Date: Thu, 9 Jan 1997 22:20:07 -0700 (MST)
From: Marc Slemko <marcs@znep.com>
To: new-httpd@hyperreal.com
Subject: Re: 1.3 veto ?
In-Reply-To: <Pine.BSF.3.91.970110035437.14526A-100000@localhost>
Message-ID: <Pine.BSF.3.95.970109221029.1363C-100000@alive.ampr.ab.ca>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: new-httpd-owner@apache.org
Precedence: bulk
Reply-To: new-httpd@hyperreal.com

On Fri, 10 Jan 1997, Rob Hartill wrote:

> On Thu, 9 Jan 1997, Chuck Murcko wrote:
> 
> > +1 for this. I agree with Alexei. We're fundamentally changing the core code
> > for 2.0, and the faster we do that, the faster we can move on with
> > enhancements. The proxy could use a lot of stuff now, but it will only get
> > bug fixes until 2.0, since it has to become HTTP/1.1 compliant first of all.
> > The proxy goal for 1.2 was to become solid for the protocols it supports, and
> > that's done. We really need to get 1.2 solidified and out the door, at this
> > point.
> 
> If 2.0 can be turned around in the short time claimed, then that's
> great. As Alexei says, 1.2's 1 month turned into 6. What will 2.0's
> 6 months turn into ?
> 
> <rant>
> Nobody seems to take any notice of timetables - that's the major problem IMO.
> Another problem is the sheer lack of input/feedback from 90% of the people
> on this list. Unless tey are contributing in some other way, I think the
> least everyone can do is testcode snapshots every week or two, even if it's
> just to see if things compile. 1.2 would be released by know had it gotten
> the wider testing it could have.

I think the Apache development team is extremely small for the number of
users Apache has.  That certainly may be partly due to the type of
software it is, but I think it is also largely due to the fact that
someone needs to know the right white magic to join the mailing list, etc. 
Offhand, I can't recall anywhere that interested and capable people are
invited to help in development or even told how to if they really really
want to.  This isn't necessary a bad thing, and I have seen advantages to
the smallish membership this list has in the short time that I have been
subscribed, but it does likely result in less development. 

> What makes people think things are going to be better developing 2.0 ?
> 
> Is anyone prepared to set a timetable *and* make sure we stick to it ?
> If there is, we might not need 1.3, if there isn't .....
> </rant>

Perhaps branching the tree now, before the final 1.2 release is out, and
letting interested parties go on with 2.0 and other changes would have
several effects:

	- stop 1.2 from holding everything up
	- stop people from becoming as bored with endless silly
	  discussions about things like snprintf
	- result in _less_ pressure to add things to 1.2 because it would
	  be "old stuff" and people could get interested in the 2.0 tree.

OTOH, it could also fragment development, result in no attention to
critical things that have to be finished up before 1.2 comes out and
destroy everything on top of making some people really confused about cvs. 

There are a good number of patches coming up that would fit quite nicely
into the 1.x series, but are too late for 1.2.  Perhaps a better list of
official and semi-official patches could be a substitute for adding some
of it (the things that add good features which are somewhat specific in
use) to 1.2 or 1.3?