[PATCH in crew] help: document '@' bookmark in 'help bookmarks' and 'help clone'

Sean Farley sean.michael.farley at gmail.com
Mon Jan 28 12:36:52 CST 2013


On Mon, Jan 28, 2013 at 12:22 PM, Matt Mackall <mpm at selenic.com> wrote:
> On Mon, 2013-01-28 at 11:58 -0600, Sean Farley wrote:
>> On Sun, Jan 27, 2013 at 9:34 PM, Kevin Bullock
>> <kbullock+mercurial at ringworld.org> wrote:
>> > On 27 Jan 2013, at 7:06 PM, Antonio Zanardo wrote:
>> >
>> > Hello,
>> >
>> > That's nice to have this feature documented, but I fail to understand
>> > clearly why this feature is useful, i.e., what are the use cases?
>> >
>> > Does the bookmark "@" exist to serve as a starting point for the
>> > contributors of a project to know which revision to base their patches?
>> >
>> >
>> > Yep, that's about right. It's the revision you want people to get by default
>> > when they clone your project. It could represent your latest integrated,
>> > tested, non-experimental code. This is particularly useful in the presence
>> > of other bookmarks that might mark experimental or unstable work (topic
>> > branches).
>>
>> I am very interested in the topic branch workflow but the presence of
>> '@' still has some quirks:
>>
>> 1) 'hg clone -U <url> && hg up' still updates to 'default' even when
>> '@' is present. I believe this should update to '@' (if it exists),
>> right? The BookmarksPlan mentions this at the bottom: "Bare update
>> with no active bookmark updates to the implicit branch-default
>> bookmark if it exists."
>
> Notabug. With clone -U, you get the null checkout, which is on the
> default branch. Updating takes you to the tip of the default branch.

Well, I'm trying to get a bookmarks based workflow and don't see how I
can "enable it." By that, I mean I'm willing to do something outside
the norm (either enable a setting, or adding the '@' bookmark, etc.).

>> 2) 'hg push -B <bookmark>' pushes everything + bookmark. It'd be nice
>> to just have a way to push the bookmark (+ancestors).
>
> hg push -r <bookmark>?

Dammit. I don't know why I don't test these things out before I send the email.


More information about the Mercurial-devel mailing list