Re^2: Does hv_delete() decrement the reference count of the SV * it returns?

Thanks tachyon!

Not to nitpick, but don't you think this is a little unacceptable? If this reference counting scheme is going to work, don't the docs have say which routines mess with their arguments refcounts and which don't?

Is there anyway I could get this info added to perlapi and perlguts? I've found other inconsistencies and omissions in the Perl C documentation too.

Comment on Re^2: Does hv_delete() decrement the reference count of the SV * it returns?

Replies are listed 'Best First'.
Re^3: Does hv_delete() decrement the reference count of the SV * it returns? by tachyon (Chancellor) on Oct 14, 2004 at 03:54 UTC
Perl 5 guts are inconsistent and perl's ref counting is somewhat sub optimal, this is part of the desire for a total overhaul. If you wanted to submit updates to the documentation you would send them as patches to p5p. Here are the patching guidelines Perlguts.pod is a real doc in the pod/ dir. Perlapi.pod is autogenerated by embed.pl. NB: embed.pl also autogenerates perlapi.c Read the header of perlapi.c to find the correct files to edit that eventually become perlapi.pod. cheers tachyon	[reply]

Replies are listed 'Best First'.

Re^3: Does hv_delete() decrement the reference count of the SV * it returns?
by tachyon (Chancellor) on Oct 14, 2004 at 03:54 UTC

Perl 5 guts are inconsistent and perl's ref counting is somewhat sub optimal, this is part of the desire for a total overhaul. If you wanted to submit updates to the documentation you would send them as patches to p5p. Here are the patching guidelines Perlguts.pod is a real doc in the pod/ dir. Perlapi.pod is autogenerated by embed.pl. NB: embed.pl also autogenerates perlapi.c Read the header of perlapi.c to find the correct files to edit that eventually become perlapi.pod.

cheers

tachyon

[reply]