Proposed bookmarks plan posted to wiki

Matt Mackall mpm at selenic.com
Mon Jun 11 17:37:55 CDT 2012


On Fri, 2012-06-08 at 14:42 -0500, Kevin Bullock wrote:
> === push ===
> 
>  * `hg push -r BOOKMARK` should automatically sync the bookmark too, without having to specify `-B`.

Not clear. Pierre-Yves seems to have a use case where maybe he wouldn't
desire that.

> Discussion: It should be possible to push a bookmark and all the
> changesets leading up to it in one command. Right now, `hg push -B
> BOOKMARK` pushes _all changesets_, and then the given bookmark, which
> is too much.

That -B behavior sounds like a bug, and doesn't align well with the docs
at all. Really sounds like it ought to do what you're proposing for -r.

>  * Warn (with hint) when pushing changesets that are bookmarked without pushing the bookmarks (suggested by [[marmoute]]: http://markmail.org/message/pxemyessq6ek7ynw).

I don't think a warning is sufficient. Given our non-existent support
for unpushing, this warning will usually translate as "you idiot, you
just pushed the code you were trying to keep local".

Sometimes our "no new heads" restriction will prevent this class of
accident, but I think we should aim to do it more generally.

> This is a compromise between local-only bookmarks and topic-branch
> bookmarks.

Not sure what the distinction you're drawing here is.

-- 
Mathematics is the supreme nostalgia of our time.




More information about the Mercurial-devel mailing list