[Bioperl-l] move ancient branches to attic

Chris Fields cjfields at illinois.edu
Fri May 14 15:34:41 UTC 2010


On May 14, 2010, at 8:56 AM, Hilmar Lapp wrote:

> 
> On May 14, 2010, at 9:32 AM, Jay Hannah wrote:
> 
>> You don't find large lists of probably dead things annoying?
> 
> 
> Not if they're not in the way of my executing my workflow effectively. Keeping a room super-tidy that you don't ever live in is a waste of energy.
> 
> As an analogy, Google Mail keeps all your dead email (email you delete). Forever. Not because they think most of what you delete you shouldn't have deleted, but because it costs so little, and can be so efficiently managed for the few things that you do decide to recover a year later that it's not worth for you as a user to spend any brain cycles on which emails you should physically delete and which you should only "archive".
> 
> Likewise, I don't see the gain that outweighs the brain cycles and careful consideration that would have to go into deciding which branches to delete, which ones to move into an "attic", and which ones to keep around. If you don't want to see them, simply clone and wipe them away. Life can be so easy :-)
> 
> 	-hilmar

I tend to fall in the middle here, in that it would be nice to clean out feature branches that have been merged back in and relegate all older branches to an attic.  Moving branches is as easy as 'git branch -m foo attic/foo'.  I'm not in favor of removing branches that haven't been merged back, unless they're deemed unnecessary by the core devs.  

re: removing feature branches, this is something we have talked about doing in the past on svn, but is a bit trickier at the moment as the git repo doesn't currently indicate if/when specific svn branches were merged to HEAD.  We still have read-only access to our svn repo to determine that if needed. 

So far, though, I haven't seen much in the way of indicating what some regard as 'feature' (removable) vs 'attic' (old but retained).  That discussion needs to happen on list.

chris






More information about the Bioperl-l mailing list