Including ESS with Emacs

Including ESS with Emacs

[Alex Branham]

[-- Attachment #1: Type: text/plain, Size: 938 bytes --]

Hello emacs-devel:

I'm Alex Branham, one of the core developers of Emacs Speaks Statistics
(ESS), an elisp package that provides support for a few statistical
languages (R, julia, SAS, and BUGS/JAGS). You can find more info on our
website[1] or clone the code from github.[2]

We'd like to contribute ESS to the FSF so that it can become part of
Emacs. My understanding is that the copyright assignment process might
be the most difficult part. ESS is a very (30ish years) old project, and
so has many contributors, some of whom may not be able to sign the
paperwork.

Is there a process for figuring out who exactly needs to sign the FSF
copyright paperwork? Is it just authors of the current code according to
git-blame, or everyone who's contributed 15 lines of code, or something
else?

Thanks,
Alex

Footnotes:
[1]  https://ess.r-project.org/

[2]  https://github.com/emacs-ess/ESS or git clone https://github.com/emacs-ess/ESS.git

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 487 bytes --]

Re: Including ESS with Emacs

[Stefan Monnier]

> We'd like to contribute ESS to the FSF so that it can become part of
> Emacs.

Yay!
[ I think it'd be best to include it into GNU ELPA rather than into
  Emacs, tho.  ]

> My understanding is that the copyright assignment process might
> be the most difficult part.  ESS is a very (30ish years) old project, and
> so has many contributors, some of whom may not be able to sign the
> paperwork.
>
> Is there a process for figuring out who exactly needs to sign the FSF
> copyright paperwork? Is it just authors of the current code according to
> git-blame, or everyone who's contributed 15 lines of code, or something
> else?

It should be all authors who contributed more than 15 lines-or-so of the
current code.  `git blame` can be a good starting point, but note that it
can be misleading: someone who just adds a `save-excursion` and reindents
all the code will cause git-blame to hide the contributors of all the
reindented code.

The way I went about it in the past is to collect all the known authors.
Then I remove those who already have signed paperwork, and those who
contributed less than 15 lines of code (by looking at the
corresponding commits).
Then I try to contact all those that remain.

This often involves a fair bit of work and time, to try and find current email
addresses, to get an answer, to sign the paperwork, to get the FSF to
process the documents, ...

After that, there are typically some authors still pending, either
because they refuse to sign, or because you can't find a valid email, or
because they don't reply.

At that point, you dig into the code in more detail, to try and see
which part of their contribution remains in the current code and whether
that can fall under the "15-lines worth of code" (where a renaming of
a variable counts as 1 line even if it touches 100 lines).  That often
lets you remove a few more from the pending list.

Finally, there will probably be a few hold outs who contributed
a non-trivial amount of code which is still present but whose copyright
paperwork we can't get.  At that point, we have to take a yet deeper
look at what to do with it: rewrite the code, drop it altogether, or
move it to a separate package.

Note that if at the last stage there is still a major contributor, it
might render the whole effort pointless.  So it's best to start with the
most prolific contributors.


        Stefan

Re: Including ESS with Emacs

[Alex Branham]

[-- Attachment #1: Type: text/plain, Size: 2658 bytes --]


On Sun 25 Nov 2018 at 10:45, Stefan Monnier <monnier@iro.umontreal.ca> wrote:

>> We'd like to contribute ESS to the FSF so that it can become part of
>> Emacs.
>
> Yay!
> [ I think it'd be best to include it into GNU ELPA rather than into
>   Emacs, tho.  ]

Right, I was including ELPA as "part of Emacs". We'd be completely happy
being in "core" Emacs or on ELPA, whatever the maintainers think is
best.

> It should be all authors who contributed more than 15 lines-or-so of the
> current code.  `git blame` can be a good starting point, but note that it
> can be misleading: someone who just adds a `save-excursion` and reindents
> all the code will cause git-blame to hide the contributors of all the
> reindented code.
>
> The way I went about it in the past is to collect all the known authors.
> Then I remove those who already have signed paperwork, and those who
> contributed less than 15 lines of code (by looking at the
> corresponding commits).
> Then I try to contact all those that remain.
>
> This often involves a fair bit of work and time, to try and find current email
> addresses, to get an answer, to sign the paperwork, to get the FSF to
> process the documents, ...
>
> After that, there are typically some authors still pending, either
> because they refuse to sign, or because you can't find a valid email, or
> because they don't reply.
>
> At that point, you dig into the code in more detail, to try and see
> which part of their contribution remains in the current code and whether
> that can fall under the "15-lines worth of code" (where a renaming of
> a variable counts as 1 line even if it touches 100 lines).  That often
> lets you remove a few more from the pending list.
>
> Finally, there will probably be a few hold outs who contributed
> a non-trivial amount of code which is still present but whose copyright
> paperwork we can't get.  At that point, we have to take a yet deeper
> look at what to do with it: rewrite the code, drop it altogether, or
> move it to a separate package.
>
> Note that if at the last stage there is still a major contributor, it
> might render the whole effort pointless.  So it's best to start with the
> most prolific contributors.

