Brian Ingerson <[EMAIL PROTECTED]> writes:
[...]
> Actually, if working at ActiveState implied anything, it would imply
> that I'm a platform agnostic programmer. We have Linux, Solaris, AIX and
> HPUX ports of many products. At least of Perl. It should be obvious from
> reading my many postings and examples that I do most of my work on Unix.
> I have a VMWare Windows machine that I use once in a while to test
> Inline and other things.
OK, sorry about that cheap shot.
> > So here's my situation. At work I use Unices running on 4 different
> > architectures (MIPS R10k, "DEC" alpha, x86, UltraSPARC). ***BUT*** I
> > only have one home directory (it's stored on some other box, and each
> > computer mounts it via the NFS[1].
>
> I assumed NFS. My point was that .Inline/ does not need to be in your
> home dir. It could be in your project dir, which could be OS specific.
My project dires are never OS specific. Everything I write has to run
under (at least) these 4 platforms. Ideally it will also work under
other Unices; I just can't test that easily.
> Or, you could always do something like:
>
> use Inline Config => DIRECTORY => "/home/ariel/.Inline/$^O";
That's reasonable (though I'd prefer to use something that gave the
platform name, so it's %Config time again).
> or even this if you don't want to change your source code:
>
> export PERL_INLINE_DIRECTORY=~/.Inline/$OSTYPE
Now *that* I like!
> I don't want Inline directory names to grow any bigger since I spent so
> much time getting them small. And there seems to be plenty of
> alternatives here.
Sounds good to me.
[...]
> We have plenty of time to figure this out. I don't plan on implementing
> it for a while.
OK, it seems more discussion is warranted. Here's the complete
question as I understand it, for the mailing list:
Sometimes the *compilation* will succeed but Perl still won't be able
to load the generated module, due to incorrect configuration. An
example is if we incorrectly set the LIBS configuration option. Upon
fixing the configuration options, Inline won't recompile (because the
code hasn't changed) and we still won't be able to run the module.
What should happen? Should Inline know which configuration options
actually affect the compiled binaries? Do we know the list (LIBS is
on it, but what else)? Or should the developer use the FORCE config
option is these cases?
[...]
--
Ariel Scolnicov |"GCAAGAATTGAACTGTAG" | [EMAIL PROTECTED]
Compugen Ltd. |Tel: +972-2-5713025 (Jerusalem) \ We recycle all our Hz
72 Pinhas Rosen St. |Tel: +972-3-7658117 (Main office)`---------------------
Tel-Aviv 69512, ISRAEL |Fax: +972-3-7658555 http://3w.compugen.co.il/~ariels
