Message-ID: <36078EE8.7A3A726C@Golux.Com>
Date: Tue, 22 Sep 1998 07:50:00 -0400
From: Rodent of Unusual Size <Ken.Coar@Golux.Com>
Organization: The Apache Group
MIME-Version: 1.0
To: new-httpd@apache.org
Subject: Re: cvs commit: apache-1.3 WARNING-NT.TXT
References: <9809211416.aa28345@paris.ics.uci.edu>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: new-httpd-owner@apache.org
Precedence: bulk
Reply-To: new-httpd@apache.org

Amazing.  For once I wasn't part of this slugfest.  Not that I feel left
out, mind you -- it's rather refreshing to be on the sidelines. :-)

Roy T. Fielding wrote:
> 
> >I would have thought that removing a part of Apache code should be treated
> >the same as adding code, in that both are changes to the same rules. Thus
> >the removal of code is subject to both commit-then-review rules, and
> >subject to potential veto.
> 
> If we are talking about a feature, sure.  But this is documentation.
> Why should we include misleading documentation in our product just
> because one person thinks it still applies?  I wouldn't mind it if
> there was enough information provided so that I can know when it will
> be removed in the future.

It looks like it's not just one person who feels that way.  Both Paul
and Marc do.  So do I, to a lesser extent.  On the other hand, why should
we remove an agreed-upon warning just because one person feels it's
misleading and *doesn't* apply?

I would say that if there are significant pieces of functionality
that work on Unix but not on Win32, then we are justified in
calling the Win32 release 'not ready for prime time.'

Two items I *know* are broken in Win32:

1. "Alias /foo/ z:/" doesn't work.  Ken Parzygnat has determined that
   this is due to a bug in the (you guessed it) filename canonicalisation
   routines, and I think he has a fix.
2. The whole extension-versus-shebang issue.  Bill Stoddard has stuff
   for this; I think the brouhaha kept it from being committed.

Things I *think* are broken:

1. The proxy.  (I may be wrong; has this been made to compile/work
   on Win32?)

Things that IMO should be changed, and hence are semi-broken:

1. All the [mis]features {en,dis}abled through compile-time -D
   flags ought to be moved to run-time configuration directives.
   We've agreed long since that building on Win32 is not something
   we expect people to do as a rule, and this restriction makes
   Win32 users second-class citizens.

Let the people who have objections to removing the 'beta' label
identify the specifics that they feel need to be fixed for them
to approve its removal.  Mine are above.  These should probably
go into the STATUS file.

I'm also in favour of the warning showing up in its own separate screen,
not as part of the README, during the installation.  The tone
could be softened so as to not scare people off, but I personally
don't think the Win32 version is out of the woods yet.  I feel
uncomfortable labeling it as being of the same quality as the
Unix release.  Maybe gamma rather than beta? :-)

Back to the safe sidelines..

#ken	P-)}

Ken Coar                    <http://Web.Golux.Com/coar/>
Apache Group member         <http://www.apache.org/>
"Apache Server for Dummies" <http://Web.Golux.Com/coar/ASFD/>