>
> It'd be no problem to make up a ppm package of Embperl based
> on the cvs version - I'll do that over the weekend ... To avoid
> confusion as to versions, would you recommend, Gerald, replacing
> the current Embperl ppm package I now have with one based on the
> cvs sources (I do that for mod_
On Thu, 1 Nov 2001, Gerald Richter wrote:
> >
> > It seems $fdat stop working after a few requests.. this varys..
> > I use a system like this: http://index.html?id=p34|l2 (id id=page | 34
> language = 2
> > this works works for several hits and than suddenly fdat{id} is empty.
> >
>
> Do you
>
> I take it CVS is something complicated?
No, the only thing is that you need a compiler. You can download a snapshot
and just install it just like the normal release. See INSTALL.pod for
detailed instructions how to compile it for ActiveState Perl (You need MS
VC++ 6.0, because ActiveState Per
Hi!
>
> Which is more expensive (memory and processing time)
>
There is no (real) difference in memory usage
> I want to use some functions on two pages, but one, say A is very rarely
> executed, B is very heavily. Is it better to have embperl subs in B and
use
> them in A throught Execute with
Okay I think I better wait and see if Randy can make a ppm file of the latest version.
For now I just leave the "maxrequestsperchild" at 5.. it seems to be working okay..
but very slow..
I take it CVS is something complicated? I do remeber I saw Apache::Xsession somewhere
with ppm files I am
Hi!
Which is more expensive (memory and processing time)
I want to use some functions on two pages, but one, say A is very rarely
executed, B is very heavily. Is it better to have embperl subs in B and use
them in A throught Execute with import=>1 or put subs into C and import to
both?
Or write
>
> It seems $fdat stop working after a few requests.. this varys..
> I use a system like this: http://index.html?id=p34|l2 (id id=page | 34
language = 2
> this works works for several hits and than suddenly fdat{id} is empty.
>
Do you do any other requests inbetween ? May some request that in
>
> Something I've found useful in vaguely similar situations is to do your
> base.epl Execute with an 'output' argument, which stuffs it all into a
> scalar for you instead of outputting it directly. You can then check
> whether the include is a full html page or just a fragment and munge
> appro
