On Wed, Mar 18, 2026 at 11:21:28AM +0800, Kevin J. McCarthy wrote:
Okay, you don't have to twist my arm. I resisted formatters myself for a long time already, but started to feel like "everyone else was doing it" and thought I'd give it a look.Before I drop the proposal, though, are there any Mutt devs/contributors who would like to play devil's advocate?
Thanks Robin (yes your voice counts too), Will, Steffen, and of course Alex for continuing the discussion some more. I was getting mostly negative feedback and wanted to know if *anyone* felt there was a benefit to formatters.
Just looking over the initial results of my trial run last weekend again, there are some iffy results, but overall it wasn't that terrible. I suppose how you feel about the result depends on how picky you are about formatting, and how well your format style can be expressed in the .clang-format config.
The mutt code is already on the messy side. It's a mostly-Allman style with a space in front of funcalls. But otherwise is a mish-mash of pointer-operator locations, cast styles, and other miscellanea. I think a little iffy results would be okay if it meant smoothing out the inconsistencies.
My only problem is the gettext calls are spaced too: "_ (" and "N_ (". I was initially thinking about following up with a sed script to fix those. But now, I'm actually leaning towards getting rid of the space before parens completely. It's not a conventional style, and combined with nested gettext calls (or any nesting really), looks terrible.
This weekend, I'm planning on releasing 2.3.1 from stable. After that, I think I'll change the config to remove the space before funcall, run it again, commit to a branch and announce it here. You all can take a look and tell me what you think.
-- Kevin J. McCarthy GPG Fingerprint: 8975 A9B3 3AA3 7910 385C 5308 ADEF 7684 8031 6BDA
signature.asc
Description: PGP signature
