You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@httpd.apache.org by André Malo <nd...@perlig.de> on 2003/01/13 01:13:00 UTC
move RFC 1413 out of core?
hmm, I'd like to move the ident code out of core into a separate module,
say "mod_ident", since it's a feature, which is hardly used in more than 1%
of the cases, where the apache httpd will be used. But before doing any
effort: are there any objections on this?
implementation planned that way: mod_ident registers an optional function
in core, which will be called from the ap_get_remote_logname if available.
(that api function should probably stay in core, if other modules rely on
it, shouldn't it?)
It also should handle the %...l logformat in mod_log_config, so that this
specific logformat function will also moved to mod_ident.
Opinions?
nd
--
print "Just Another Perl Hacker";
# André Malo, <http://www.perlig.de/> #
Re: Review desired - mod_ident
Posted by André Malo <nd...@perlig.de>.
* Justin Erenkrantz wrote:
> --On Thursday, January 16, 2003 5:09 PM +0100 André Malo
> <nd...@perlig.de> wrote:
>
>> Another question: I guess, the change does require an MMN bump,
>> doesn't it?
>
> Well, I'm not real sure if we consider core_dir_config a part of our
> public structures. I really don't see why a module would be playing
> around in there.
Oh, I didn't think about the core_dir_config. I thought about the (then)
missing ap_rfc1413 function (and the global timeout variable, anyway).
nd
--
print "Just Another Perl Hacker";
# André Malo, <http://www.perlig.de/> #
Re: Review desired - mod_ident
Posted by Greg Ames <gr...@apache.org>.
Justin Erenkrantz wrote:
> --On Thursday, January 16, 2003 5:09 PM +0100 André Malo <nd...@perlig.de>
> wrote:
>
>> Another question: I guess, the change does require an MMN bump,
>> doesn't it?
>
>
> Well, I'm not real sure if we consider core_dir_config a part of our
> public structures. I really don't see why a module would be playing
> around in there.
>
> Anyone? -- justin
I agree with Justin. IMO any core_*_config is private to the core. I recall us
putting stuff in the core_request_config to keep it out of the request_rec for
example. Dunno if any modules are being bad.
Greg
Re: Review desired - mod_ident
Posted by Justin Erenkrantz <je...@apache.org>.
--On Thursday, January 16, 2003 5:09 PM +0100 André Malo
<nd...@perlig.de> wrote:
> Another question: I guess, the change does require an MMN bump,
> doesn't it?
Well, I'm not real sure if we consider core_dir_config a part of our
public structures. I really don't see why a module would be playing
around in there.
Anyone? -- justin
Re: Review desired - mod_ident
Posted by André Malo <nd...@perlig.de>.
* Justin Erenkrantz wrote:
> --On Wednesday, January 15, 2003 2:32 AM +0100 André Malo
> <nd...@perlig.de> wrote:
>
>> I'd like to let someone review before committing, because it's at
>> least a bunch of core changes.
>
> Looks good upon cursory review. I think we could optimize some of
> the socket calls (short read/write cases), but that was there before.
> We could always clean that up later. (sscanf - ugh!)
>
> My only real complaint is the tabs/formatting. =) There seem to be
> lots of whitespace issues. (Again, there before, but we ought to
> clean it up when we place it in the new location.)
ah well, c&p from rfc1413.c ;-)
I'll run a tab killer and whitespace cleanup before committing. For better
reproducing of the changes, I think, the code cleanup is better to apply
later.
Another question: I guess, the change does require an MMN bump, doesn't it?
nd
--
Da fällt mir ein, wieso gibt es eigentlich in Unicode kein
"i" mit einem Herzchen als Tüpfelchen? Das wär sooo süüss!
-- Björn Höhrmann in darw
Re: Review desired - mod_ident (was: move RFC 1413 out of
core?)
Posted by Justin Erenkrantz <je...@apache.org>.
--On Wednesday, January 15, 2003 2:32 AM +0100 André Malo
<nd...@perlig.de> wrote:
> I'd like to let someone review before committing, because it's at
> least a bunch of core changes.
Looks good upon cursory review. I think we could optimize some of
the socket calls (short read/write cases), but that was there before.
We could always clean that up later. (sscanf - ugh!)
My only real complaint is the tabs/formatting. =) There seem to be
lots of whitespace issues. (Again, there before, but we ought to
clean it up when we place it in the new location.) -- justin
Review desired - mod_ident (was: move RFC 1413 out of core?)
Posted by André Malo <nd...@perlig.de>.
* Justin Erenkrantz wrote:
> --On Monday, January 13, 2003 1:13 AM +0100 André Malo <nd...@perlig.de>
> wrote:
>
>> hmm, I'd like to move the ident code out of core into a separate
>> module, say "mod_ident", since it's a feature, which is hardly
>> used in more than 1% of the cases, where the apache httpd will be
>> used. But before doing any effort: are there any objections on
>> this?
