httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Roy T. Fielding" <>
Subject Re: ap_ vs apr_ revisited.
Date Tue, 25 May 1999 00:03:40 GMT
>If you voted the last time, please vote again so I don't have to dig
>through archives for your vote last time.

I am not inclined to vote on this kind of thing, but I guess I should
mention my preference and give some rationale.

>I would like to see the prefix for apr functions to be:
>    [X] ap_       [ ] apr_        [ ] I don't care.

The big problem with meaningful prefix names is that if we later
decide that a function belongs in some category other than APR, or
that some function left out of APR should be moved over, then we
have to change the name of the function and provide a backwards
compatibility define and otherwise muck up the code for what really
is a rather pointless exercise in futility.

It has been my belief all along that the only reason to prefix names
is to prevent name conflicts with other people's code.  Whether or not
APR is done as a separate project, there exists no scenario in which
we would allow ap_foo_bar and apr_foo_bar to coexist in our codebase
and not point to the same function, since to do otherwise invites
suicide on the part of some hapless code readers.  These projects have
the same namespace, for all practical purposes, regardless of how we
prefix the names.  Maybe if you called it moccasin I'd feel differently,
but I'm too lazy a typist to suggest that as a prefix.

Having said that, I also need to say that I normally wouldn't care how
this particular argument is settled, but it should never be propagated
to the module developers list for voting.  This is the only mailing list
that is allowed to make technical decisions for the Apache HTTP Server
Project, period.  If someone is not on this mailing list and fully informed
by its discussions and entertainment, then they aren't qualified to have an
opinion on any technical issue regarding Apache, no matter how well
intentioned and intelligent they may be.  Asking for a person's opinion
without the full context of this list is about as useful as a lawn dart
contest at a football game, and almost as dangerous to those of us
on the field.

In any case, I suggest that the people actively working on APR just
go with something they feel comfortable with right now and ignore the
long term implications until we have to worry about a release.  Be happy.


View raw message