Re: "community" for the RFC series

[Stephen Farrell <[e-mail]>]

Hiya,

On 04/10/2019 19:21, Christian Huitema wrote:
> 
> On 10/4/2019 2:31 AM, Stephen Farrell wrote:
>> On 04/10/2019 08:51, Christian Huitema wrote:
>>> I have heard Brian Carpenter's argument that if there is not an
>>> authorship community, there is a readership community. That leaves me
>>> skeptical. Clearly, authors and publishers should care about their
>>> readership, and I wish we had better ways to assess the impact of our
>>> publications. But passive readership does not create a community, no
>>> more than me reading ITU publications makes me part of the ITU
>>> community. What creates a community is engagement, contributions and
>>> sharing.
>> I guess I disagree with you there Christian - ISTM that
>> at the very least, people who read RFCs and write related
>> code that is part of many network stacks, but who do not
>> engage with the IETF or RFC editor at all, do deserve more
>> consideration than you imply. I can see arguments for a
>> bigger set of people deserving consideration but omitting
>> the above example set seems just broken to me.
> 
> 
> Sure, 

As in: you agree that particular set of people do deserve
consideration in the upcoming discussion? If so, I think
that's correct.

But, that implies we should also be thinking if there are
other sets of folks (adding up <<7.7 billion:-) who similarly
deserve consideration.

> but if they don't somehow communicate, how do you know they are there?

For the set I mentioned? Do you really doubt they're there?
I do not.

> And if they do communicate, the question is "with whom"? 

I don't think that's a question we need address at all. That's
just their business.

> Where do they
> send the message saying that they are trying to implement protocol FOO
> but they don't get what section 3.1.5 of RFC XXXX really means?
> Slashdot? Stack overflow? Some Reddit group? 

Yes. stackexchange etc. mostly at present it seems. It'd
once have been things like Dr. Dobbs magazine and/or
comp.<various> via usenet I guess.

> Actually, it would be very
> nice if the IETF had a documented feedback channel for such exchanges.
> That would be a nice way to grow the community.

We did discuss exactly that (a stackexchange <something>) at
the IAB retreat - not for IETF process-wonkery like this, but
for protocols that are widely used like HTTP. I think someone
there was gonna look into it but can't recall if anything's
been done since. I can totally such a channel working for HTTP
or maybe DNS, but I'd guess not for this discussion. Might be
worth a try regardless though, even if it didn't work for this
discussion, we'd learn HOWTO.

I do however have a glib answer for how we could try communicate
with people who read some RFCs but who don't otherwise get at all
involved... we write an RFC. Maybe one describing the upcoming
discussion and try see if promoting that in various places gets
any feedback. So there may be a case for writing charter-like
text for the upcoming discussion in an RFC (I guess that'd be
an IAB-stream RFC if we did it).

For other subsets of people, other mechanisms will I guess be
needed, hence me wondering if it'd be good to try characterise
better who might really be in that "bigger" community. We seem
to have identified at least one set of people already so maybe
it'll not be too hard to find more sets or agree we've found
enough.

Cheers,
S.



> 
> -- Christian Huitema
> 
>