>
> Sounds decent. I wouldn't even turn it on by default.
>
> Of course, this is 2.1+ only. -- justin
well, been there, done that ;-)
I've attached the full patch against HEAD. Changes to the rfc1413.c/core
code:
- added IdentityCheckTimeout directive (defaults to 30)
- the ident lookup is plugged via an optional function
- the function ap_rfc1413 no longer exists (to avoid failures, resp. force
compile errors if someone trys)
- -> removed rfc1413.c and rfc1413.h
- put mod_ident.c into modules/metadata (ok? or is there a better place?)
I'd like to let someone review before committing, because it's at least a
bunch of core changes.
(module documentation follows then.)
nd
--
s s^saaaaaoaaaoaaaaooooaaoaaaomaaaa a alataa aaoat a a
a maoaa a laoata a oia a o a m a o alaoooat aaool aaoaa
matooololaaatoto aaa o a o ms;s;\s;s;g;y;s;:;s;y#mailto: #
\51/\134\137| http://www.perlig.de #;print;# > nd@perlig.de
Re: move RFC 1413 out of core?
Posted by Justin Erenkrantz <je...@apache.org>.
--On Monday, January 13, 2003 1:13 AM +0100 André Malo <nd...@perlig.de>
wrote:
> hmm, I'd like to move the ident code out of core into a separate
> module, say "mod_ident", since it's a feature, which is hardly
> used in more than 1% of the cases, where the apache httpd will be
> used. But before doing any effort: are there any objections on
> this?
Sounds decent. I wouldn't even turn it on by default.
Of course, this is 2.1+ only. -- justin
Re: move RFC 1413 out of core?
Posted by André Malo <nd...@perlig.de>.
* Mads Toftum wrote:
> On Mon, Jan 13, 2003 at 02:19:26AM +0100, André Malo wrote:
>> oh darn, that's a weird definition. Ok, then I think, I'll leave it in
>> mod_log_config (which uses the api function anyway). Thanks for the hint
>> ;-)
>>
> Couldn't this be fixed by either changing the logformat to have - in that
> place, or by just substituting a - whenever ident is not turned on?
> Having a module for it might be a bit overkill though ;)
the functionality of "determine the ident user" is not in mod_log_config
(of course) but the definition of the "%l" format itself. My first thought
was to move the format definition also to mod_ident, which is - as Joshua
pointed out - not a good choice, since the %l format will be needed by
default.
So it will stay in mod_log_config, which uses the already existing api call
to get the value. Then it doesn't matter if mod_ident is present or not,
because if the api call returns no value it will be automatically set to
"-" in the log (which is the current behaviour, too). The format definition
code is quite small, so the little overhead in mod_log_config doesn't
really count.
nd
--
"Die Untergeschosse der Sempergalerie bleiben währenddessen aus
statistischen Gründen geflutet." -- Spiegel Online
Re: move RFC 1413 out of core?
Posted by Mads Toftum <ma...@toftum.dk>.
On Mon, Jan 13, 2003 at 02:19:26AM +0100, André Malo wrote:
> oh darn, that's a weird definition. Ok, then I think, I'll leave it in
> mod_log_config (which uses the api function anyway). Thanks for the hint
> ;-)
>
Couldn't this be fixed by either changing the logformat to have - in that
place, or by just substituting a - whenever ident is not turned on?
Having a module for it might be a bit overkill though ;)
vh
Mads Toftum
--
`Darn it, who spiked my coffee with water?!' - lwall
Re: move RFC 1413 out of core?
Posted by André Malo <nd...@perlig.de>.
* Joshua Slive wrote:
> On Mon, 13 Jan 2003, André Malo wrote:
>> It also should handle the %...l logformat in mod_log_config, so that this
>> specific logformat function will also moved to mod_ident.
>
> The only problem is that %l, even though it is almost always empty, is
> included in the common log format. If we create mod_ident and then don't
> include it by default, the CLF won't work in the default config.
oh darn, that's a weird definition. Ok, then I think, I'll leave it in
mod_log_config (which uses the api function anyway). Thanks for the hint
;-)
nd
--
>kann mir jemand sagen, was genau @-Domains sind?
Ein Mythos. Ein Werbetrick. Verarsche. Nenn es wie du willst...
-- Alexandra Buss und Björn Höhrmann in dciwam
Re: move RFC 1413 out of core?
Posted by Joshua Slive <jo...@slive.ca>.
On Mon, 13 Jan 2003, André Malo wrote:
> It also should handle the %...l logformat in mod_log_config, so that this
> specific logformat function will also moved to mod_ident.
The only problem is that %l, even though it is almost always empty, is
included in the common log format. If we create mod_ident and then don't
include it by default, the CLF won't work in the default config.
Joshua.