On Thu, 2004-09-23 at 20:11, Mark Stosberg wrote:
> > The order is only important in rare cases, for example: A database
> > connection needs to be closed at the teardown stage, but maybe the
> > session object needs to be flushed at the teardown stage as well... in
> > this case, the order can be important if the session uses the database,
> > so the database would register a hook at the teardown stage as LAST or
> > even REALLY_LAST since it is a special case. I'd be happy enough to
> > drop the REALLY_LAST and REALLY_FIRST sections to simplify things, since
> > I think it would work fine with only 3 options (FIRST, MIDDLE, LAST), or
> > (FIRST, DONTCARE, LAST).
>
> For reference, I looked up some the Apache documentation about this.
> Here's one page I found:
>
> http://httpd.apache.org/docs-2.0/developer/hooks.html
>
> I noticed that in general the "really_first" and "really_last" positions
> weren't highlighted. I think I would like to drop them, at least
> initially. MIDDLE is always described as the "don't care" option, thus I
> think is better named DONTCARE. (I realize I'm just re-stating what Cees
> has stated are reasonablie ideas above).
I've been pondering this for a while now and I have come to two
conclusions:
1. Making DONTCARE a synonym for MIDDLE is a very good idea. I think
that DONTCARE is more meaningful in a lot of ways but I don't think we
should drop MIDDLE. Maybe consistently use DONTCARE in the docs and just
leave MIDDLE in the code for people that are familiar with Apache. Maybe
make a small note about it being synonym in the docs but that's it.
2. We should leave REALLY_FIRST and REALLY_LAST in the code and maybe
make a small note about it in the docs but don't make a big deal about
using them. There are perfectly good reasons for using them, they should
be used sparingly though. Maybe like MT does with priority 0 and 11 we
should throw an error if more that one callback is registered to them.
--
Tony Fraser
tony@xxxxxxxxxxxxx
Sybaspace Internet Solutions System Administrator
phone: (250) 246-5368 fax: (250) 246-5398
---------------------------------------------------------------------
Web Archive: http://www.mail-archive.com/cgiapp@xxxxxxxxxxxxxxxxx/
http://marc.theaimsgroup.com/?l=cgiapp&r=1&w=2
To unsubscribe, e-mail: cgiapp-unsubscribe@xxxxxxxxxxxxxxxxx
For additional commands, e-mail: cgiapp-help@xxxxxxxxxxxxxxxxx
|
