perl-modperl-cvs mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Doug MacEachern <>
Subject Re: cvs commit: modperl-2.0/lib/ModPerl
Date Fri, 14 Apr 2000 22:43:06 GMT
> I don't want to get involved in an unproductive semi-religious war or
> anything

so then why did you?

>, but I'd just like to caution against change for (what seems
> little more than) the sake of change.

re-read my message, it is for more than the sake of change.
> I think it was our Lord Chancellor in about 1980 who said:
> "If a thing ain't broke, don't mend it".

i consider ~30Mb processes broken, i plan to do everything in my power to
fix that.  e.g. "ain't broke", it does what it says it will
do.  what it doesn't tell you is how much bloat it adds to the runtime.
as far as i'm concerned, that needs mending.  mod_perl-2.0 will either
have it's own version of Exporter written in C (which is experimental in
1.x) or will not export anything.  i haven't decided which yet, but i have
decided that the standard will not be used.
> The issue of compile and execution times is a red herring.

says who?

> If you care that much about CPU cycles you'll at least get out
> Kernighan & Ritchie.

really?  does that explain how Perl is implemented internally and what i
can do to squeeze the best performance out of Perl?
> One of the most frequent causes of software failures is fiddling about
> with the code to make it prettier or to satisfy some other lust.  If
> it works and has been tested by thousands of people, please leave it
> alone unless you have a *very* good reason to brea^H^H^H^H touch it:)

this has nothing to do with being pretty or satisfying lust.
modperl-2.0 is a complete re-write, it's only been tested by me so far.
regardless, using 'constant string' instead of "constant string" isn't
going to break anything.

View raw message