From: Luiz Eduardo Campos Date: Wed, 25 Mar 2026 17:04:03 GMT Subject: Re: [RFC PATCH 1/1] add -p: support discarding hunks with 'x' Message-ID: In-Reply-To: Hi Ben, Thanks for the feedback! (Sending this message again because I had not configured my email client to send it plain text, so I apologize if you received duplicated content) > One feature the Fugitive Git client for Vim supports is to discard > hunks; when it does so, it also prints a message explaining how to > recover the hunk if you need it. That’s a great point. I hadn’t considered recoverability in this initial version. Storing the discarded hunk as a blob (or otherwise making it recoverable) seems like a useful safety measure. > If it's never been in the index, it can be impossible to recover! Agreed — this is probably something that should be addressed before considering this feature complete. > PS How different is this from "git restore -p" ? My understanding is that `git restore -p` already allows discarding changes interactively, but it requires a separate pass. The goal here was to allow discarding during `git add -p`, so users can decide what to do with each hunk in a single pass. That said, I’m not yet sure whether integrating this into `add -p` is the best approach, or if this should be handled differently. Thanks again for the insights! Luiz Em qua., 25 de mar. de 2026 às 12:44, D. Ben Knoble escreveu: > > On Wed, Mar 25, 2026 at 3:53 AM Luiz Campos wrote: > > > > When using `git add -p`, users can stage or skip hunks, > > but cannot discard unwanted changes from the working tree. > > > > Introduce a new 'x' action to discard the current hunk by > > reverse-applying it. > > > > This idea was suggested in a previous mailing list discussion: > > https://lore.kernel.org/git/X%2FiFCo0bXLR%2BLZXs@coredump.intra.peff.net/t/#m0576e6f3c6375e11cc4693b9dca3c1fc57baadd0 > > > > Feedback is very welcome. > > > > Signed-off-by: Luiz Campos > > One feature the Fugitive Git client for Vim supports is to discard > hunks; when it does so, it also prints a message explaining how to > recover the hunk if you need it. > > I think it writes the file as a blob to Git's database before > restoring from the index. > > I'm not suggesting Git copy this necessarily, but we might want to > consider how to help folks when they lose a hunk they didn't mean to. > If it's never been in the index, it can be impossible to recover! > > PS How different is this from "git restore -p" ? > > > -- > D. Ben Knoble