RE: Versioning between checkout|update, commit

[Donald Sharp \(sharpd\)]

The problem with this methodology is that it's very easy for stuff to
get checked in that will break the build for other people.  Isolation is
*good*.  Allowing people to save work that may not be finished is also
good( although maybe not as good ). 

donald

-----Original Message-----
From: address@hidden
[mailto:address@hidden On Behalf Of Andy
Jones
Sent: Thursday, April 01, 2004 10:05 AM
To: Hugh Sasse Staff Elec Eng; address@hidden
Subject: RE: Versioning between checkout|update, commit



>> Some shops also implement a handoff mechanism that divorces the 
>> notion of "latest committed" from "candidate for integration".  That 
>> allows the developers to commit with impunity without fear that the 
>> world would see something inappropriately.
>
>Yes, some places seem to do this with different branches...

You don't *have* to use branches.

Here we tag each release.  We also tag the *next* release, and this tag
gets moved and updated as part of the testing process.

The developers are free to commit code at any time they like, knowing
that it will only appear in the test environment once the release tag
has been moved to that version.

The only time they would need to make a branch would be if they were
developing something that clashed with other development work.  However
this is handled, it will still need a merge at some point - the problem
is not CVS, but the work itself.

Andy Jones
Tapestry Software. 



_______________________________________________
Info-cvs mailing list
address@hidden
http://mail.gnu.org/mailman/listinfo/info-cvs