16 messages in com.perforce.perforce-userbranching question| From | Sent On | Attachments |
|---|---|---|
| Kurt...@best.com | 11 May 1998 12:57 | |
| Davi...@home.chat.net | 11 May 1998 14:18 | |
| Kurt...@best.com | 11 May 1998 14:38 | |
| Guid...@python.org | 11 May 1998 14:54 | |
| Ping...@MIT.EDU | 11 May 1998 15:38 | |
| Brad...@email.mot.comBrad_Appleton-GBDA001 | 11 May 1998 15:45 | |
| Step...@vance.com | 11 May 1998 17:14 | |
| WesP...@xmission.com | 12 May 1998 00:48 | |
| Kurt...@best.com | 14 May 1998 10:18 | |
| WesP...@softweyr.com | 14 May 1998 10:53 | |
| Mark...@glyphic.com | 14 May 1998 11:19 | |
| Brad...@email.mot.comBrad_Appleton-GBDA001 | 14 May 1998 11:43 | |
| Davi...@home.chat.net | 14 May 1998 11:56 | |
| Marc...@mpath.com | 14 May 1998 12:18 | |
| Brad...@email.mot.comBrad_Appleton-GBDA001 | 14 May 1998 12:30 | |
| Davi...@home.chat.net | 14 May 1998 14:23 |
| Subject: | branching question![]() |
|---|---|
| From: | Kurt...@best.com (Kurt...@best.com) |
| Date: | 05/14/1998 10:18:22 AM |
| List: | com.perforce.perforce-user |
Thanks to all who responded; I got a lot of very thoughtful replies.
I read the white paper again and I think I get it now, except for one thing: I'm not convinced that I understand why "90% of SCM process is enforcing codeline promotion to compensate for the lack of a mainline". In fact, I'm not sure I get the purpose of the main- line.
In our environment, trying to keep the most active project in the mainline means that developers are creating new clients every 2-3 months, new codelines are being created more often than that, and much whining is heard.
This is what we're probably going to end up doing (Thanks for the nice ASCII graphic!):
Wes Peters - Softweyr LLC (softweyr at xmission.com) typed this ...
mainline ---+-------------------------------+----------- release_3 \ release_4 \ release_3.0 +------ release_4.0 +--- \ \ release_3.1 +------- release_4.1 +------- \ release_3.2 +---- \ release_3.3 +------------
My question about this model concerns integrations. Should integ- rations from the release_3.3 branch go to 3.2, then 3.1, touching each branch on the way to the mainline? It seems like the policy has to take into account whether the change is new development or a bug fix, and that means that our automated integration system won't work, creating the opportunity for more whining. "A tool is only as good as you use it", and if developers are spending their time whining, we're not using the tool optimally.
Thanks again!
---Kurt




