Re: [patch] require 5.6.1 in autogenerated .pm files

Doug MacEachern wrote: > On Mon, 4 Mar 2002, Stas Bekman wrote: > > >>since we use 'our' in the autogenerated .pm files, requiring 5.6.1 is >>probably a good idea: >> > > no need. .pm's that have a .xs are installed in $version/$arch > so anybody using < 5.006 would have to try really hard to

Re: [Patch] gcc3 -Wall : 'deprecated use of label at end of compound statement'

Philippe M. Chiasson wrote: > gcc3 seems to have changed -Wall behaviour a bit and now considers that > the 'use of label at end of compound statement' should be deprecated. > > Small patch to fix this and everything builds find with gcc3 thanks gozer, it should be fixed now. the ignores patch

perl-5.7.3 is out

perl-5.7.3 is out. It passes all tests with threads enabled flawlessly :) So if you were afraid to upgrade your bleadperl it's good time to do it now. Though it seems that LWP now has something to solve (this is with the latest version 5.64): Use of inherited AUTOLOAD for non-method HTTP::Respo

list of contributors/ changes log

[crossposting dev list, since it's relevant] one of the issues to be resolved on the new mod_perl site (read: documentation) is to provide a mechanism for generating the Changes and (optionally Contributors) list. With mod_perl 1.x source this has been done manually, and the same with the mod_

Re: usage of safemalloc

On Wed, 6 Mar 2002, Stas Bekman wrote: > Doug, didn't you say that we should always use apr_.?alloc functions for > malloc'ing? Or is it OK to use it? yes, in general, apr_.?alloc should be used. > I see it used twice in the mod_perl sources: > > 1. in modperl_perl_destruct() in src/modules/p

Re: [patch] better diagnostics on typemap conversion failure inModPerl/TypeMap.pm

On Tue, 5 Mar 2002, Stas Bekman wrote: > Currently if you use an unknown type with xs_generate you get: > > writing...WrapXS/typemap > Odd number of elements in hash assignment at lib/ModPerl/TypeMap.pm line > 151. > > which is not very helpful and just shows that the code is flowed. > > Here

RE: cvs commit: modperl/eg README httpd.conf.pl perl_sections.txt perl_sections_2.txt perlio.pl registry.pl startup.pl test.pl

On Mon, 4 Mar 2002, Geoffrey Young wrote: > I'd take care of it, but I didn't see a commit email come through for the addition >of the STATUS > file, even though I got the "mailing commit message..." message from log_accum.pl, >so maybe > something is up with my access yet... i think brian or

Re: [patch] require 5.6.1 in autogenerated .pm files

On Mon, 4 Mar 2002, Stas Bekman wrote: > since we use 'our' in the autogenerated .pm files, requiring 5.6.1 is > probably a good idea: no need. .pm's that have a .xs are installed in $version/$arch so anybody using < 5.006 would have to try really hard to shoot themselves in the foot here. t

Re: make source_scan doesn't work

On Mon, 4 Mar 2002, Stas Bekman wrote: > I've simply patched C::Scan not do die on the parsing problem. Also > added local() fix. This is against http://www.apache.org/~dougm/Scan.pm nice work. i've updated ~dougm/Scan.pm with your patch. -

usage of safemalloc

Doug, didn't you say that we should always use apr_.?alloc functions for malloc'ing? Or is it OK to use it? I see it used twice in the mod_perl sources: 1. in modperl_perl_destruct() in src/modules/perl/modperl_perl.c 2. in mpxs_apr_uuid_alloc() in xs/APR/UUID/APR__UUID.h why? Thanks! __