httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From j.@pa.dec.com (Jim Gettys)
Subject Re: New to this list
Date Wed, 22 Jul 1998 14:37:30 GMT

> Sender: new-httpd-owner@apache.org
> From: Rodent of Unusual Size <Ken.Coar@golux.com>
> Date: Wed, 22 Jul 1998 10:19:01 -0400
> To: new-httpd@apache.org
> Subject: Re: New to this list
> -----
> David Southwell wrote:
> >
> > I wasnt suggesting you areeed to support em--
> > In fact you could put those archives under the general heading:
> >
> > Older releases which are no longer supported-
> >
> > This would then make very clear what was being supported and what
> > was not!!
> 
> No, it doesn't - as experience has shown us repeatedly.  If people can
> find old crufty software, they'll install it - and complain when it
> doesn't work as they think it should - regardless of where they
> found it.
> 
> > > > Can I  advocate keeping a good public archive  with adequate
> > > > pointers to it from the main and the dev sites?
> > >
> > > Archive of what?  Mail from this list?  It's there.  Old releases?
> > > Not interested.  Something else?  Maybe - but what? :-)
> >
> > Well its your shout in the end anyway and if you are not interested I
> > suppose that is the end of the discussion.
> 
> So you're asking about an archive of the software kits.  That's what
> *I* was asking - what you were asking to be archived.
> 
> > If I can not bring to empathise with an alternative view point - well
> > that`s life *GRINZ*
> 
> I'll take this out of order.
> 
> > Archives of old releases can be extremely useful and they are generally
> > self maintaining - the only disadvantage is that they take a bit of drive
> > space -
> 
> You don't seem to be empathising with *us* and our experience with
> this.  Archives of old betas have been shown to be far, far more
> trouble than they're worth.  Drive space is the least of the issues,
> believe me.
> 
> So convince us: How can keeping obsolete beta kits be 'extremely
> useful'?  Make a valid case and it will definitely be considered;
> but if it's unfounded opinion it's going to lose against our hard
> experience every time..
> 
> #ken	P-)}
> 
> Ken Coar                    <http://Web.Golux.Com/coar/>
> Apache Group member         <http://www.apache.org/>
> "Apache Server for Dummies" <http://WWW.Dummies.Com/

Certainly Ken's experience jibes with mine when I dealt with the X Window 
System distribution. With the exception of keeping the previous stable 
version during a beta test, experience has shown keeping old ones around 
(particularly in an archive with stable names) to be a pretty bad idea 
(for free software, anyway).  

"Support" resources are very scarce, and handling bug reports on old software 
is something for larger commercial companies to do (and something some 
such companies are good at and can make money at; it is why you pay for 
software "support" if you are a commercial company). It is hard enough 
to get bugs out of the next version, and spending time to make sure that 
the N'th report of the same, fixed, bug subtracts from forward progress.

And the Web has made the problem worse: Today, it is all to easy for there 
to be a hyperlink to an out of date version on a page in a random web 
site, and rather than going back to the authoritative site (which they 
may not know in the first place, and be feeling too lazy to figure out), 
people often follow it, and end up with the old version.

If there is an archive at all, one should be careful to break the
links periodically by renaming it, to avoid this problem of links to
antique software.

-- 
Jim Gettys
Digital Industry Standards and Consortia
Compaq Computer Corporation
Visting Scientist, World Wide Web Consortium, M.I.T.
http://www.w3.org/People/Gettys/
jg@w3.org, jg@pa.dec.com
 

Mime
View raw message