Re: name for A..B ranges?

From: Junio C Hamano <gitster@pobox.com>
Date: Jun 25, 2016, 19:49 UTC
Message-ID: <xmqqwpldcamb.fsf@gitster.mtv.corp.google.com>
In-Reply-To: <E61B46FFA8874DD3973AA96BE5B36790@PhilipOakley>

"Philip Oakley" <philipoakley@iee.org> writes:

Show 9 quoted lines

>> Yup, I think "range" is the commonly used word in discussions here.
>> When inventing A...B as a new thing in addition to A..B, we called
>> the former "symmetric difference", and what is implied by that is
>> the latter is "asymmetric difference"; we do not say that unless we
>> are contrasting between the two, though.
>>
> I asked because the man page does indicae that it (A..B) is a special
> sort of revison range and "there is a shorthand for it", but then
> didn't have a way of naming it.

I do not see "is a special sort of revision range" improved in your two patches, though.

Knowing that A..B is merely a short-hand for ^A B is important to understand how revision ranges work (e.g. "A..B C" is not "union of A..B and C"), so I think it is worth addressing if the existing description appeared to you that it may confuse readers.

Show 7 quoted lines

> The symmetric difference is then brought in as a further similar
> notation. There are a number of Stackoverflow questions about the
> differences betwee 'two dots' and 'three dots' as well, so having a
> word/phrase for it could help.
>
> I was thinking that maybe "single-sided difference (two dots)" maybe
> one choice that is relatively neutral (or even a "two-dot range"...).

When contrasting .. and ..., we have always used "asymmetric" vs "symmetric". I'd prefer to see usnot invent new phrase nobody has used, which leads to unnecessary confusion and learning burden.