Re: bettering open source involvement

[Alia Atlas <akatlas@xxxxxxxxx>]

Mike,

On Fri, Jul 29, 2016 at 9:56 AM, MH Michael Hammer (5304) <MHammer@xxxxxx> wrote:

Alia,

 

Thanks for the thoughtful response. Submitting a draft is indeed a reasonably quick process. But that is only the starting point. My IETF experience is mainly in the area of email authentication so perhaps that colors my perception. I can’t think of anyone that didn’t walk away from the MARID working group (SPF and SenderID) without a sour taste in their mouth. It can be summed up by saying politics and religious wars. I look at how long it took for DKIM to go from draft to standard – without looking up the exact dates I’ll call it something around 8 years. ADSP was a painful experience all the way around and a time suck. I’ll not go into DMARC which has become implemented widely yet ran into fierce resistance within portions of the IETF community. I’m part of the DMARC team that came out with it and the goal was to open up something that was working among private peers so that any person organization of any size could benefit. Instead of working to help address the corner cases, there was an intense effort on the part of some to kill it off and/or stonewall despite increasing acceptance and implementation in the wild.

What I know about the DMARC and email space is merely the public discussions - and I know DMARC was rather sensitive on a number of different aspects.   Of course, it was also trying to solve a problem that is increasingly urgent and frustrating.  Sometimes, there's a tendency here to go for the best being the enemy of the good - but also sometimes there's a tendency to try and push ahead with a point solution that can cause active collateral damage.  I would like to assure you that much of what the IETF works on isn't as challenging to come to consensus on an approach as DMARC was.

 

It appears that rather than trying to find ways to reduce friction, the attitude is that people must dedicate their lives to the IETF alter in order to get things done. Part of this is because to some extent IETF is driven by people who are paid by their organizations to be full time IETFers. My company allows my participation in IETF working groups (as well as other places) but does not “sponsor” it. This isn’t a complaint but rather, recognition of reality. Is it any wonder that so many people who get involved in the IETF because of one particular thing end up walking away from standards work? 

It's actually a rather fascinating assumption - because I don't actually think that there are that many people who are full time IETFers.   In my current role as AD, I basically am; I've considered whether I could decrease that a bit.   There are at least 3 ADs who are not spending full-time on IETF.   Before I was an AD, I was never full-time on IETF work.  It was one of my tasks and focuses (and something I cared deeply about personally), but I wasn't full-time.   I do think there are some IETFers who are close to full-time because of the number of things that they engage in, but not very far from most.

If anything, that most people aren't engaged full-time can be seen in the hiccups and slow-downs of discussions when folks have to dig out from their day jobs to reload context and respond.

 

I wonder if it would be productive to have a survey to collect statistics about participation workload and other common assumptions.  Maybe we'd cut each other a bit more slack if we realized that most of us are in the same boat as far as time and personal commitment?

 

Perhaps my perspective is somewhat jaundiced yet I continue to participate in working groups.

 

I’ve been working on some ideas (surprisingly not in the email or security arenas) that would at first glance appear to be naturals for bringing to the IETF yet I am hesitant because I would likely expire of old age before seeing them get through the process. Life is too short. 

I can't speak for the ART area, of course, but I really don't think that it has to take that long.  One of the parts of shaping new work can be setting expectations on the time to get them done.  Personally, I'd much rather discuss the work and concerns around timeliness than not see the work because of assumptions that we can't improve.

 

Just a few thoughts before I duck and run for cover from an anticipated backlash.

Thoughtful conversation and honest discussion shouldn't cause horrible backlash.

Personally, I'm always extremely interested to hear how others are perceiving things

and really appreciate your engagement.

Regards,

Alia

 

Mike

 

From: Alia Atlas [mailto:akatlas@xxxxxxxxx]
Sent: Friday, July 29, 2016 9:04 AM
To: MH Michael Hammer (5304)
Cc: Melinda Shore; Brian E Carpenter; Suzanne Woolf; IETF discussion list

Subject: Re: bettering open source involvement

 

Hi Melinda & Michael,

 

On Fri, Jul 29, 2016 at 1:51 AM, MH Michael Hammer (5304) <MHammer@xxxxxx> wrote:

> -----Original Message-----
> From: ietf [mailto:ietf-bounces@xxxxxxxx] On Behalf Of Melinda Shore
> Sent: Thursday, July 28, 2016 10:31 PM
> To: Brian E Carpenter; Suzanne Woolf
> Cc: IETF discussion list
> Subject: Re: bettering open source involvement
>
> On 7/28/16 1:06 PM, Brian E Carpenter wrote:
> > And there's our problem, right there. Protocols without APIs are
> > pretty much useless these days. IPv6 without a socket API would have
> > been an abject failure. Without RFC 2133, RFC 2292 and their
> > successors, who knows how the POSIX and Winsock support for IPv6 would
> > have turned out?
>
> Not specifying APIs in the IETF clearly doesn't mean that there are no APIs,
> clearly.
>
> I'm certainly open to the possibility that we start tackling APIs but I'm not
> sure it's a terrific idea.  For one thing, we already have too much work.  For
> another, I'm not sure we'd produce particularly good APIs. It's a different skill
> from developing and specifying network protocols.  And thirdly, I'm not
> convinced that the people implementing our protocols would want IETF-
> developed APIs.
>
> This is completely subjective but my own sense is that the
> #1 problem we have related to open source projects we take years to
> produce specifications.
>

This! +1000

 

That certainly aligns with what I've heard as well, but can I poke into a bit more.

I know that, for instance, I can get a draft from written to the RFC Editor in 6 weeks,

if it isn't controversial.   Most of that time is to allow adequate review at the WG, IETF

Last Call, directorates and IESG levels.

 

My sense is that the rest of the time goes to the WG process which has aspects of:

   a) Getting people interested in the idea

   b) Having folks cycle in and out of paying attention and commenting

   c) Having authors/editors be distracted and unresponsive.

   d) Having WG Chairs be distracted/unresponsive and want more discussion first.

   e) Avoiding having actively hard discussions about contentious points.

   f) (sometimes) waiting for implementations to sanity-check

 

It feels like a WG group or topic in a fixed area with agreement could get past many of these slow-downs - if there were general agreement and a culture in that WG of doing so.

 

We aren't, after all, doomed to repeat the delays of the past :-)  which isn't to say that this would be easy.

 

What do you think?  Are there factors that I'm missing?   Is there a technical topic where there could be enthusiasm to push?

 

Regards,

Alia