Re: [PATCH] doc: add note about ignoring --no-create-reflog

From: Junio C Hamano <gitster@pobox.com>
Date: Feb 1, 2017, 23:27 UTC
Message-ID: <xmqqefzhmr02.fsf@gitster.mtv.corp.google.com>
In-Reply-To: <20170201231939.hxhhujpzyb2cqq7a@sigill.intra.peff.net>

Jeff King <peff@peff.net> writes:

> Should this perhaps say "currently" or "this may change in the future",
> so that people (including those who might want to fix it later) know
> that it's a limitation and not intentional?

Good point.

Show 9 quoted lines

> I'd also probably say it a little shorter, like:
>
>   The negated form `--no-create-reflog` only overrides an earlier
>   `--create-reflog`, but currently does not negate the setting of
>   `core.logallrefupdates`.
>
> I guess that really isn't much shorter (I wondered if you could cut out
> the "overrides --create-reflog" part, since that is the normal and
> expected behavior, but I had trouble wording it to do so).

I had the same trouble wording. Another thing I noticed was that I deliberately left it vague what "default" this does not override, because it appears to me that those who do not set logallrefupdates will get the compiled-in default and that is also not overriden.

IOW, "does not negate the setting of core.logallrefupdates" will open us to reports "I do not have the configuration set, but I still get reflog even when --no-create-reflog is given".

   The negated form `--no-create-reflog` currently does not negate
   the default; it overrides an earlier `--create-reflog`, though.

perhaps?