2.0 docs todo:
---
mp2 Utilities need to be documented (a separate document?):
mp2bug - (should also be mentioned in user/help/help.pod)
*** General items ***
---
Add a note somewhere that explains the layout of the documentation, in
particular:
api/ - documents the API, fully covering all options.
user/ and devel/ are tutorials and don't necessarely cover all the
features of a discussed module. but should instead point to the API
manpage for a complete info.
The point is that we want to avoid duplication of the documenation, in
order to make it easier to maintain the docs and keep them in sync.
---
under
SetHandler modperl
the
Options +SetupEnv
has no effect. Instead either use perl-script or call
$r->subprocess_env in the void context.
---
On the cleanups from Bill Rowe's email (need to integrate this into
the cleanup docs, the hook that we don't have yet):
Cleanups are a LIFO stack. When a pool is freed, first all sub-pools
are freed. Then all cleanups within that pool are run (the cleanups
in the sub-pool already ran when those sub-pools were freed.)
So if you register a cleanup, then create an apr object in the same
pool, those apr object's cleanup runs before your registered cleanup.
If you create an apr object and register your own cleanup, you still
have that object until your registered cleanup is finished.
---
* this is probably a documenation issue. Several phases are of type VOID
and docs say:
Handlers of the type C<VOID> will be I<all> executed in the order they
have been registered disregarding their return values. Though in
mod_perl they are expected to return C<Apache::OK>.
may be we shouldn't mark them as VOID then, but RUN_ALL? this is in
user/handlers/intro.pod
---
* the issue with crypt_r: in 5.8.0's each perl clone allocates 128K
for crypt_data. This fixed in 5.8.1 and 5.9.0 (#19119 (maint) and
#19122 (blead)), however this could be quite a waste for those using
5.8.0. perhaps adding a note to performance chapter will do. For
more info see the thread:
http://archive.develooper.com/perl5-porters@perl.org/msg93846.html
---
---
---
---
---
---
---
---
---
---
---
---
---
---
