[patch] fix bug 48136 bzr status, diff, etc do not work properly after remote push

[John Arbash Meinel]

Aaron Bentley wrote:
> John Arbash Meinel wrote:
>>> Aaron Bentley wrote:
>>>
>>>> John Arbash Meinel wrote:
>>>>
>>>>>> Aaron Bentley wrote:
>>>>>>
>>>>>>
>>>>>>> Using heavyweight checkouts instead of lightweight ones with remote
>>>>>>> branches is just a convention.  So doing this does require a connection
>>>>>>> to the branch, which is a new requirement.
>>>>>> Well, at present because of the way 'BranchReference' is defined, we
>>>>>> always connect to the remote branch, for any lightweight checkout.
>>>> Right.  That's why I phrased it as I did.
>>>>
>>>
>>> ...
>>>
>>>
>>>> out of date (use "bzr update"):
>>>>    branch is at: "Fixed a really nasty bug"
>>>>
>>>> But I'm fine to put it in as the warning, and then worry about tweakage.
>>>>
>>>> Aaron
>>>
>>> Ooooh a summary line of the last commit. Fancy. :) (I really do like it).
>>>
>>> So, would you be +1 if we got rid of the warning in 'diff', and then
>>> encourage Cheuksan to extend his patch to change 'bzr status' to conform
>>> to this?
> 
> Yes, that would suit me well.
> 
> Aaron

Alright. Just to keep things rolling, that is what I did. I merged
Cheuksan's patch, reverted the diff changes, and submitted it to the pqm.

Just to mention, though. That 'bzr status' *also* doesn't need to
connect to the remote branch. It does pretty much the same thing as bzr
diff, only in summary mode. So it actually has even less need to connect
to the remote branch, since it never needs to extract the file texts.

I suppose if you have pending merges, it needs to connect to get the log
messages, etc.

Though dirstate will actually be caching all of the parent trees, and
could conceivably cache the commit messages as well. So neither diff nor
status would have any need to connect to the remote branch.

This isn't truly valid, since you really need to cache the log messages
for all of the ancestry of revisions that you might be merging, etc.

And of course, both of them need to connect to the branch if you supply
any sort of revision specifier. Which is also interesting, in that if
you have an out-of-date working tree, you could do 'bzr status -r -1'
and it should report changes.

Anyway, I'd like to request that Cheuksan update the 'bzr status' code,
so that it also checks the master if it exists. And changes the output
to be like Aaron outlined.

But for now, 'bzr status' is a lot more helpful when the tree is out of
date.

John
=:->

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 254 bytes
Desc: OpenPGP digital signature
Url : https://lists.ubuntu.com/archives/bazaar/attachments/20061027/2874212f/attachment.pgp