Hi Mike, thanks for the suggestions! Are you suggesting lua because of the process startup time of python? That sounds reasonable, if a bit unfortunate for us. We have python modules for globbing source modules (we have nearly 100 supported compiler/platform combinations). Perhaps in the longer term, we could embed tup into python as a module. That might also address the "tup without Tupfiles" issue also - the python script could specify some other file to look for to get build targets.
The reason I went for separate groups is to maximize the parallelism. If we have 100 libraries, a single group would prevent any compilation before the autogeneration has finished for all 100 libs if I understand correctly. Thats kind of a killer if an exe only depends on 2 of those libraries, say. It looks like the lua integration is worth a closer look - to be honest I just skimmed it because I didn't think it could do more than the normal rules. Perhaps the lua api won't go *all *the way though? It still wants files scattered around the filesystem to create targets. Or does it make any difference to tup if I generate a massive list of rules in the root tupfile? I suppose that could work. Stephen. -- -- tup-users mailing list email: [email protected] unsubscribe: [email protected] options: http://groups.google.com/group/tup-users?hl=en --- You received this message because you are subscribed to the Google Groups "tup-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
