On Mon, Nov 24, 2003 at 08:40:16PM -0800, Edward S. Peschko wrote:
> btw - what's the best way to trace through par to see what its doing ?

In debug mode, break at PAR::find_par and start from there?

> For example, if you compile some Inline'd code, and then move the executable that 
> you compiled into a different location, then the executable bails with:
> 
> sh: syntax error at line 1: '(' unexpected.
> 
> And autoloader doesn't seem to like CODE references  very much, I'm
> getting statements like:

Ho hum.  Normally AutoLoader works just fine with PAR, iff we pack all
the .al files along.  I hadn't tried Inline'd code -- and since Brian
mentioned that Inline will soon produce intermediate XS code anyway,
I'm not highly motivated to hack PAR into support the current Inline.pm.

But I'll check it in a couple days when I have time.

> (
> ps - all-in-all the whole coderef thing seems to be a tenuous fit at best, if 
> you simply packed perl up, tacked on the modules at the end and untarred them into
> a specified location you could probably avoid lots of the headaches, but then 
> again, I'm sort of hesitant to do that. I'd rather make the present scheme work
> for all cases.
> )

That is true.  But that's cumbersome and in some cases inpractical for
an end user to do.  Also, nothing's preventing you from:

    unzip pp_produced.exe;
    perl -Ilib script/main.pl

Which amounts to the same thing. :-)

Thanks,
/Autrijus/

Attachment: pgp00000.pgp
Description: PGP signature

Reply via email to