[rfc-i] [Ietf-and-github] RFC Editor & Github

Eric Rescorla ekr at rtfm.com
Thu Feb 21 20:10:30 PST 2019


On Thu, Feb 21, 2019 at 7:04 PM Joel M. Halpern <jmh at joelhalpern.com> wrote:

> The RFC Editor (more specifically the production center) already
> conducted an experiment in using github in conjunction with the editing
> process.  While I will leave it to Heather to provide details, the
> simple result was that it turned out to be VERY painful.
>
> Please do not assume that using github / git for that part of the
> process is a trivial or easy step for us to take.
>

As one of the main people involved in that experiment, I have mixed feelings
about it.

On the one hand, it was quite painful. On the other hand, in many respects,
it was quite a bit easier for me than previous experiences with AUTH48
because
it made it possible to precisely localize and respond to every proposed
change.

A full analysis of the situation is certainly out of scope for this WG, but
in short,
one difficulty with the AUTH48 process is the intermixture of different
kinds of
changes, for instance:

- Boilerplate-type changes
- Reformatting
- Copy-editing
- Pagination/typesetting
- Semi-substantive changes by the AD and editors

Having to work with these all at the same time makes it quite difficult to
properly
assess and manage each change. This is the kind of thing that can be
improved
by tooling, but it also requires a certain level of discipline about
dividing up each
type of change that we didn't do in the previous github experiment, so I
think that
any future experiment would need some more planning along these lines.

-Ekr



Yours,
> Joel
>
> On 2/21/19 9:25 PM, Martin J. Dürst wrote:
> > On 2019/02/22 07:06, Mike Bishop wrote:
> >> Ideally, I'd see AUTH48 as a PR (perhaps from a private repo). One of
> the things we'll need to work out is to what extent that's feasible for the
> RFC Editor.
> >
> > I'm of course not in any way speaking for the RFC Editor. Also, I'm
> > myself still struggling with git and github, so I don't want to push
> > (sic) it on anybody else.
> >
> > But I think it would be highly beneficial for the RFC Editor to try to
> > slowly move to (also use) git(hub). They may take quite some time
> > because they have other things on their plate, and have to make sure
> > they move carefully, and establish a process that works for them.
> >
> > But compared to the OLD-NEW style of talking about diffs, pull requests
> > are way more precise, and that's what matters for their work. If they
> > can reduce the number of hickups in the back-and-forth with the editors
> > in Auth48, that may help them quite a bit.
> >
> > Regards,   Martin.
> > _______________________________________________
> > rfc-interest mailing list
> > rfc-interest at rfc-editor.org
> > https://www.rfc-editor.org/mailman/listinfo/rfc-interest
> >
>
> _______________________________________________
> Ietf-and-github mailing list
> Ietf-and-github at ietf.org
> https://www.ietf.org/mailman/listinfo/ietf-and-github
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.rfc-editor.org/pipermail/rfc-interest/attachments/20190221/90aa77e0/attachment-0001.html>


More information about the rfc-interest mailing list