I'm coming down on the side that the strong language against changing
drafts is unjustified.
Moreover, I prefer the name change because at the moment one could
implement the -02 draft, looking for "_ssp" and applying that algorithm,
*and* the -03 draft, looking for "_asp" and applying that algorithm, *and*
the -04 draft, looking for "_adsp" (or whatever it will be) and applying
that algorithm. One could moreover put the selection in the hands of the
user *or* implement all of them and return some sort of consolidated
result. Who knows? We might even be able to collect statistics that
reveal meaningful conclusions about one algorithm over the other.
Changing the draft (which is, after all, only a draft and thus
experimental) shouldn't shatter any production implementations. Ripping
out -02 support and adding -03 didn't take me very long at all, maybe an
afternoon, and it also wouldn't be a giant deal for me to exhume the -02
code and run them in parallel.
I've left my _ssp records up in my test domains too, so interoperability
at least with my installations shouldn't be broken even if they change the
name another dozen times.
This answers JD's question:
Are there any current implementations of asp in the wild?
Yes, mine is one. There may be others, but I've not heard of any.
_______________________________________________
NOTE WELL: This list operates according to
http://mipassoc.org/dkim/ietf-list-rules.html