Wow.. i'm out of touch for a couple of days and what a string of information to digest when i get back...
I add my thanks to Kepa as well - and i'm old enough to have experienced the beginnings of DNS. I think the main reason that it was rapidly accepted and flourished is that nobody really wanted to continue to carry the burden of being a heavy-hitter email switch - there's no money in it (and we are very thankful to those that did step up to the plate early on). William is correct in his assumption that CHs will not be willing to share addresses with competitors, and they certainly won't share them with their clients (like us), since we are interested in submitting directly to those payers of our larger volumes of claims. I've already been down that street... I really do like the DNS concept, however, and while it may not be a short-term solution implementable this year, it has an excellent potential. I don't believe the security concerns pose an insurmountable barrier, and I believe that there can be some protection built in if our provider numbers (which are publicly available) can be used in some way by payers to validate our transmissions (perhaps as part of a PKI identity check?). Anyway, without trying to solve the problems, i'd like to weigh-in in favor of pursuit of the DNS concept -- even if it has to be a "phase 2". One other thought on which identifier is in the ISA - I believe it was Rachel who was favoring having the endpoints identified there rather than the single-hop communicating parties. This makes sense to me, and is certainly similar to how we handle it in the HL7 world on the MSH. The concern i have, however, is that it precludes a receiver from being able to check to see if they were actually the intended party to receive the message. I could accidentally send a package meant for payer A to CH B that also submits to payer A. CH B would have no way to detect that this is a mistake, and would proceed to process the payload. Worse, I could send a package meant for CH B to CH C, who has some, but not all of the payers in their stable as well, and end up with a mess that would take weeks to unravel... Dave Minch T&CS Project Manager John Muir / Mt. Diablo Health System Walnut Creek, CA (925) 941-2240 -----Original Message----- From: William J. Kammerer [mailto:[EMAIL PROTECTED]] Sent: Wednesday, January 30, 2002 5:33 PM To: WEDi/SNIP ID & Routing Subject: Re: some history on email addresses and thoughts for a proposal Thanks to Kepa Zubeldia and Dick Brooks for their enlightening historical perspectives on the evolution of DNS. I'm thankful the graybeards of our industry are around to share with us memories of the old days, for surely otherwise, hubristic and callow youths like myself would make the mistake of thinking history began with them. I do see another problem with the DNS proposal, besides the security concern posed by Dick and the time constraints raised by Rachel Foerster: would you share your customer list? If not, you can see why VANs and Clearinghouses might be loathe to make their complete customer list available on distributed DNS servers or an LDAP directory. I thought I was pushing it when I proposed auto-interconnection back on 20 January, whereby one VAN or CH could ask another whether an entity identified by a certain identifier was accessible via their system. I could see VANs or CHs fighting that small reform, unless otherwise compelled by HIPAA law to share that information. Never have I imagined that a switch would expose the contents of their entire directory (or customer base) - their "family jewels," if you will - in an open directory! William J. Kammerer Novannet, LLC. +1 (614) 487-0320
