From owner-freebsd-hackers Sat Sep 27 19:33:22 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id TAA29692 for hackers-outgoing; Sat, 27 Sep 1997 19:33:22 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id TAA29687 for ; Sat, 27 Sep 1997 19:33:19 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.7/8.6.9) with ESMTP id TAA27234; Sat, 27 Sep 1997 19:33:00 -0700 (PDT) To: Eivind Eklund cc: hackers@FreeBSD.ORG Subject: Re: How do I check out a snapshot? In-reply-to: Your message of "Sat, 27 Sep 1997 17:29:23 +0200." <199709271529.RAA11811@bitbox.follo.net> Date: Sat, 27 Sep 1997 19:33:00 -0700 Message-ID: <27230.875413980@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-hackers@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk An individual tag operation isn't expensive so much as is tagging *every single file* in the src/ hierarchy, essentially what we have to do at each major release. Adding "merge me" tags on a case by case basis would probably be a little different since you'd expect that to happen over a long period of time (the life of the branch you're merging into, basically) and to small collections of files in each instance. That'd make it much less of an impact on all those users out there who are cvsup'ing daily, and not quite what I was talking about below. I believe someone was asking why each SNAP couldn't be reproduced exactly by checking out a tree with the "snap tag", and to do that I'd essentially need to tag all the sources (and, if one wanted to be _really_ anal, the ports) once a day. That would be far too expensive and the lockdown time on the repository to create the tags would not be appreciated either. Jordan > > Yes, I tag and then I build from that tag after resyncronizing my > > local repository. > > > > Unfortunately, tags are expensive and you don't just lay them down > > for fun. > > In what way are release tags expensive? I've been thinking of using > tags to make a commit a single operation instead of a bunch of changes > just connected by the commit-log, and for storing meta-information for > merges. With this and some merge-related tags (for storing > meta-information) syncing RELENG_* and -current the way we want should > become quite a bit easier. Are the tags so expensive this would be > non-feasible? I thought they would be a comparatively cheap way of > storing minor amounts of meta-information. > > Eivind.