Re: [GSWHackers] Re: statistics

[David Ayers]

Nicola Pero wrote:

> > I'm not sure who you meant by 'you'... but if you meant people in general...
> > I'm sure most people know be my earlier posts, but to shed some more 
> > light... I'm working on makefiles and headers to build two seperate sets 
> > of GSWeb frameworks with clean namespaces. I already have a running 
> > version but I'll wait to submit it until I've got it's usability to be 
> > very easy, straight forward and configurable. This also depends on work 
> > that is currently being done by Nicola to support multiple 
> > FRAMEWORK_NAME entries in makefiles.
> >    
> This might take some time before it's done.
>
> You can still implement it without having the multiple FRAMEWORK_NAME
> entries.  For example (untested code, just drafting to give an idea) -
Thanks Nicola...

> [... snip ...]
> Please note that this is quite Ok :-) ... except it's going to be less
> efficient than having support for multiple FRAMEWORK_NAMEs in gnustep-make
> - but I don't think you should care as we're talking of little
> inefficiencies.
I'm not worried about effiency, I'm just worried about usability and 
maintainability.
> You can also have it only at the top-level - if you have a single
> top-level GNUmakefile which uses aggregate.make and SUBPROJECTS, it's much
> simpler and nicer -
> [... snip ...]
> This is quite nicer.  I'd definitely suggest using this second way if you
> can.
Well in this case I do have an aggrefate.make but I would probably have 
to extract the mechanism into a makefile fragment so that make 
invocations in the subdirectories also work correctly and all use the 
same configured default.
If this is important to someone else I'll gladly look into preparing it 
ASAP. But as it seems everyone else seems contempt with the current 
hybrid framworks and I have local solution, there seems to be no rush. 
I'd rather wait to finish the final patch and invest my time on a 
different constructin site right now (improving EOQualifer qualifier 
format parsing).
Or is there someone actually waiting for this namespace split?

Cheers,
Dave