Bazaar User Guide: Workflows

Matthew D. Fuller fullermd at over-yonder.net
Thu Mar 27 23:49:39 GMT 2008


On Fri, Mar 28, 2008 at 09:39:27AM +1000 I heard the voice of
Ian Clatworthy, and lo! it spake thus:
> 
> I think the underlying reason is one that you guys have been raising
> a lot lately: mental model. For Git and Mercurial, the underlying
> model is a vanilla DAG and commands map to DAG operations. For
> Bazaar, the underlying model is different. [...]

+1.  More to the point, in git/hg, it seems that there is thought of
"_THE_ DAG".  And that leads to the basis of criticisms along the
lines of "the Unnecessary Merge Revs(tm) interfere with keeping _the_
DAG in sync", from not really grokking the reasons we consider the
local views[0] advantageous.

I ponder from time to time how best to explain this.  I think the best
starting point is to emphasize the meaning of local views on "_this_
branch".  And since we're on workflows, hierarchial workflow models
are an excellent place to show this; the 'collapsing' aspect of the
local view in progressively "higher" branches can make the
progressively "higher" views of history more understandable.

Sadly, this is one of the things on my list to braindump somewhere in
my Copious Free Time...




[0] The other problem is when 'local view' is taken too literally, and
    combined with the predisposition toward '_THE_ DAG', and taken to
    mean seeing the same DAG differently.


-- 
Matthew Fuller     (MF4839)   |  fullermd at over-yonder.net
Systems/Network Administrator |  http://www.over-yonder.net/~fullermd/
           On the Internet, nobody can hear you scream.



More information about the bazaar mailing list