httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "David Reid" <dr...@jetnet.co.uk>
Subject Re: APR_LOCAL/APR_REMOTE
Date Mon, 06 Nov 2000 23:17:24 GMT
Hmm, I don't really mind which way, though I do feel passing a flag would be
"cleaner".  Adding a lot of internal functions just seems messier somehow...

Anyway, I'll get the patch done and see what we think...

david
----- Original Message -----
From: <rbb@covalent.net>
To: <new-httpd@apache.org>
Sent: Monday, November 06, 2000 8:23 PM
Subject: Re: APR_LOCAL/APR_REMOTE


> On Mon, 6 Nov 2000, Greg Stein wrote:
>
> > Using flags/parameters to differentiate semantic meaning is rather icky.
> > You're only saving one function, yet making the API more complicated and
> > cumbersome.
>
> I completely disagree.  If anything this makes the API cleaner.  You are
> setting address information.  That is done with one function, with a flag,
> not two functions.  This basically says all sockets have two sides, local
> and remote information.  If you want to set information about the socket,
> use one function to do it, specifying which part of the socket you are
> modifying.  This goes along with setsocketopt and getsocketopt.
>
> Ryan
>
> >
> > If the internals have duplicate code, then adjust the internals.
Reflecting
> > it out into the API is a backwards step.
> >
> > Please don't make this change :-)
>
>
____________________________________________________________________________
___
> Ryan Bloom                        rbb@apache.org
> 406 29th St.
> San Francisco, CA 94131
> --------------------------------------------------------------------------
-----
>
>


Mime
View raw message