launchpad-dev team mailing list archive

[Barry Warsaw <barry@xxxxxxxxxxxxx>]

On Nov 25, 2009, at 9:28 AM, Francis J. Lacoste wrote:

> The Bugs case is different in that it's a bug number (which happens  
> to reuse
> the database identifier). The bug number is well-exposed in the UI.  
> The reason
> it was used was also because a bug could live in many context and it  
> is
> impractical to give a nickname to all bugs and try to manage a  
> coherent
> namespace on top of that.

From a usability standpoint, I've always (20 years at least :)  
thought that bug numbers are a convenient short hand for  
communication, but really insufficient.  For example, if you ask me  
whether I fixed bug 12345 last month, I say "huh? which one was  
that".  I'll have to look it up to see the summary at which point I'll  
(probably ;) remember the details of the bug.

OTOH, it's great to be able to type "lpbug 12345" in my browser and  
let my shortcut go right to that bug page.  It's great to be able to  
just mention "Bug 12345" in a merge proposal or in an irc channel and  
have the 'bots linkify that.  So the bug number is a great handle in  
some cases but incomplete in others.

I'm not sure whether this helps the discussion, except to add that I  
do generally give informal nicknames to my branches when I'm working  
on a bug.  I used to do something like:

% bzr branch devel bug-12345

but then I realized that if that branch lived on my local filesystem  
for too long, I'd completely forget what it was about, especially if  
there were many other such branches.  Now I've gotten in the habit of  
doing something like:

% bzr branch devel 12345-addmember

to give me more of a clue as to what the branch is about.  I consider  
the 'addmember' bit a bug nickname, though it's really tied more to  
the branch.  (Which can get into trouble if I have more than one  
branch that is associated with a bug, i.e. for looms or pipes.)

-Barry

Attachment: PGP.sig
Description: This is a digitally signed message part