[J3] (SC22WG5.6209) Re: [EXTERNAL] Re: RE: [ukfortran] October meeting visa invitation letter
Bill Long
longb at cray.com
Fri May 1 16:50:19 EDT 2020
> On May 1, 2020, at 2:10 PM, Ondřej Čertík via J3 <j3 at mailman.j3-fortran.org> wrote:
>
> Dear Committee,
>
> Bryce, thank you for this detailed explanation how C++ does it. That's precisely what I am trying to get the Fortran Committee to eventually do also. But step by step in an evolutionary matter, so we first created the J3 GitHub repository, tried it for a few papers, it was a success, then we'll try for more papers, until we get there.
And where is “there”? discussing and submitting all the papers on GitHub?
>
> A few people suggested this is a revolution. It's not. This is a well tested workflow, from other committees, as well as our own Fortran Committee on a few papers. Revolution would be if we got rid of the j3-fortran.org website and did not post papers there at all, and only used GitHub. That's not what we are doing. We are supplementing the primary place j3-fortran.org to discuss papers with a secondary place at GitHub to track each idea / papers and have a platform for discussion / work between meetings. But at meetings, we are still using j3-fortran.org to submit papers and to vote on them.
And, yet, replying to these emails, not to mention discussion of papers, is already taking WAY too much time away from day jobs. If the goal is to have the “community” composed mainly of retired people, and ones who have very lax, or no, job, then it might be reasonable. But that excludes a lot of people.
>
> Steve contemplated perhaps hosting some board ourselves. Please don't do that, unless you understand well all the Cons of that approach:
>
> * we will have to build a community again there --- it was a *lot* of work that I and others had to put into the J3 GitHub site. Somebody will have to put in the same amount of work to make it a success.
No. You just put a message on the current GitHub saying that the project has moved, and click here to go there. Same community. No new building. Just one click.
>
> * it will not be as easy for people to contribute, because people have to create yet another login that they have to remember, and not just committee members, but every single member of the wide community that we want to attract. All of them already have logins at GitHub. In fact, every voting member of the J3 committee already has an account at GitHub, see here where I collected all the GitHub IDs: https://github.com/j3-fortran/fortran_proposals/issues/155, the only voting members that are not there yet is Dan, Van and Bob. I am happy to help each of them personally (off the mailinglist).
No one remembers logins and passwords anymore. Keychains, or Firefox remembers them. I use Githib so infrequently that when I did log in, Forefox remembered, and I got through. I would have never “remembered” my old login information. And, for someone who insists on typing in the information, just make the login name and password the same as you currently use on the LANL GitHub site and there is nothing new to remember. Odds are good that is what most people would do anyway.
>
> * it will be yet another place to discuss things.
No, not if the old site is shut down. Everything happens at the new place. Only one place. And not hosted at a problematic location. (Except that we already have other places, like the Fortran90 email list.)
>
> * We have to actively maintain the service. That's non-trivial time that somebody has to invest into it.
We only have to maintain the new service.
>
> Bryce explained in another email that there is no perfect solution, and in practice, GitHub does work well. If, however, there comes up an insurmountable obstacle in practice (not in theory) why we cannot use GitHub, only then let's bite the bullet and move to our self-hosted GitLab, that Reuben and I offer to maintain for the Committee, that will provide all the features that we need from GitHub, but it will be open source software, hosted by ourselves (but with all the Cons mentioned above).
I thought it was Steve, with Reuben’s help, who was going to manage the GitLab site.
>
> Until then, please let's just continue using GitHub. It works really well, and instead of starting over, let's rather keep building the community and invite them to submit and collaborate with us on even more papers for the next meeting.
A lot of people are are not using the GitHub site regularly. For various reasons. They (we) have day jobs, and it is a lot less efficient than face-to-face at the physical meeting, for starters.
Cheers,
Bill
>
> Thank you,
> Ondrej
>
> On Thu, Apr 30, 2020, at 5:53 PM, Bryce Adelstein Lelbach aka wash via J3 wrote:
>> There are plenty of ISO and INCITS working groups that conduct
>> business on GitHub.
>>
>> For ISO WG21 and INCITS PL22.16 (C++), we have mailing lists, and we
>> make extension use of GitHub.
