From: Dragan Simic <dsimic@manjaro.org>
Date: Mon, 06 May 2024 18:45:00 GMT
Subject: Re: [PATCH v4 07/14] builtin/config: introduce "list" subcommand
Message-ID: <3d84f2eac9b03eda814332d73b5d5ba6@manjaro.org>
In-Reply-To: <04d001da9fe3$e9c2eef0$bd48ccd0$@nexbridge.com>

On 2024-05-06 20:33, rsbecker@nexbridge.com wrote:
> On Monday, May 6, 2024 1:13 PM, Junio C Hamano wrote:
>> - Declaring that "git restore" and "git switch" were failed
>>   experiments and deprecating them.
> 
> This restore and switch have broad use in my community. I do not
> consider them failed experiments at all, but essential in scripts and
> usage. Removing these would block migration to git 3, in my view. I
> have actually been thinking about doing a YouTube video on these
> commands.

Huh, I was also really surprised to see "git switch" and "git restore"
on the planned deprecated feature list.  For example, I use "git switch"
literally all the time, and "git restore" also sees a fair amount of 
use.
Thus, I think that deprecating these two features is a really bad idea.

>> - Declaring that "git submodule" was a failed experiment and
>>   deprecating it.
> 
> This would be a very high-impact decision. Aside from my community's
> use (and my own company's dependence on submodules), this will break
> large numbers of GNU-based projects that use submodules for
> distribution, including build-aux. I would suggest staying away from
> this decision. Submodules have definite value. Please keep them.

I agree that deprecating "git submodule" isn't a good idea.