Re: Compiling GNUstep on darwin (again)

[Nicola Pero]

> Seems like I am the only one who is struggling at the NeXT runtime front. 
> Since I am just a little java-programmer, I am often lost in the jungle 
> when it comes to difficult problems what in return means that there is 
> only very slow progress made.
> 
> Since you say that GNUstep runs on Darwin using the GNU runtime, I ask my 
> self if it is worth to keep the effort of porting GNUstep to the NeXT 
> runtime. As I can see from my current point of view that would possibly 
> mean to make some incisive changes to some of the GNUstep Classes since 
> the NeXT runtime works just differently (For instance I think it is not 
> possible to send messages to itself during initializing classes since the 
> NeXT runtime keeps everything in a cache and looks there first before it 
> sends a message. If the class is not in the cache it tries to initialize 
> it. So if you send messages during initializing it looks again and tries 
> to initialize again over and over).
> 
> So what do you think? Keep the effort or just stop and be happy with what 
> we have (GNU runtime with gcc 3.1 which comes out with jagu-ire). I don't 
> think that I am able to cope the thing alone, I am just not experienced 
> enough (and miss some deeper insight into the design principles of 
> GNUstep).

If you feel you're wasting your time and not making any actual progress
(or very little progress compared to the effort you are putting into it),
well yes I would second leaving it alone :-) and maybe trying to do
something different, where you feel you can get a better reward (in terms
of results and satisfaction) for the efforts you make. :-)

Thanks for your support and contribution !