Re: CVS update: subversion/subversion/libsvn_fs/tests skel-test.c

On Sun, Oct 08, 2000 at 12:40:49PM -0500, Jim Blandy wrote:
>...
> It looks to me like most calls to apr_register_cleanup are just going
> to pass apr_null_cleanup as the exec handler. Rather than having a
> useless argument in every call, wouldn't it be better to use a
> separate function for registering exec cleanups?

But that isn't the Way It Has Been Done(tm)

:-)

Actually, that makes very good sense. I took a look thru the Apache code and
most of them *do* pass apr_null_cleanup. However, the cleanup code in the
pools lends itself better to the dual-function that we have now. I'd need a
bit more thinking to figure out whether they're easily separable (I'm
thinking they are).

The bigger issue may be changing the prototype. I can update Apache and SVN
no problem, but it will impact third party module writers. Of course, they
have a number of issues already with porting to Apache 2.0, so this may not
be a big deal :-)

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/