Re: [PATCH 1/2] add -p: mark split hunks as undecided

From: Junio C Hamano <gitster@pobox.com>
Date: Feb 21, 2025, 21:31 UTC
Message-ID: <xmqq34g79e8k.fsf@gitster.g>
In-Reply-To: <43a0592a462cf68bcfdc54373da2319431c3c1ca.1740149837.git.gitgitgadget@gmail.com>

"Phillip Wood via GitGitGadget" <gitgitgadget@gmail.com> writes:

Show 10 quoted lines

> From: Phillip Wood <phillip.wood@dunelm.org.uk>
>
> When a hunk is split each of the new hunks inherits whether it is
> selected or not from the original hunk. This means that if a selected
> hunk is split all of the new hunks are selected and the user is not asked
> whether or not they want to select the new hunks. This is unfortunate as
> the user is presumably splitting the original hunk because they only
> want to select some sub-set of it. Fix this by marking all the new hunks
> as "undecided" so that we prompt the user to decide whether to select
> them or not.

Good. I am very sure that the design of the current behaviour goes back to the very original version of "add -p" with hunk splitting I invented; I simply never considered a workflow where people may first select and say "oops, let me take it back and redo it". What I am getting at is that I do not think the current behaviour is something I designed it to be with too much thought, and debeting if it makes sense or it would be better to force them to be undecided is probably a good thing to do now.

Having said that, I have one small concern about forcing them to be undecided. This now allows it to

 1. Add the whole hunk 
 2. Go back (with K) to that already chosen hunk
 3. Split

and makes the resulting minihunks more obvious, as you do not have to use the uppercase J/K to visit them.

But if one is very used to do this intentionally (as opposed to "oops, let me take it back"), this would be a usability regression. "Ah, here is a big hunk with 10 changes, most of which I like, but one of the lines I do not want to include" in which case I may do the "Add the hunk to grab 10 changes, visit that decided-to-be-used hunk, split, and then visit the one minihunk that I want to eject and say 'n'". This makes the workflow simpler and more stupid by requiring the 9 minihunks to be chosen individually after splitting.

So, I dunno.