Date: Mon, 2 Feb 1998 21:07:24 -0700 (MST)
From: Marc Slemko <marcs@worldgate.com>
To: new-httpd@apache.org
Subject: Re: apache/linux modules
In-Reply-To: <Pine.LNX.3.96.980202225342.26063Q-100000@localhost.localnet>
Message-ID: <Pine.BSF.3.95q.980202210405.6464C-100000@valis.worldgate.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: new-httpd-owner@apache.org
Precedence: bulk
Reply-To: new-httpd@apache.org

On Mon, 2 Feb 1998, Cristian Gafton wrote:

> On Mon, 2 Feb 1998, Marc Slemko wrote:
> 
> > You are right, no one has really cared.  Because no one here seems to use
> > it.  We haven't had much in the way of people building third party
> > distributions in the list speaking up about what they need. 
> 
> That's beyond muy point of understanding. Everybody (every third party
> shipping customized apache server) preferred t go through the pain of
> maiking the user install a new server, try to merge it's changes for the
> config files into the new ones provided, etc, instead of preferring to ask
> the user "what version of apache are you running and on what platform" and
> upon receiving the answer ship a shared object with a five-liner install
> script and make adding a new, very cool apache feature a very easy task ?
> 
> You mean, there were no such vendors screaming loud about having such a
> nifty feture ?

Nope.  I didn't hear anyone from RedHat saying a word.  I didn't hear any
other vendor saying a word[1]

> Then maybe is not such a cool thing to have, after all.

No, more like vendors just think we are a hidden unresponsive mass of
something that can't be altered so they have to.  Or that they would lose
their cool "edge" if they didn't hack it in themself.  Vendor
contributions from packaging efforts (or any other efforts) have been very
minimal. 

On Mon, 2 Feb 1998, Cristian Gafton wrote:

> So - dynamic stuff for 1.3 - is that all right ?

I think there is general agreement in concept that it is reasonable and
good, although people may not go for the changes required to autoload
modules based on anything except the filename for the name of the module
symbol, but until the actual code is in our faces... no once can give any
definite answer. 

[1] I think someone said a bit about the FreeBSD port, which probably
pushed the ifdefs in faster...