Re: [EAI] Rechartering

[Shawn Steele <[e-mail]>]

We definitely want the "right" result :)  My request isn't "just to have a schedule", but rather so that there's something we can plan to.  Also, if "right" takes much longer, it won't matter.  There's a huge user segment that currently doesn't have effective email support because they aren't literate in the Latin script.  It'd be nice if those people could be enabled to experience what we take for granted, and IMHO it's worth risk of not being perfect in order to support them in the next year instead of longer.

I have two concerns regarding scheduling:

1) I hear from CNNIC that China is creating standards based on the Experimental RFCs, due to be published in November.  We may not care or approve of that behavior, but it may be a standard my products have to support, so it'd be nice if it was more in-sync with IETF.  I realize this isn't this working group's problem, but it does impact the behavior of mail in China.

2) Several of our product units are starting to look at "vNext"  For planning purposes I have to guess about when EAI'll finally become a standard.  Mostly it comes down to "is this standard going to happen soon enough that we need to address it in this product cycle?  Or is it going to be uncertain enough that we'll punt it to next release cycle."

As you're probably aware, some products' release cycles take several years.  So if there aren't standards, or at least nearly finished drafts of standards, in the next 4-6 months, then it is likely that EAI will miss our next release cycle.  Even if we started now, it could take some time for EAI support in our products.  If this release cycle is missed, it could be 5+ YEARS before our customers see complete support.  EAI touches almost every product we make, even XBox.  It's a huge amount of work to "do it right".

I think everyone's customers need EAI support sooner rather than later, and I'm trying to build a case for earlier support, but there's no point to us working hard to build something that isn't standardized.  And it's hard for me to tell some executive "We've gotta jump on this" if the only schedule the WG has posted is obviously years behind schedule and there's no end-game in sight.

My thinking:

* Some drafts are in last-call, so that "phase" is pretty much complete.
* 1-2 months for experimenting & creating the report, maybe Aug-Sept?
* Could rechartering happen in parallel?, so end of Sept?
* Then, assuming the Experimental RFC's didn't reveal any gaping holes, could they be fast tracked in 3 months or so?  Maybe by Jan 2010?

Picking months:

* August 2009 - Finish Last Calls
* August - September 2009, continue experimenting and create the report
* September 2009, Recharter for standards track.
* January 2010.  Core RFCs stable, preferably done or last-call.

If the WG doesn't think we're going to have anything by middle 2010, please let me know & I'll stop bugging our product units.  I know IETF efforts are typically very slow, but I think that this schedule would be achievable, with the "right" results, IF we are willing to commit to the work and solving non-ASCII users needs sooner rather than later.

-Shawn



-----Original Message-----
From: Harald Alvestrand [mailto:[e-mail]]
Sent: Friday, July 17,  2009 9:43
To: Shawn Steele
Cc: Alexey Melnikov; [e-mail]
Subject: Re: [EAI] Rechartering

Shawn Steele wrote:
> That's good.  Unfortunately I can't attend :( If the WG can provide a roadmap/schedule at Stockholm, that'd help my planning efforts :)
>
The biggest obstacle to getting a roadmap/schedule has been that we have
very little feedback from implementors on when they expect to feel that
they have tested the implementations in practice enough to give feedback
on the documents.

The most critical part is getting the experience of how non-EAI users
(and their software) react to downgraded messages.

I'm very unwilling to put up a schedule just to have a schedule -
experience shows that this easily gets into the mode where the chair
pushes the writers, who then write whatever is needed to get the work
done, no matter what the "right" result should be.

We need more mail from our implementation teams.

                         Harald