Best policy is to use switch_ and not use the lazy apr_.
I tend to agree, at least from the standpoint of a non-coder who wants (or needs) to read the source code.
I cannot think of a good reason, maybe others can -- but,
I can't either. To preserve modularity, and to make it easier to migrate away from apr if that need should arise, a clean and consistent namespace seems appropriate.
If you find occurrences of non-conformance, patches are welcome!
Sounds good. I'm studying up on apr (and thusly switch_xxx) and if/when I'm comfortable submitting patches I will most definitely do so. -MC
m
On Thu, 8 Nov 2007, Michael Collins wrote:
Guys,
bkw gave me some homework and now I have a few questions. I notice
that
in switch-types.h there are all sorts of typedef struct statements that create a switch_ version of the corresponding apr_ type. However, I've also noticed that scattered throughout the source that there are numerous occasions where the apr_ namespace is used even when there is a switch_ equivalent available. Example: using apr_pool_t instead of switch_memory_pool_t in mod_openmrcp.c.
Questions: Is it FS standard policy/best practices always to use the switch_ namespace? Are there legitimate situations where one would prefer apr_ instead of switch_?
Thanks,
MC
P.S. - If there are cases where apr_ is used when there are switch_ equivalents, does that mean there is an impending codebase cleanup project to make everything consistent?
_______________________________________________ Freeswitch-dev mailing list Freeswitch-dev@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-dev UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-dev http://www.freeswitch.org