>>>> "Hrvoje" == Hrvoje Niksic
<hniksic(a)xemacs.org> writes:
Hrvoje> Michael Sperber <sperber(a)informatik.uni-tuebingen.de> writes:
>>
>>>> "Hrvoje" ==
Hrvoje Niksic <hniksic(a)xemacs.org> writes:
Hrvoje> >
Hrvoje> Personally I still prefer Kyle's original proposal of
Hrvoje> "dumping" C code and linking against it as part of the
Hrvoje> XEmacs build process (which would be *really* portable), but
Hrvoje> I could live with the current pdump architecture as well.
Hrvoje> >
> That's nice in theory---in practice, there are complications
because
> many C compilers run quadratically in the size of things like array
> initializers.
Hrvoje> But you don't have to initialize everything in one array. Or did I
Hrvoje> misunderstand you? Anyway, I've seen code that pulls it off, so I
Hrvoje> know it's possible in practice as well as in theory. I agree that it
Hrvoje> might be harder with something of the size of XEmacs.
Yeah, me too. (Scheme 48 and scsh do this.) But splitting your data
into several arrays is annoying.
--
Cheers =8-} Mike
Friede, Völkerverständigung und überhaupt blabla