Ambiguous link fields in revlog entries

Matt Mackall mpm at selenic.com
Thu Jul 17 14:09:40 CDT 2008


On Thu, 2008-07-17 at 13:50 -0500, David Baum wrote:
> I was looking at the changelog/manifest/revlog code and believe there
> is some ambiguity regarding the link field in an entry's header...

Yep.

> The link field for that manifest node is ambiguous.  Alice will have
> recorded it as pointing to her changeset, while Bob will have it
> pointing to his changeset.  What are the consequences of this?  Is the
> only requirement that the link field point to some changeset that uses
> the manifest?  When sending a changegroup, does it matter which link
> field is used?  Am I missing something?

Ideally, you point to the first such changeset.

-- 
Mathematics is the supreme nostalgia of our time.



More information about the Mercurial-devel mailing list