Thanks, this is very helpful! I'll work on generating such a list of
contributors. What is the best way to go about telling people how to do
the copyright process? I know for a contribution to Emacs they email
yall requesting the copyright form but it might be annoying for yall to
deal with all of ESS's contributors. Does it make sense for you to email
me the form and then when I reach out I can include the form with my
email?

Thanks again,
Alex

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 487 bytes --]

Re: Including ESS with Emacs

[Paul Eggert]

Alex Branham wrote:
> Is there a process for figuring out who exactly needs to sign the FSF
> copyright paperwork? Is it just authors of the current code according to
> git-blame, or everyone who's contributed 15 lines of code, or something
> else?

If we accept just the latest version of ESS, we'd need permission only from 
people who have made significant contributions to that version. We wouldn't need 
permission from people who made past contributions that haven't survived into 
the current version.

Unfortunately, though, the question of "who contributed to this line of code?" 
cannot in general be answered by "look at the git blame output for this line", 
as the line may have most recently been touched by person X even though its main 
author was person Y earlier.

I don't know how many contributors we're talking about here, but one possible 
option is to ask for permissions from everybody, and then if you can't get 
permission from persons P, Q and R then you look at every commit that they made 
and see whether any significant part or those commits have survived into the 
current version. If some bits of code have survived, take them out or rewrite 
them from scratch (ideally via a cleanroom approach).

> Does it make sense for you to email
> me the form and then when I reach out I can include the form with my
> email?

Sure, here's the form I usually use. Other forms might apply depending on your 
contributors' concerns.

https://git.savannah.gnu.org/cgit/gnulib.git/plain/doc/Copyright/request-assign.future

Re: Including ESS with Emacs

[Richard Stallman]

[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

Stefan has it right.

  > Finally, there will probably be a few hold outs who contributed
  > a non-trivial amount of code which is still present but whose copyright
  > paperwork we can't get.  At that point, we have to take a yet deeper
  > look at what to do with it: rewrite the code, drop it altogether, or
  > move it to a separate package.

At that point, how about telling me about the situation and I will see
what I can do.

-- 
Dr Richard Stallman
President, Free Software Foundation (https://gnu.org, https://fsf.org)
Internet Hall-of-Famer (https://internethalloffame.org)

Re: Including ESS with Emacs

[Stefan Monnier]

I'm continuing this discussion off-list,


        Stefan


Alex Branham <alex.branham@gmail.com> writes:

> On Sun 25 Nov 2018 at 10:45, Stefan Monnier <monnier@iro.umontreal.ca> wrote:
>
>>> We'd like to contribute ESS to the FSF so that it can become part of
>>> Emacs.
>>
>> Yay!
>> [ I think it'd be best to include it into GNU ELPA rather than into
>>   Emacs, tho.  ]
>
> Right, I was including ELPA as "part of Emacs". We'd be completely happy
> being in "core" Emacs or on ELPA, whatever the maintainers think is
> best.
>
>> It should be all authors who contributed more than 15 lines-or-so of the
>> current code.  `git blame` can be a good starting point, but note that it
>> can be misleading: someone who just adds a `save-excursion` and reindents
>> all the code will cause git-blame to hide the contributors of all the
>> reindented code.
>>
>> The way I went about it in the past is to collect all the known authors.
>> Then I remove those who already have signed paperwork, and those who
>> contributed less than 15 lines of code (by looking at the
>> corresponding commits).
>> Then I try to contact all those that remain.
>>
>> This often involves a fair bit of work and time, to try and find current email
>> addresses, to get an answer, to sign the paperwork, to get the FSF to
>> process the documents, ...
>>
>> After that, there are typically some authors still pending, either
>> because they refuse to sign, or because you can't find a valid email, or
>> because they don't reply.
>>
>> At that point, you dig into the code in more detail, to try and see
>> which part of their contribution remains in the current code and whether
>> that can fall under the "15-lines worth of code" (where a renaming of
>> a variable counts as 1 line even if it touches 100 lines).  That often
>> lets you remove a few more from the pending list.
>>
>> Finally, there will probably be a few hold outs who contributed
>> a non-trivial amount of code which is still present but whose copyright
>> paperwork we can't get.  At that point, we have to take a yet deeper
>> look at what to do with it: rewrite the code, drop it altogether, or
>> move it to a separate package.
>>
>> Note that if at the last stage there is still a major contributor, it
>> might render the whole effort pointless.  So it's best to start with the
>> most prolific contributors.
>
> Thanks, this is very helpful! I'll work on generating such a list of
> contributors. What is the best way to go about telling people how to do
> the copyright process? I know for a contribution to Emacs they email
> yall requesting the copyright form but it might be annoying for yall to
> deal with all of ESS's contributors. Does it make sense for you to email
> me the form and then when I reach out I can include the form with my
> email?
>
> Thanks again,
> Alex

