problem for us, so it might be
> seen by TrollTech as a means to promote Qt. After all, other people are
> creating these languages (of which at least one is wonderful) for free.
> Has this been discussed anywhere already ?
Yes, at aKademy see above. I think the best way to reduce the risk of being 
the first to do commercial development is to hire the author (myself and/or 
Alex) as a consultant. Then we can be on site, fix any bugs as they arise and 
train people on the team. I don't think I would charge for a commercial 
license under that arrangement, although you would still need to buy 
development licenses from Trolltech, just as if you were doing C++ 
development. I believe the PyQt license is cheaper than the standard C++ one, 
but it doesn't allow you to do C++ development, only python.

>
> PS Of course, what I would really love to see is the replacement of Windows
> with Linux desktops, but I don't realistically think I can wait that long.
> But when it happens, I'll be ready.
With Qt everything you write is cross-platform, so you aren't locked in. And 
also you can develop on Linux and then deploy on Windows, which would make 
for happy programmers like ourselves.

Regards
-- Richard