Re: Some random thoughs and questions on the future of GNUstep

[Helge Hess]

On Sep 30, 2006, at 16:45, Nikolaus Waxweiler wrote:
> Now, why CMake when we already have got GNUstep Make?  First, I must
> admit that I have not used CMake extensively -- barely at all,
> actually.

Now thats a very bad position to advocate for anything. Really. What  
about first learning CMake and then showing a prototype on how things  
work and would look like? :-)
That would be a much better basis to bring up such a proposal.

> There are a few reasons why I propose it anyway, apart from
> a big project like KDE using it.

Well, GNUstep doesn't have the major issues KDE wanted to solve with  
CMake. So its a really bad example.
What I miss about your proposal is outlining actual problems GNUstep  
has and which would be solved by using cmake (and how).
In my understanding the major issue KDE had was autotools and not  
make. Obviously GNUstep make projects do not usually need or use  
autotools.
[and in fact we could improve even more on this by exposing configure  
information collected at gstep-make time to gstep-make projects]

> - CMake works out dependencies of files automatically, so you can run
>    several concurrent build processes ("make -j 4"). I heard there  
> were
>    problems in that area with GNUstep Make.

If you read Nicola's responses to that topic (all archived, yes), you  
know the answer to this one. No need to guess (when I hear "I  
heard" ;-).

And BTW: a major difference is that as a KDE developer you can't  
really live w/o -j because the compiler is so slow compared to cc1obj  
(w/o an icecream cluster you really spend loads of time just  
compiling). For GNUstep it would be just a nice to have, but its not  
a major issue.

> The general idea is to move away
> from homegrown reinventions of the wheel and letting others to that
> work for you.

What I like about GNUstep make is that its _not_ a reinvention of the  
wheel but a very clever use of the existing stuff. In fact you could  
possibly say that CMake is a reinvention of gstep-make? Probably  
around much longer ;-)


Anyway, would be interesting to see what people come up with :-)  
GNUstep-make is certainly not perfect but quite nice overall. I would  
be interested to see how a typical GNUstep make package would look  
like in cmake.

Greets,
  Helge
--
Helge Hess
http://docs.opengroupware.org/Members/helge/