Qt licensing hell!

[Boudewijn Rempt]

Mats Wichmann <mats at laplaza.org> wrote:
> On 24 Aug 2001 05:18:32 GMT, boud at rempt.xs4all.nl (Boudewijn Rempt)
> wrote:

> I believe Boudewijn is in the best position (outside TrollTech) to
> comment, he seems to have been working on this /forever/.

:-) - it was such a nice break when I was able to spend a week or so
on investigating microthreads, instead of thinking about GUI's. The
result will be up at InformIT at 9/10.

> :You cannot use your app in a commercial setting yourself, but
> :people who download it can.  --

> Now there's a nice confusing restriction.

It's caused by the fact that the people at TrollTech have some
difficulty wrapping their minds around the existence of a complete
Python binding of Qt - hacking Python source is so much easier than
hacking C++ source, that a lot of restrictions just don't make sense
anymore.

Imagine the following situation. I develop Kura, a nice, GPL'd
application for professional linguists. Not having been paid for
linguistic work since I was a student, I can distribute and hack
kura using the non-commercial Windows Qt. (Actually I'm using Linux,
and my last Windows-bearing hard-disk crashed a few weeks ago, but
that's not material.) 

If some professional linguist downloads Kura, he can use it in the
course of his work - which I couldn't do if I suddenly got a job
as a fieldworker in the eastern regions of Nepal. However, he cannot
hack the source if he's using Windows on his laptop - he can if he
uses Linux. If he wants a new version, he's got to tell me what he
wants, perhaps even write a patch, but he has to wait for my
new release before he can use it.

Of course, this is all quite theoretical: the nice people at TrollTech
are quite reasonable, and just want to make sure they get their share
out of the profit people make with their tools, but the letter of the
license causes all kinds of complex situations.

-- 

Boudewijn Rempt  | http://www.valdyas.org