[PATCH] backout: make help more explicit about what backout does

timeless timeless at gmail.com
Wed Feb 2 05:12:06 CST 2011


On Wed, Feb 2, 2011 at 10:23 AM, Gilles Moris <gilles.moris at free.fr> wrote:
> On Wednesday 02 February 2011 09:10:27 am Jonathan Nieder wrote:
> Why are you switching to the past tense just here? May be:
> +    Otherwise, a merge IS necessary, so you should inspect ITS result
> +    and then commit it.

I think probably 'needed' instead of 'necessary'.

Probably: 'Otherwise hg needs to merge the changes and you need to
inspect them before committing.

>> +    If the --merge option is supplied, the next commit will have two
>> +    parents: the old parent of the working directory and a child
>> +    of REV that simply undoes REV.  Otherwise, the next commit will
>> +    have one parent, maintaining a linear history.
> Tentatively:
> + have JUST one parent, ...

Here, "Otherwise" is not a good word, since we're now talking about a
default case (iiuc).

Use <By default> and list that branch first:

<By default> without specifying --merge, the pending changeset will
have one parent.

With --merge, the ...

>> +    Before version 1.7, the default behavior (without --merge) was

Before version 1.7, the default behavior was equivalent to specifying --merge.

>> +    different. To restore the previous default behavior, use
>>
>>      :hg:`backout --merge` and then :hg:`update --clean .` to get rid of
>>
>>      the ongoing merge.

'ongoing' is not a word you want to use. And I don't understand what
the 'and then ...' stuff is doing, please explain it to the list and
someone can try to suggest something.

> Please feel free to discard those non English speaker comments if you think
> they don't make sense.

Gilles: nope, in this case I was going to comment on those things but
you saved me that effort. :)


More information about the Mercurial-devel mailing list