I'm not addressing this specifically to Pat since he knows it all perfectly.
SNA "addressing" relies on two components - and it uses names not numbers. Within an enterprise, a naming authority is assigned which allocates - these days - just 8-character LU names.[1] The enterprise in turn is supposed to seek permission from a higher naming authority to provide it with a so- called "network identifier" (NetId) with which to "qualify" the names within the enterprise. IBM elected to be the "higher naming authority" and structured the NetId so that the first two characters indicated the country. (I believe there's some standard for this country code.) The next 6 characters are then available for the IBM entity responsible for the country to assign on request. Obviously the major enterprises can "cherry-pick" their codes but sequential allocation could apply at some level. In principle this offers 3 518 743 761 (39 to the power of 6)[2] NetIds within each country code - sufficient down to the level of the individual citizen even for countries such as China and India but additional "country codes" could be assigned if necessary. With IPv6 IP has removed the intrinsic limitations of IPv4, where the bountiful numbers (256*4 = 4 294 967 296) absolutely, that is, worldwide, are drastically reduced by having to structure the allocation of numbers in order to support IP routing - as well as assigning numbers within the range for special purposes such as broadcasting within a LAN "network". Since neither the assigned IP address number nor, when used extensively, the SNA name is or could be, respectively, "user-friendly", a friendly name to address/unfriendly name service is/could apply to both. Chris Mason [1] Some may argue, even using SNA official documents, that the CP name also needs to be considered. It has been a feature of my teaching of APPN that the CP is just an LU which happens to use certain restricted mode names when performing functions needed for the operation (and management) of APPN protocols. When other mode names are used, that same session end- point entity (LU/CP) can perform application duties. It's only because, alone among APPN platforms, VTAM can't actually do that - except for the APING application - that architecture had to get so convoluted regarding the CP vis- à-vis the LU. [2] 39 derives from 10 numbers, 26 alphabetic characters and the three "specials", "$", #" and "@". On Sun, 9 Aug 2009 16:42:39 -0500, Patrick O'Keefe <[email protected]> wrote: >On Sun, 9 Aug 2009 16:03:33 -0500, Paul Gilmartin <[email protected]> wrote: > >>On Sun, 9 Aug 2009 11:57:15 -0400, Anne & Lynn Wheeler wrote: >>> >>>possibly SNA organization viewed it as competition (even tho SNA had >>>nothing to do with networking). >>> >>Now I'm confused. What does the initialism "SNA" stand >>for? >> >>Or, while this list is focused on initialism pedantry, >>is it possible that there's another "SNA" than the one >>I found at the top of a Google search? >>... > >Well, the N is "Network", not "Networking", but I don't think that >clarifies anything. Lynn apparently has some very specific defibition >of "Networking" in mind. His comment may be accurate (His comments >usually are.) but I'm not sure what that comment really was. > >SNA does not have a universal addressing scheme and IP does. >Perhaps that was the point. But SNI provided that in sort of the >same way that NATing allows interconnection of 2 IP networks that >use private IP addresses. > >SNA does not provide a universal name space for resources, but >neither does IP. The Domain Name space used by IP hosts is >not provided by, or dependent on IP. > >Pat O'Keefe ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
