Subject: Re: how to go forward from 19981020-SNAP?
To: None <email@example.com>
From: John Darrow <John.P.Darrow@wheaton.edu>
Date: 04/22/1999 18:06:37
In article <9904221755230D.firstname.lastname@example.org>, Dave McGuire wrote:
>On Thu, 22 Apr 1999, Ross Harvey wrote:
>>Having said that, arguments have been advanced periodically that we should
>>put up the exact source the snapshot was made from; if people really care
>>we can certainly do that.
> Speaking only for myself and a few coworkers, we definitely care. We tend to
>stick with a given snapshot for quite some time, and having those sources
>around tends to make life much, much easier.
> -Dave McGuire
This was one of the deficiencies I've run into numerous times, that of no
easy way to produce the source tarballs that a given snapshot built from.
It was one of the things I addressed in pr misc/7089. I basically extended
the distrib/sets/lists stuff to include the sources, and added a make
variable to cause the build process to generate the source tarballs after
it generated the binaries (along with a separate makefile target to just
build the source tarballs). I haven't had a chance to update the
modifications recently, as I was down with the flu, and since then I've
been fighting with cvs (which doesn't seem to have any good way to deal
with things when the vendor you've been tracking via cvs import (i.e. TNF)
decides to branch their sources, and you need to follow (including local
patches) _both_ branches...)
Computing Services, Wheaton College, Wheaton, IL