Re: Including ESS with Emacs

[Alex Branham]

[-- Attachment #1: Type: text/plain, Size: 520 bytes --]


On Sun 25 Nov 2018 at 19:40, Richard Stallman <rms@gnu.org> wrote:

>   > Finally, there will probably be a few hold outs who contributed
>   > a non-trivial amount of code which is still present but whose copyright
>   > paperwork we can't get.  At that point, we have to take a yet deeper
>   > look at what to do with it: rewrite the code, drop it altogether, or
>   > move it to a separate package.
>
> At that point, how about telling me about the situation and I will see
> what I can do.

Will do, thanks!

Alex

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 487 bytes --]

Re: Including ESS with Emacs

[Augusto Stoffel]

There has been a discussion about including ESS in GNU ELPA a while back
[1] but apparently that effort has stalled due to the copyright
paperwork.

Maybe it should be added to NonGNU ELPA?

[1]: https://lists.gnu.org/archive/html/emacs-devel/2018-11/msg00400.html

Re: Including ESS with Emacs

[Philip Kaludercic]

Augusto Stoffel <arstoffel@gmail.com> writes:

> There has been a discussion about including ESS in GNU ELPA a while back
> [1] but apparently that effort has stalled due to the copyright
> paperwork.

Do you know if the situation has changed since 2018?

> Maybe it should be added to NonGNU ELPA?
>
> [1]: https://lists.gnu.org/archive/html/emacs-devel/2018-11/msg00400.html

Re: Including ESS with Emacs

[Stefan Monnier]

> There has been a discussion about including ESS in GNU ELPA a while back
> [1] but apparently that effort has stalled due to the copyright
> paperwork.
> Maybe it should be added to NonGNU ELPA?

Actually, it stalled later, in 2020 for no good reason.
AFAIK the code was pretty much ready.  In my mail archive it seems in the
last communication with Alex we were discussin details about a cleanup
patch to make ESS buildable "in the usual way" (rather than via ESS's
own makefile, and without breaking if julia-mode is not installed, or
something like that).

Maybe I just dropped the ball.  Alex?  Eric?
I also added Lionel in the Cc since `git log` suggests that he's doing
a fair bit of maintenance lately.


        Stefan

Re: Including ESS with Emacs

[Alex Branham]

[-- Attachment #1: Type: text/plain, Size: 988 bytes --]

Everyone who has contributed to the current instance has signed the
paperwork, so as soon as we release a new version we can add ESS to ELPA.

On Tue, Sep 13, 2022, 5:26 PM Stefan Monnier <monnier@iro.umontreal.ca>
wrote:

> > There has been a discussion about including ESS in GNU ELPA a while back
> > [1] but apparently that effort has stalled due to the copyright
> > paperwork.
> > Maybe it should be added to NonGNU ELPA?
>
> Actually, it stalled later, in 2020 for no good reason.
> AFAIK the code was pretty much ready.  In my mail archive it seems in the
> last communication with Alex we were discussin details about a cleanup
> patch to make ESS buildable "in the usual way" (rather than via ESS's
> own makefile, and without breaking if julia-mode is not installed, or
> something like that).
>
> Maybe I just dropped the ball.  Alex?  Eric?
> I also added Lionel in the Cc since `git log` suggests that he's doing
> a fair bit of maintenance lately.
>
>
>         Stefan
>
>

[-- Attachment #2: Type: text/html, Size: 1353 bytes --]

Re: Including ESS with Emacs

[Stefan Monnier]

> Everyone who has contributed to the current instance has signed the
> paperwork, so as soon as we release a new version we can add ESS to ELPA.

Excellent.  I'll take a look at adding to GNU ELPA as soon as possible,
Thanks,


        Stefan

Re: Including ESS with Emacs

[Philip Kaludercic]

Alex Branham <alex.branham@gmail.com> writes:

> Everyone who has contributed to the current instance has signed the
> paperwork, so as soon as we release a new version we can add ESS to ELPA.

Is there any reason it couldn't be added directly?

Sv: Including ESS with Emacs

[Eric Lindblad]

[-- Attachment #1: Type: text/html, Size: 1666 bytes --]