[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
SEUL: Dispelling some fallacies about Qt.
I have seen too many absurdities about Qt. So it is time to put
things right. Note than I am on GNOME side against KDE.
-Why GPL is irreversible
-Qt in GPLed programs
-Qt in commercial and shareware
-Qt against Motif
-The _real_ danger of Qt (What happens if the licence changes).
1) Why GPL is irreversible
Imagine this. Linus trying to make Linux commercial. (No he will not
try, he is a nice guy). The facr is than Linux contains many parts
who are not from Linus so Linus is not sole owner. Because software
derived from a GPLed program must remain GPLed Linus could not base on
present Linux because of all the GPLed but non-Linus parts in Linux.
2) Qt license:
You can redistribute Qt. You can run programs with it. You can use
it for writing free programs (the distribution of source code is the
criteria for free in Qt license). You cannot modify it. You have to
pay for using it in commercial software (I am not shocked by this
restriction). Last time I checked you had to pay if you were writing
_any_ (free or not free) kind of MS Windows software. The later I
find an excellent idea :-)
3) Qt in GPLed programs.
Can you include Qt in a CDRom? YES. Infomagic does. The licence
allows redistribution and it does not enforce giving Qt for free.
Can you include KDE or any other Qt-using programs in a CDROM? Yes.
GPLed and LGPLed software are explicitly allowed to use Qt for free.
And Qt licence does not condition its use to the fact than the program
must be distributed gratis (it would no longer be GPL) but in the
distribution of source code
4) Qt in commercial and shareware.
Here you have to pay. The cost of Qt is not too much of an issue in a
comercial firm: a programmer costs over 150$ a day so if Qt allows
shortning the time of development then the cost is rapidly recovered.
However the cost of Qt would be an obstacle for people developping
shareware, specially for those people for whom the shareware licenses are
only a nice income bonus and not their main activity.
5) Qt against Motif in commercial software.
Motif is cheaper for developpers but Motif makes for a longerr
development cycle than Qt. More important Motif using programs do not
sell well in Linux world because you are forced to use static linking
thus making a big and slow software. On the other hand every Linuxer
can have Qt for free if all it does is running programs with it, so a
Qt based program will sell better than a Motif one in Linux (not in
commercial Unixes) because you can ship dynamically linked versions.
6) The _real_ danger of Qt (What happens if the licence changes)?
If tomorrow Qt license is modified so you cannot get it freely what
happens? Nothing in a short term. You have a legal copy of Qt and Qt
license allows redistribution so we can continue including a copy of
Qt along each copy of KDE. We can continue including it in CDROMs.
We can continue writing GPLed programs with Qt. Nothing changes.
What is wrong then? The problem is than we are not allowed to modify
Qt so programs using Qt would be hampered by a frozen library. The
effect would only be noticeable after some years when the free Qt
would have become obsolete compared to modern toolkits. That would
not be a problem if only a few small programs a la EZPPP used Qt, the
problem is than the project becoming the standrd desktop for Linux
will cause than most of ythe X software for Linux will be sooner or
later be rewriten for its toolkit. So if KDE becomes the standard
desktop for Linux sooner or later most Linux X software would use QT
and then if Qt changes licence that would be bad for Linux. Bad but
not catastrophic: remember than all waht happens is a frozen library.
_NOTHING_ more. Linux people would have to rewrite Qt based software
but they would have plenty of time for doing it.
So I oppose KDE but please not be hysterical about the subject. I do
not want than distributions base their desktop on KDE. But pricvately
I use KDE for impressing Windows users. And it is very efficient in
this role. :-)
Conclusion: The problem is not so much Qt as KDE. KDE would be a
heavy weight who can satellize all X Linux around it and so around Qt.
That is the reason it is important to have an alternative to KDE, and
having that alternative win.
Jean Francois Martinez
==================== The Linux. Use the Linux, Luke! =======================