Changesets feature complete

Aaron Bentley aaron.bentley at utoronto.ca
Sat May 20 23:24:06 BST 2006


Aaron Bentley wrote:
> Michael Ellerman wrote:

>> # bzr cset -r 1..-1 > patch
>>
>> I was expecting this to be equivalent to the concatenation of all the
>> individual changesets between 1 and -1. But it's not quite. The last
>> changeset seems to contain stuff from way back in the past, which is
>> confusing because it means the message doesn't accord with the diff.
>> I'm not sure if that's the expected behaviour or not.
> 
> 
> This is by design.  'Changeset' is a really bad word to describe what 
> these things have become-- they're more like a collection of revisions 
> in a textual format.

I should clarify that it's by the *current* design.  I'm not stuck on it.

I want mailing changesets to be a viable alternative to publishing 
branches, so I think we need to do this *some of the time*, at least. 
But perhaps that should be reserved for the "bzr cset -r 6 
TARGET_BRANCH" form.

Making it optional would mean there are two kinds of changesets, which 
makes bzr a bit more complicated.

It's a fine balancing act, because changesets need to be both 
human-readable (so they're acceptable on mailing lists) and complete (so 
they aren't inferior to publishing branches).

Suggestions are definitely welcome.  Unless they involve -p1 patches :-P

Aaron




More information about the bazaar mailing list