On Jun 19, 2013, at 10:01, Paul Hoffman wrote:
But there is no EPP WG. And WEIRDS is supposed to only be forward-looking,
not dealing with practices with the current protocol.
Brief history and then maybe a point. (Written as one of the co-chairs of the
PROVREG WG.)
In December 2000 a BoF was held to start the work on EPP at the San Diego
meeting. A few months passed by before the PROVREG WG was formalized. I was
told by a then-IESG member that the decision to form the group was contentious
because EPP was to be a limited scope (B2B if you will) protocol and not
something for general use.
The WG progressed by 2003 to have a nearly complete specification when it ran
into IESG pushback. I was told that the group was nearly shutdown despite the
WG having come to consensus. In March 2003 the group met for the last time,
nailing down an acceptable (to the IESG) set of documents.
The group formally closed in 2004 once the drafts passed the RFC Editor queue.
At the time, the queue was quite lengthy (in duration) and in fairness there
was a dependency on a trailing document.
In 2009 (5 years hence) at a ccTLD registry meeting the chorus of "EPP stinks"
got quite loud. What made this irritating was the EPP had progressed to Full
Standard (Proposed, Draft and then Full) in full accordance with IETF
procedures and policies. So I took on a little exercise to figure out why, on
the one hand, operators dissed EPP but the protocol engineers elevated it.
First, once the PROVREG WG was shutdown the mail list was kept open. At the
same time (2004-2009) the industry EPP that supports blossomed. The new
majority (operators coming on line then) were not aware of the mail list as the
list had no cover, no publicity, no "home" if you will. This pretty much was
the root cause, as it spawned a side effect.
The material deficiency surrounding the chorus of "EPP stinks" was due to the
new majority not having access to the originators of the documents and the lack
of publicity for the "5th" document. There was conventional wisdom built of
how to extend the protocol that was not in-line with the intent of the design
and this was causing, well, chaffing. Was there a document describing what
should have been done? Yes, the "5th" document. The first four EPP documents
were the ones elevated through the standards process while the 5th was just
informational. As the first 4 were elevated, the 5th was forgotten, again,
because the link to it was lost. (RFP's would cite the 4 latest ones, the Full
Standard, and not the original 5th.)
Once the new majority was "rejoined" with the mail list communications were
re-opened between the new majority and the originators. The technical issues
were quickly solved (at one of the "formal" Bar BoFs in spring 2003/Anaheim).
With the publicity surrounding some open-source EPP software the plethora of
extensions became less of a concern to the operators. (Software trumps spec!)
To underscore the point, despite sporadic traffic in the 2004-2009 timeframe,
today the list is moderately active. There are new phases of operations that
EPP supports, so there is constant growth in engineering.
At the next round of the ccTLD meeting that started my task, all of this was
explained. The alternative was, continue on with this re-patched relationship
or try to invent EPP 2.0 - meaning a whole new development, deployment, and
adoption effort. The consensus was, forget EPP 2.0, the situation was now
"good to go." There's been no more choruses of "EPP stinks."
Looking back in hindsight, what would help is to have some means for the IETF
to provide a maintenance vehicle for it's products. Or realize that the
"waterfall model" that seems to be in place is no longer appropriate. (As if
you've never heard that before!) The world changes (the new majority) but the
IETF acts as if "once it's an RFC it is done."
This is an example of an ICANN initiated piece of work that barely got into the
IETF, the IETF completed it in a way that has benefit beyond ICANN (meaning
many ccTLDs have adopted it on their own accord), but the IETF didn't make it
easy and didn't help the deployment. I hope the latter phase isn't repeated
with the WEIRDS WG and RDAP.
PS. With WEIRDS there's a much more substantial industry and majority than
"back in the day". Just something to keep in mind.
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Edward Lewis
NeuStar You can leave a voice message at +1-571-434-5468
There are no answers - just tradeoffs, decisions, and responses.