httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ben Laurie <>
Subject Re: Apache 2000, err Apache 2.0 gets real
Date Mon, 02 Aug 1999 12:14:45 GMT
"Roy T. Fielding" wrote:
> >I'll try to make sure this never happens again. As I'm sure you know,
> >fighting emacs autoindent is hard work, and I haven't got the time for
> >it for such wholesale changes.
> Yep, that's why I don't use emacs (well, that and control-key-itis).
> >> protected the global names from collisions,
> >
> >Proposal? They are, after all, just function names. How are any global
> >functions protected from collisions (ap_ is used already, so you've lost
> >me slightly)?
> I meant the symbol names -- IMPLEMENT_HOOK is a pretty universal concept.
> Likewise for the other HOOK macros.  How about AP_HOOK_DO, or anything
> else that is short and specific to Apache.  I just don't want to make
> that kind of global change after setting up a nice clean repository.

Ah, okeydoke. I'll do it to the current versions.

> >> and included documentation,
> >
> >Let us know when you've finished. More seriously, if we could agree on
> >an in-code docco method, I'd probably do some. I can't be bothered with
> >parallel docco. Alternatively (since I can predict no-one else will
> >bother) I suppose I can just write comments.
> Well, I'll put it this way.  I can't be bothered to understand your code
> until you at least explain what it is doing, why it is doing it, and
> where others are supposed to use it.  If the only docs I have is the code
> itself, then the changes are not maintainable and won't be distributed
> as Apache.  This is one of the key parts of the Apache architecture you
> are replacing, so I expect at least what RST wrote for API.html.

Actually, the _architecture_ is mostly unchanged, in that the hooks get
called in the same places in more-or-less the same way, so the existing
documentation stands, with minor name changes. What has changed, of
course, is how modules hook into it, and how you implement a new hook.
These I will document in some way. Existing undocumented hooks are not
my responsibility and the introduction/non-introduction of a new hooking
mechanism does not influence their documentation state.

>  Actually,
> I expect a lot more than that, since in the past the lack of documentation
> has been a sore point, but what I won't allow is for the situation to get
> any worse.

Seriously, if you want this to improve, it is my view that inline docco
is needed. There's lots of ways to do this. We should choose one.




"My grandfather once told me that there are two kinds of people: those
who work and those who take the credit. He told me to try to be in the
first group; there was less competition there."
     - Indira Gandhi

View raw message