martinvonz added a comment.
In https://phab.mercurial-scm.org/D2013#34320, @spectral wrote:
> In https://phab.mercurial-scm.org/D2013#33867, @martinvonz wrote:
> > I wonder if we should instead have a --draft option for this. Reasons:
> > - If we ever add a fourth phase (like Jun's proposed "archived" phase),
then --no-secret doesn't clearly indicate "draft", it could just as well be
> > - Actually, we of course already do have a third phase. One could imagine
a "hg commit --public", although that's probably not useful enough to warrant
its own option, but it seems to suggest that "--no-secret" doesn't necessarily
> > - I find this tri-state boolean weird. "--secret" kind of defaults to
off, but it can be made "more off" with "--no-secret".
> Yeah, I wasn't sure I liked it when writing it, and I'm fine with changing
it, but do we really want a proliferation of flags? Perhaps we want a generic
'phase' flag, so one can specify -s or --secret (BC and it's the "most common"
case), and --phase <public|draft|secret|archived> for more advanced use-cases?
This runs into a couple small problems, specifically that there's now "more
than one way to do it [specify secret]", it's a lot of typing, and I don't
think we should abbreviate it `-p` (it just doesn't feel like it's going to be
common enough to warrant any abbreviation, let alone 'p', which could stand for
phase, or patch, or path, or a number of other words that we might eventually
add to `commit` and be sad about not being able to abbreviate in the obvious
I think "--phase <phase>" makes sense, but it is a little long as you said.
Since the real goal of this series is to preserve the phase on `hg split`, I'll
see if I can do that by adding general support to scmutil.cleanupnodes()
instead (Kyle and I already talked about this offline).
To: spectral, #hg-reviewers
Cc: martinvonz, mercurial-devel
Mercurial-devel mailing list