My understanding of STD 66 is that commas should bind to path segments, rather than delimit paths entirely, but I'm happy to be corrected if I'm wrong. Also, I don't see any problem with having commas or even slashes in a branch name, that's what percent-encoding is for.
So, if the plan is that ",foo" in a URL indicates a branch name of "foo", then I'd expect URLs to be parsed like:
I wasn't at Strasbourg. Did we decide against a key/value syntax (like ",branch=name"), or is it still intended that we might support e.g. ",revid=AAAA..." at some point?
(What happened to the promised update to doc/developers/colocated-branches.txt?)
My understanding of STD 66 is that commas should bind to path segments, rather than delimit paths entirely, but I'm happy to be corrected if I'm wrong. Also, I don't see any problem with having commas or even slashes in a branch name, that's what percent-encoding is for.
So, if the plan is that ",foo" in a URL indicates a branch name of "foo", then I'd expect URLs to be parsed like:
* http:// host/path, branchname host/path host/path, branch% 2Cname host/path host/path, branchname/ README. txt host/path
* branch location: http://
* branch name: branchname
* http://
* branch location: http://
* branch name: branch,name
* http://
* branch location: http://
* branch name: branchname
* path inside branch: README.txt [e.g. I want to be able to use this with bzr log, bzr cat, and bzr annotate]
I wasn't at Strasbourg. Did we decide against a key/value syntax (like ",branch=name"), or is it still intended that we might support e.g. ",revid=AAAA..." at some point?
(What happened to the promised update to doc/developers/ colocated- branches. txt?)