--On Tuesday, January 5, 2021 21:43 +0000 Stephen Farrell
<stephen.farrell@xxxxxxxxx> wrote:
>
> Hi Kyle,
>
> On 05/01/2021 17:46, Kyle Rose wrote:
>> Again, no one in any WG is required to interact with GitHub
>> to contribute.
>
> I don't think the above is true in practice, at least in some
> cases. In one case where I'm implementing a draft I've found
> the github issue discussion effectively excludes me from
> really being part of the discussion.
>
> Github issue discussions do work fine for many people, and
> could work fine for me some other time (e.g. if I and others
> involved were in similar TZs and online at similar times),
> but my experience is that even with the best will in the
> world, the split-discussion or exclusionary effect PHB
> complained of can and does happen.
>
> I don't think this is only because the github->mail tooling
> is worse than useless (for Rich: few mins ago, having been
> offline for an hour or two, I got 40 mails for one repo to
> which I'm subscribed;-), I think there are problems with
> timezones too (hence this mail).
>
> I don't have a good solution for that. Equally, I'm not
> asking that we stop with git or github - I just think we
> need to do more work to better take into account that the
> github issue discussion workflow is not everyone's workflow.
Right. And, while both have been said a little differently
before (in this threat and in the discussion leading to RFC
8875, let me add two comments to Stephen's:
(1) At least for someone with a moderate level of skill (I can't
speak for those with high levels of skill) active github
discussions, especially tracker-type ones, are far easier to
follow if one is doing so more or less continuously than they
are for someone who is looking in much less frequently. For
better or worse, when email shifts from fewer messages in which
people are listening carefully to each other to rapid
turnarounds of often short and repetitive messages with signs of
increasing tempers and frustration, email gets very hard to
follow too. However the email usually still seems easier to get
the gist of what is going on by sampling a stream of messages
than trying to decode the history of github discussions and
changes. I don't know whether 24 messages on this theme in a
tad over six hours is an illustration of anything.
(2) There is a risk in the IETF process (and many like it) of
reaching consensus by some core group (who basically agree
within the group) driving everyone else off. A group that
agrees on everything but small details, and maybe on those, can
be extremely efficient. After all, controversies take time to
resolve and the resolution process may not be efficient at all.
But one that gets that efficiency by excluding (even if
completely unintentionally or by choices of style, tools, and
meetings and real-time communications over working
asynchronously is often not, over time, doing the IETF or the
Internet any favors.
john
