Date: Fri, 21 Nov 1997 10:39:11 -0600 From: Richard Wackerbarth <rkw@dataplex.net> To: chad@dcfinc.com Cc: nate@mt.sri.com (Nate Williams), mike@smith.net.au, freebsd-stable@freebsd.org Subject: Re: Version Resolution? Message-ID: <l03110707b09b66ad2958@[208.2.87.4]> In-Reply-To: <199711211532.IAA05903@freebie.dcfinc.com> References: <199711210529.WAA13140@mt.sri.com> from Nate Williams at "Nov 20, 97 10:29:17 pm"
next in thread | previous in thread | raw e-mail | index | archive | help
At 9:32 AM -0600 11/21/97, Chad R. Larson wrote: >I don't think you're listening. The ID gets changed as part of a *commit*. >Each change state of a given tree is uniquely identified. There are still "race" conditions that we cannot solve by any means that does not serialize the update/snapshot process. Since we cannot get an absolutely accurate stamp, I think that an "approximate" stamp done by a cron job is both adequate and simpler. >The ID is kept in a source module (most seem to be voting for >newvers.sh). CVS changes that module during a commit, but doesn't RCS >the change (to avoid the file growing without bounds--history of all >these timestamps is meaningless). Any change to newvers.sh other than >the timestamp would still be RCSd. Fundamental Rule of Modular Programming: It is not a good idea to have two mechanisms altering the same file. It is much better to have "newvers.sh" (or anyone else who cares) include a file which contains the timestamp information. [Been there, done that...] Richard Wackerbarth
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?l03110707b09b66ad2958>