Re: gnustep-make experiment

[Nicola Pero]

> Right, and also, frankly, there's a very important non-technical aspect 
> to this as well; Way more developers know what the heck .pc files and 
> pkg-config are, as well as how they work, than a funky GNUstep.conf 
> file, which, while somewhat self-explanatory, definitely is proprietary. 
> pkg-config is commonly used and makes things easier for us.
>
> As we move towards a totally FHS-capable GNUstep, this is something 
> which should be taken into account, IMHO.

Thanks ... it's certainly an aspect. :-)

There is also another important non-technical aspect, which is the amount of 
dependencies
that you have, and the robustness of your GNUstep configuration/installation 
process.

If we had gnustep-make depend on pkg-config, then you wouldn't be able to use 
GNUstep
unless you installed pkg-config first.

That seems a step backwards to me; adding more dependencies makes it more 
difficult
to install things.  End users are adversely affected.

At the moment, gnustep-make depends on nothing, and doesn't even need building. 
 You
just configure it and install.  That's very simple and very unlikely to break. 
:-)

That's key.  We should keep trying to simplify our configure/build/install 
process
so that more people that try to try GNUstep actually manage to try it ... they 
are
much more likely to stay. :-)

So I don't want to add extra dependencies that provide nothing of value just 
because
"way more developers know" about GNOME development tools than about GNUstep
development tools.  And if you go down that route, you'll end up using glade 
instead
of Gorm! ;-)

Thanks