GNU bug report logs - #40764
New package: r-restrserve

Previous Next

Full log

Message #31 received at 40764-done <at> debbugs.gnu.org (full text, mbox):

From: zimoun <zimon.toutoune <at> gmail.com>
To: Oleg Pykhalov <go.wigust <at> gmail.com>
Cc: Ricardo Wurmus <rekado <at> elephly.net>, Eric Brown <ecbrown <at> ericcbrown.com>,
 40764-done <at> debbugs.gnu.org
Subject: Re: [bug#40764] New package: r-restrserve
Date: Tue, 15 Dec 2020 15:35:45 +0100

Hi,

On Mon, 14 Dec 2020 at 12:54, Oleg Pykhalov <go.wigust <at> gmail.com> wrote:

> It adds ‘base-commit’ and optionally ‘prerequisite-patch-id’ (see [1]), which will
> help you to checkout base-commit and apply a patch, but still you will
> need to rebase on origin/master with solving conflicts AFAIK.

“with **hypothetically** solving conflicts”.  Well, it is by-design
because of Git.

It is still better than the current situation IMHO where it is boring to
apply old patches.


> Also, on git version 2.29.2 --base requires an argument.  If we apply
> the patch on 43946 issue, we should mention this in example.

Agree.

> And, IMHO, the --base flag should be mention inside the first paragraph
> of [2], where format-patch is recommended for patch submission.
> Otherwise people will send with and without --base, which probably
> will not hurt, but still.

Yes, the first paragraph seems better.

More people will send with base-commit and easier automation will be.
For example, we could imagine that the testing build-farm builds the
patch applied at the base-commit, so the reviewer can use the substitute
instead of burning CPU (sometimes the test suite is really resource
hungry).  The reviewer checks the compliance and rebase if needed.


All the best,
simon

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.