emacs-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: GNU ELPA package discoverability


From: Tim Cross
Subject: Re: GNU ELPA package discoverability
Date: Tue, 26 May 2020 10:24:01 +1000



On Tue, 26 May 2020 at 01:20, Eli Zaretskii <address@hidden> wrote:
> From: Tim Cross <address@hidden>
> Date: Mon, 25 May 2020 10:21:47 +1000
> Cc: Richard Stallman <address@hidden>, Emacs developers <address@hidden>
>
>  Aren't these questions answered in ELPA's README?
>
> No it doesn't. It does answer some of them, but not others (and is probably not
> the right place to answer some of them).
>
> The ELPA README is certainly a start with respect to basic workflow. Information
> which it lacks that a new developer may want to know or which needs
> clarification includes
>
> - Who can get push permission to the ELPA git repository?
> - How do you request push permission?
> - The README is a little weak on process when you don't have push permission
> - The instructions re: email to emacs-devel for package developers who do not
>   have push permission does not indicate how to provide the package
>   sources/updates. It says that someone will push it for you, but if you
>   don't have push permission to the git repo, how do you get your code in there
>   to begin with?

So basically what's missing is the write access issue.  That should be
a single sentence: we don't have ELPA write access, only access to the
entire Emacs repository, so they need to request membership in the
Emacs project.

That would certainly be a good start. However, is that a maintainable approach. 

Assume we are successful in getting more packages into ELPA, increasing the discoverability of appropriate packages without the need to add repositories like MELPA. Will all of those developers be entitled to write access to the GNU git repository? What about Richard's proposed non-copyrighted repository? Perhaps now is the right time to look at the architecture and consider breaking off ELPA into a separate authentication realm? 

> Questions about what can/should go into ELPA, what should be included in Emacs
> core and what cannot go into ELPA are not addressed at all (the README is
> probably not the right place for this information)

It's quite expected that this is not described, because we are still
arguing about that.

I think this is an important argument to resolve in order to address the other issues that have been raised, such as improving package discoverability or implementation of a non-copyrights assigned repository. ELPA has existed for quite some time now and we still don't have a clear definition of what should go into it. How long do we need to argue about it before making a decision? Is anything being recorded regarding the various arguments or is it just endless unconnected threads in the mail list? In other words, how far have we moved towards a consensus?

--
regards,

Tim

--
Tim Cross


reply via email to

[Prev in Thread] Current Thread [Next in Thread]