Re: "Write a new package" culture instead of patches?

[Clément Pit-Claudel]

On 18/05/2020 10.49, Eli Zaretskii wrote:
>> From: Philippe Vaucher <address@hidden>
>> Date: Mon, 18 May 2020 07:41:42 +0200
>> Cc: Eli Zaretskii <address@hidden>, "address@hidden" <address@hidden>
>>
>> That said, for those living on github/gitlab/etc compared to ELPA you
>> feel at home... you just open issues, make pull requests & get
>> answered there, you feel "welcome". On ELPA/emacs-devel you don't feel
>> as welcome because of copyright assignments / subscribing to a mailing
>> list / having to create patches and send an email, that plus usually
>> the first answer you receive is that you did your commit message all
>> wrong and that it follows complex rules in a tone that is more serious
>> and "hard work" than what you get on MELPA.
> 
> I think you make the MELPA rules sound easier, and our rules sound
> harder, than they actually are.  I suggest to scan the archives for
> proposals to add new packages to ELPA, where you will see neither the
> need to subscribe to this list, nor the need to create patches and
> email them, nor "all wrong" responses with a certain "tone".  At least
> not in general.

Part of the problem could be perception?  MELPA does an incredible job at 
explaining the process to get accepted (and indeed they get slightly more than 
one new package per day). 

To address this, what about (1) adding a prominent link to elpa.gnu.org on 
www.gnu.org/software/emacs, and (2) on elpa.gnu.org, removing "To contribute a 
new package refer to the README" and replacing it with a new section titled 
"Contributing packages"? 

That section would include the part of the README that deals with new packages, 
and would highlight external branches more prominently, since those are closer 
to MELPA's way of doing things.

How does pushing to these branches work, btw? Do we have a way on savannah to 
give a user push access to a single branch? Or do we give savannah access to 
everyone who submits a package?

Clément.