Tom Christiansen writes: > We've been down this route. It doesn't help the way you think it does. > These are merely wafer-thin wrappers about syscalls. It's Perl's > complete infrastructure support system you're seeing, and that you > will not reduce. Actually, if we can split compiler from runtime, Apache could choose to embed only the runtime engine (no eval()s permitted, obviously) with a corresponding drop in memory consumption. I don't know how you'd find out what amount of /usr/bin/perl5 is runtime vs compiletime. But I'd like to know. Nat
- New Perl rewrite - embedded Perl Matthew Gillman
- Re: New Perl rewrite - embedded Perl Dan Sugalski
- Re: New Perl rewrite - embedded Perl ye, wei
- Re: New Perl rewrite - embedded Perl Tom Christiansen
- Re: New Perl rewrite - embedded Perl ye, wei
- Re: New Perl rewrite - embedded Perl Tom Christiansen
- Re: New Perl rewrite - embedded Pe... Nathan Torkington
- Re: New Perl rewrite - embedde... Dan Sugalski
- Re: New Perl rewrite - embedded Perl Ken Fox
- Re: New Perl rewrite - embedded Perl Russ Allbery
- Re: New Perl rewrite - embedded Perl Jarkko Hietaniemi
- Re: New Perl rewrite - embedded Perl Dan Sugalski
- Re: New Perl rewrite - embedded Pe... Bennett Todd
- Re: New Perl rewrite - embedde... Russ Allbery
- Re: New Perl rewrite - embedde... Hildo Biersma
- Re: New Perl rewrite - emb... Bennett Todd
- Re: New Perl rewrite - emb... Dan Sugalski