tag points / branch points

Jeremy Utley jeremy at jutley.org
Mon May 24 18:55:52 PDT 2004


Archaic wrote:

>On Mon, May 24, 2004 at 05:45:11PM -0700, Jeremy Utley wrote:
>  
>
>>Archaic wrote:
>>
>>    
>>
>>>So, what is b5_1?
>>>
>>>      
>>>
>>That's the branch containing all the stuff for the 5.x series, from when 
>>it was branched off the mainline.  The top of that branch will be v5_1_1 
>>when we release (maybe tomorrow).
>>    
>>
>
>Excuse my stupidity, but isn't branch_point_5_1 the actual branch? If
>not, what is it? I thought perhaps branch_point_5_1 was the branch, and
>b5_1 was a tag, but then I wondered why someone would tag a newly
>created branch since it would be identical.
>  
>
Nope, you branch from a tagged point in time, so branch_point_5_1 is the 
specific point in time where the branch was made.

>  
>
>>Not that I know of - you can kind of get an idea of the time frame by 
>>looking at the revision dates before and after, but not really.  Maybe 
>>someone who knows CVS better will have some additional input.
>>    
>>
>
>Which is a PITA. :(
>
>Suggestion to Matt:
>
>In the process of creating a release (making tarballs, tagging CVS,
>making news announcements, etc), could you include the new tagpoint in
>the announcement (preferably on list and on the webpage news) so that
>one can quickly get the identical xml from either the web tarball or
>CVS?  Which reminds me, previous releases had the xml tarred up and
>placed alongside the PDF, nochunks, etc tarballs. I don't see it for
>5.1. Don't know if you forgot, didn't know, or don't want to mess with
>it...
>
>  
>
Probably just forgot, honestly.  A lot of things went wrong with this 
release, we know that.  But, in our defense, it was the first time 
someone besides Gerard was responsible for managing a release.  I know 
Matt's making a set of docs for releases in the future, so I'm sure 
he'll incorporate your suggestions into it.  However, I think we'll 
pretty much follow our previous format, where the tag for a release is v#_#.

-J-



More information about the lfs-dev mailing list