He's not asking for anything. He's pointing out how a different data model allows for solving a different set of problems. The MVCC page on Wikipedia might give you a better background so that you're able to contrast the two models and determine for yourself where they are useful: http://en.wikipedia.org/wiki/Multiversion_concurrency_contro...
For a simple registrar that supports DNSSEC I've been quite happy with https://www.gkg.net/. Their prices are reasonable and their site is quick and functional, though it's navigation is a bit clumsy. I can't comment on their hosting (including DNS) as I've not used it.
I'm going to continue ignoring DNSSEC until I see it affect clients in a meaningful way. Until then I think there is other things my users would prefer I spent my time on.
The entire point of this story is that until this week your clients couldn't have used it even if they wanted to and thatchanged.
I'm not saying you have to go running off and caring about DNSSEC now, but the kneejerk "omg it's not deployed" argument really isn't helpful or insightful attached to a story about how it just finished getting deployed for real on the roots finally after years of waiting.
Your reading of my post as "omg it's not deployed" is an unwarranted exaggeration. I merely stated that for me, someone who maintains a from the wire-up DNS implementation, this baby-step doesn't change anything practical — which is something you've agreed with.
EDIT:
It goes without saying but I'll state it anyway, this cascade of down-votes is not the reaction I'd expect for making what I consider to be a pretty innocuous comment about something that potentially affects the priorities of my startup.
It goes without saying but I'll state it anyway, this
cascade of down-votes is not the reaction I'd expect for
making what I consider to be a pretty innocuous comment
about something that potentially affects the priorities of
my startup
It's probably a reaction to your advocacy for ignorance at the expense of a story marking an important milestone in DNS's capabilities.
If you insist on not caring, do so quietly. Some people around here are trying to fix things, or break them, which hopefully will result in better systems. In any case, you appear to be doing neither.
(As a side note, your assertion that I agree that this doesn't change anything practical is false, as of yesterday people can resolve domains over DNSSEC, that wasn't true last week. I consider this a practical change. I'm not a big fan of DNSSEC as a protocol, but this is a big deal(tm).)
Last week: No one could use DNSSEC. No one did. No one was affected.
Today: Many people can use DNSSEC. Some actually are. Some clients are affected. (Just obviously not yours, since you remain completely uncaring about the subject.)
Your assessment of what constitutes "many people" and mine clearly differs.
You are quite right that my clients are unaffected. I like most of the world run a mix of Windows and OS X which like the software that runs on top of them, are not DNSSEC aware (with few exceptions).
Clearly I do care about the subject and I'm not sure why you keep stating that I do not. There is a difference between ignoring a technology until it's useful and ignoring it full-stop.
My bad, I get excessively terse when I'm tired. I wasn't suggesting it out of utility but out of appreciation — nothing says thank you quite like a superfluous vanity domain.
EDIT: Just noticed my other post has garnered at least one down-vote so although I'm pretty sure this idea doesn't have legs, on the off-chance you dear reader are one of ~13 other folks who'd like to see this happen, drop me an email to express your interest.
Although the identifying info from those screenshots is scrubbed, the comments say the R-URI is basically user@ip:port. This is interesting: there's no SIP proxy in between.
Unfortunately I see "Bandwidth Limit Exceeded" when I go to that page but your comment makes me wonder whether they are doing SIP URI discovery via DNS-SD[1] given their penchant for DNS-SD.
Charlie Cheever's comment "Python data structures map well to JavaScript data structures" resonates with me. I'm building something in Erlang right now and having come from Python, it's just plain grating.
Gotta wander off on a tangent here — scribd is a beautiful site! I'm probably the last one around here to find out since this is the first time I've looked at their site since they dropped Flash (which I ignored) but wow, what a difference — if you found scribd not to your taste and haven't been back since, it's time to revisit it.
