Re: [regext] DOODLE: select your documents

[Alexander Mayrhofer <[e-mail]>]

Tobias,

Thanks for coming back to my "rant". A few observations inline: 

> However, nowadays most domain registries have withdrawn to the point of
> implementing only their own ideas or approved RFCs. This inevitably leads to
> the situation that proposals for improvement - whoever they come from -
> either have to be solved via approved RFCs, in lengthy bilateral talks /
> negotiations or through policy development.

[AM] You're unwittingly strengthening my arguments here. You're saying "we don't bother about what Standards Track implies. We just need that RFC number so that we can force the industry into doing what we want". Not because of the higher quality reviews involved in IETF work, the well established and accepted process to progress an Internet Standard, etc. It's just to "gold plate" a spec with a "label" that tricks people into believing it's actually important. 

And secondly, you're actually saying "we're doing this so that we can evade the policy development  and discussions around it". Wow. 

> The goal of the ICANN CPH TechOps Group was and is to address technical
> and operational challenges and as there seems to be no other way we
> decided to go for the standardized way. If the REGEXT Group thinks that an
> individual submission with the status informational is sufficient and this is
> fully supported by the domain registries, then I see no problem doing it this
> way and we can save time and effort here.

[AM] I don't know what the group thinks, i'm speaking as an individual here, obviously. Escrow has been implemented widely, and never made it beyond an individual draft. So, it's possible without an RFC number. Yes, there was some force involved into that, obviously, but - on the contrary 
 
> Just let me know how you want to do it.

[AM] First, I'm not disregarding your (or someone else's) work. I have written enough specifications to understand the amount of work that goes into those documents. Ironing out the details once the first implementation is done, fixing contradictions with other specifications, ironing out late corner cases.. Tons of very very useful work.

[AM] Practically, CPH techops could perfectly publish all those file format specifications as their own documents (CPHT-001... CPHT-002... etc..).  If they are good and beneficial for both registrars and registries, implementation will follow. I'm the last one to oppose standardization in that area - quite the contrary, I can think of  (and I'm working on) standardization of more items (Registry Registrar Data Group, Registry Data Nerds, etc...) - but I wouldn't think of submitting the work there to the IETF. That's my only point.

Best,
Alex