Help text improvements: mq

[timeless]

On Tue, Nov 3, 2009 at 11:49 AM, Fabian <bona+google at fabian-kreutz.de> wrote:
> #: hgext/mq.py:738
> msgid "local changes found, refresh first"
> -> local changes found, use \"qrefresh\" first

i presume we want 'hg qrefresh'

> #: hgext/mq.py:950
> msgid "cannot push to a previous patch: %s"
> -> "patch %s is already applied, use \"qpull\" to bring it to top"

i don't have a qpull, only qpush and qpop

> The subcommand qguard and the whole guard concept is not mentioned in
> the main help text.

i don't use guards at all


$ hg help qselect
hg qselect [OPTION]... [GUARD]...

set or print guarded patches to push

    Use the qguard command to set or print guards on patch, then use
qselect to tell mq which guards to use. A patch
    will be pushed if it has no guards or any positive guards match
the currently selected guard, but will not be
    pushed if any negative guards match the current guard. For example:

        qguard foo.patch -stable    (negative guard) qguard bar.patch
+stable    (positive guard) qselect stable

the example text is stale, it needs to have --stable
and it needs to be split onto two lines...

i wonder if we could change to foo[+-] instead of [+|--]foo, the
syntax seems to suck a lot.

> Is there any difference between:
> #: hgext/mq.py:965
> msgid "all patches are currently applied\n"

timeless-mbp-2:hgext timeless$ hg qpush --all
all patches are currently applied

[you asked for all but there was nothing to do]

> #: hgext/mq.py:975
> msgid "patch series already fully applied\n"

timeless-mbp-2:hgext timeless$ hg qpush
patch series already fully applied

[no patch to apply]

> #: hgext/mq.py:1733 hgext/mq.py:1895
> msgid "all patches applied\n"

timeless-mbp-2:hgext timeless$ hg qun; hg qnext
all patches applied

[there is no patch after the current patch]

i'd say they're all distinct and relatively meaningful __in context__.
there's even some hinting in the code explaining how to reach them.

> #: hgext/mq.py:1471
> msgid "save entry has children, leaving it alone\n"
> is very unclear
> "entry" is a revision, "has children" == "not a head" and "leaving it
> alone" is not a subordinate clause . I don't know if it would be
> correct to say
> --> "save revision is not a head - skipping\n"
> I have difficulties translating "save revision" or state or entry, as
> I don't understand it's meaning from the bleak "restore the queue
> state saved by a revision" documentation.

sorry, too complicated for me tonight (2:40am)

please change the subject when you change components