Re: [gtk-list] Re: To GTK, or not to GTK - that is the question



Chris.Bitmead@misys.com.au wrote:
> 
> >I'm talking about motivation. You asked why people write stuff with
> GNU/Linux in >mind. Well, if I have a GNU/Linux box, I don't care that
> Solaris has a nice CDE >environment and extending it would be cool. I am
> concerned about the fact that >GNU/Linux does not have such an environment,
> and if I wrote the environment in >such-and-such a way it would be cool.
> 
> One of the reasons that people are talking about Linux moving into the
> mainstream
> now is that it conforms to standards which are supported by other vendors
> like
> Sun, HPUX etc. I wonder if going off at a tangent is a good idea.
> 
Why do you consider this going of tangent. Motif is available,
CDE is available (commercially). So there is no problem with
adhearence to standards. On the other Hand GNU/Linux is also well
known for inovative solutions, whereby GNOME/Gtk is one of them.

> > When deciding whether to use Motif, the fact that most free systems don't
> have
> > it (and Lesstif sucked when Gtk started) is a big factor.
> 
> If Lesstif "sucked", when Gtk did not exist, then I guess there were two
> choices:
> Work on Lesstif, so that it didn't suck, or go off on a tangent.
> 
Lesstif has been continued to work on.

> >If you think about it, Motif has never been a standard on free systems.
> 
> 5 years ago, UNIX wasn't a standard on free systems, because there were no
> free systems.
> 
I take this you mean there was no freely availabe UNIX.

Ok, a short look into UNIX history tells us:
Source Code of Bell Labs Unix Version 6 was freely available (1975)
Start of Linux Development as a joint venture in the internet(1991)
386BSD0.0 from Berkley University for free (1992)

All >5

> >And Xt by itself is just not that useful or exciting;
> 
> Xt was never supposed to be "exciting". Just meant to be a standard way to
> get
> different widget sets to work with each other. There are truckloads of very
> specialised widgets out there that can work with any Xt widget set.
> 
That is an advantage that I never managed to exploit, as it either
looked
odd or things didn't work that well together after all. Mind you I've
never
been much of deep imersed Motif expert.

> >if you're dropping Motif and CDE, you lose little by reverting to Xlib and
> >starting over from there. Most people hacking Gtk apps were never Motif
> >programmers, so their was no lost-expertise cost in that respect.
> 
> Most people hacking on personal computers never used to be UNIX
> programmers.
> They hacked Apple IIs and C64s and Trash-80s. But adopting an existing
> standard was a good idea.
What standard has been adopted and by whom??

> It brought everybody together - hackers, commercial
> vendors, UNIX programmers. There ARE Motif programmers out there (lots of
> them of which I am one).
Yes, and there are many Gtk programmers as well already.

> 
> >Basically there was much to gain and little to lose by dropping Motif.
> 
> It seems like there is a lot to lose to me. Don't take this as a flame. It
> may be worth it. Maybe I will find out why as I start to use GTK. But it just
> looks to me at the moment like gratuitously breaking standards. It's not clear
> to me right now what exactly could possibly have been gained that couldn't have
> been done by extending what was standard out there already. There's just so 
> many X11 widget sets out there, it's crazy.
> 
I don't consider Gtk as a toolkit breaking a standard as it was never
intended
to work with the standard you are referring to. Rather it is a new
implementation
of a widget system.  The Gtk system has been build upon the experince
people have
been having with other widget sets, the result has been something
different, that
is evolution. There are so many people using Gtk because it offers the
advantages
of for example being: complete, easy to program, visually apealling.
Netscape just
abonded motif in favour of Gtk for their unix ports. I would think they
wouldn't
have done this if Gtk didn't offer as much functionalty as motif.



[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]