Hello Patrice,

You made some interesting points, and I have some answers for you.
But first, let me say that if I'm somewhat an advocate of the (a) vote, it 
is because I think that TrollTech did something very kind to us. 
Actually, they did something like what I wished they would do
(see my message on this list, dated of 27 august 1998
 (Message-ID:<[EMAIL PROTECTED]>),
and other of my early messages on this list). At that time, most of the 
Harmony developer were skeptical about such a change.
Torben Weis however said "a NPL like Qt could be a very good idea."
Now that it has happened, I see no point in continuing with the
original aim.

So now I'll answer your questions:

Patrice Fortier writes:
 > 1- First, if I (we?) go for Qt hacking, I can't go back to Harmony if 
 > something goes wrong. This can be a bad surprise from TT or because 

We have never be in position to prove that we  haven't read the Qt
code. So, if we actually read the Qt code, and we want to clone again
Qt in the future, I believe we could still do so, only being very 
careful with how we implement our library to never duplicate their solutions.
Not a big deal as "there's always more than one way to do it".

 > I got very disapointed by the maintainer - who is the only one who 
 > decide whoch patch can be included in the distro. FYI, this is one of the
 > reason I left Gtk, so I'll think twice before going on with Qt...

You are speaking of the official Qt only. But as you were able to fork Gtk 
if you really wanted to, you are able to create an unofficial
Qt distribution, but in the form (TT's original Qt + your (or others) patches).

 > 2- I really don't like the patch stuff, and the fact that only TT can 
 > decide which patch is good for inclusion and whoch one is not.

Isn't the linux kernel developed with patches? Contributors send their 
patches to linus, who drops them in the wastebasket, then they resend
patches, and so on... until they get accidently compiled in :)

 > If you don't agree with TT, you can't shism, like any other OSS project.
 > Then see 1.

No. Reread the QPL. You can distribute your changes, or even a
modified binary library, provided the patch is available somewhere.

 > 
 > 3- The QPL is incompatible with (L)GPL, this can be more nasty than we
 > think for a Gtk-Qt compatibility project...

Right, if this is important for you. But I don't see this issue
primordial (I'm confident that it will be solved too, as more important 
problems have already been solved).

 > 4- Qt 1.40 is not under QPL, and Qt 2.0 is not out yet. So I won't switch 
 > before 2.0 is out with the proper license (paranoiac? me? :)).

Good point :) 
But I can ensure you that this _is_ really paranoiac and that you can
trust them.

 > 5- There is currently no way to know if the QPL is a political move,
 > or a benevolent one. This means: Will the TT guys include all the patches
 > we'll submit to them (like relay++ for example), or is this just something
 > to get the "OSS" flag and all the good publicity that comes with it?
 > 

It is both. 

If you look at the way TrollTech behave, you'll see that they were
always supporting the free software ideas. That's for the benevolent
part.

It is also a political move, of course, but in the good sense :), that is,
foresight.

Their non-modification clause, which made them non-OSS, had some
advantages in their minds. 
Whether they have realized that those advantages are not really
threatened by an OSS status or that there is bigger to gain than to
loose, or both, I don't know. 
The fact is that they realized that the overall good decision for them
was to make their product modifiable and redistributable when modified. 
That is, really OSS.

 > 6- TT used to ship products already beta tested... I don't think there
 > will be cvs trees, or any "release-of-the-day" for developpment, but for
 > that point I may be wrong (any experience from the KDE guys? Did you play
 > with Qt 2.0 - except the guys working at TT of course).
 > 
 > 
 > All this makes me skeptical, and I'd like to wait a bit more before taking
 > a decision, which is a "no-return" one.
 > 

I too feel it is a no-return one, but not for the same reasons :)

-- Christian

Reply via email to