>>
>> https://github.com/cplusplus
>>
>> Every paper is tracked by a GitHub issue:
>>
>> https://github.com/cplusplus/papers/issues
>>
>> Using wg21.link, our lookup service, you can easily get a paper and
>> locate the GitHub issue and status for a paper:
>>
>> http://wg21.link/P1255 <- link to the latest revision of the paper
>> http://wg21.link/P1255R1 <- link to revision 1 of the paper
>> http://wg21.link/P1255/github <- GitHub issue and current status for a paper
>>
>> As you can see from the last link, the GitHub issue for each paper
>> includes a comment with a summary of each discussion of said paper,
>> including poll results and the guidance to authors. The GitHub issue
>> tracks the progress of the paper, from initial incubation, through
>> design review groups, through wording review groups, and ending with
>> the issue being closed after a pull request has been made to the
>> sources for the standard applying the change.
>>
>> National Body comments are processed on GitHub:
>>
>> https://github.com/cplusplus/nbballot
>>
>> Issues are tracked via GitHub:
>>
>> https://github.com/cplusplus/issues/issues
>>
>> The Latex sources for the standard are on GitHub:
>>
>> https://github.com/cplusplus/draft
>>
>> The Project Editor and his assistants manage changes to the standard
>> via pull requests:
>>
>> https://github.com/cplusplus/draft/pulls
>>
>> Major editing projects are tracked via GitHub projects:
>>
>> https://github.com/cplusplus/draft/projects/1
>>
>> Subgroup agendas for face to face meetings are created using GitHub projects:
>>
>> https://github.com/cplusplus/papers/projects/19
>>
>> ISO WG21/INCITS PL22.16 is about 1 order of magnitude larger than
>> WG5/J3; we have about ~250 people from ~15 national bodies and ~25
>> active study groups at each of our face to face meetings. We have
>> people from all sorts of backgrounds, organizations, and countries,
>> with all sorts of different requirements. Some people can't or won't
>> use any Google services; some people disable Javascript in their web
>> browser. These limitations have not prevented us from making use of
>> organizational tools. The number of such individuals is small, so
>> we're able to make accomodations, and not everyone has to use or have
>> access to everything. Some people just participate via the mailing
>> list, some people just participate at face to face meetings, some
>> people just participate on GitHub or Slack. The only people who need
>> to have access to and use all of the different services are the
>> officers and leadership.
>>
>> On Wed, Apr 29, 2020 at 11:00 AM Van Snyder via J3
>> <j3 at mailman.j3-fortran.org> wrote:
>>>
>>> On Wed, 2020-04-29 at 15:10 +0000, Bill Long via J3 wrote:
>>>> some people have an aversion to, and concerns about, products from
>>>> Microsoft (of which GitHub is one, and Skype another). Google and
>>>> Facebook have acquired tarnished reputations as well. And don’t get
>>>> the cyber security folks started about ZOOM.
>>>
>>> I believe IEEE P1722 did everything via a-mail. I believe they did not
>>> have any face-to-face meetings. Ultimately, formal "motions" were
>>> circulated, which contained edits. These were passed (or rejected) with
>>> formal e-mail ballots.
>>>
>>> I haven't asked my contacts in WG9 and ARG, but it seems the Ada
>>> committees do a lot of work via e-mail. They have a yearly meeting,
>>> usually in conjunction with Ada Europe and the Conference on Software
>>> Reliability, about three days total. I don't know for sure, but based on
>>> the number of different names I see in e-mail traffic, I think WG9 has
>>> more members than WG5.
>>>
>>> Not to say the WG5 and J3 could switch to that mode tomorrow, or that we
>>> would find it useful to move rapidly in that direction, but some people
>>> have made it work.
>>>
>>> At least we ought to try to get more done via e-mail between meetings,
>>> especially so that we don't get ambushed on Friday after working all
>>> week on a project.
>>>
>>>
>>
>>
>> --
>> Bryce Adelstein Lelbach aka wash
>> US Programming Language Standards (PL22) Chair
>> ISO C++ Library Evolution Chair
>> CppCon and C++Now Program Chair
>> CUDA Core C++ Libraries (Thrust, CUB, libcu++) Lead @ NVIDIA
>> --
>>
Bill Long longb at cray.com
Principal Engineer, Fortran Technical Support & voice: 651-605-9024
Bioinformatics Software Development fax: 651-605-9143
Cray, a Hewlett Packard Enterprise company/ 2131 Lindau Lane/ Suite 1000/ Bloomington, MN 55425
More information about the J3
mailing list