Current NSAP stuff and the original NSAP allocation scheme follows: (current NSAPs first then the old scema): -------------------------------------------------------- NSAPS HOST NAME-ADDRESS TABLE uk.ac.edinburgh.krypton uk.ac.ed.krypton LSAP: 08000901E2BA7E X 3882611000015099001001000101 B 3882611000015099001001000102 G 3882611000015099001001000103 uk.ac.edinburgh.its63b uk.ac.ed.its63b LSAP: 0207010038E37E X 3882611000015099001001000201 B 3882611000015099001001000202 G 3882611000015099001001000203 uk.ac.edinburgh.itspna uk.ac.ed.itspna LSAP: 000081D780017E X 3882611000015099001001000301 B 3882611000015099001001000302 G 3882611000015099001001000303 uk.ac.edinburgh.chapel uk.ac.ed.chapel LSAP: 08006D0004877E X 3882611000015099001001000401 B 3882611000015099001001000402 G 3882611000015099001001000403 uk.ac.edinburgh.etive uk.ac.ed.etive LSAP: 02CF1F1047357E X 3882611000015099001001000601 B 3882611000015099001001000602 G 3882611000015099001001000603 uk.ac.edinburgh.servilan uk.ac.ed.servilan LSAP: 0800141131517E X 3882611000015099001001000701 B 3882611000015099001001000702 G 3882611000015099001001000703 uk.ac.edinburgh.klingon uk.ac.ed.klingon LSAP: 00DD0043F7007E X 3882611000015099001001000801 B 3882611000015099001001000802 G 3882611000015099001001000803 uk.ac.edinburgh.phyvax uk.ac.ed.phyvax LSAP: aa???????????? X 3882611000015099001001000901 uk.ac.edinburgh.emas-a uk.ac.ed.emas-a LSAP: 0201000010007E X 3882611000015099001001001001 B 3882611000015099001001001002 G 3882611000015099001001001003 gadmin gadmin LSAP: 0800390010007E X 11 uk.ac.edinburgh.gate uk.ac.ed.gate LSAP: 0800390070007E X 00 -------------------------------------------------------- Old scheme I was asked to produce a list of all the NSAPs in use on the LAN etc. I would suggest the following strategy for the future (and refer to Adam's mail on the structure of NSAPs which I have appended): 1. Allocate some test NSAPs which can be used by anyone for testing e.g. the machine number field as 9990 to 9999? It would also be extremely useful to assign some test MAC addresses - since the York code will not permit LAN calls to something which is not configured - WHY? 2. Easy one - allocate new field for the JCMB, Engineering etc. LANs (the 15099001 in Adam's note is the ED CS LAN) - assuming that we wish to go that far in addressing. 3. Before we go any further, recode the context byte as in the NRS Lookup Protocol (Brown Book) (Sorry Adam, but I feel that this will lead to less confusion in the future when one might wish to cook or manipulate an NSAP). This would mean changes in the numbers below. Namely, 01, 02 and 03 context would become 00, 02, 03. This also implies that the non-NRS contexts in Adam's note must be re-assigned to something higher than 16 - suggest 64+n? 4. I have not yet added PADs to the scheme. Since at this time it is not clear whether we need to register all PAD ports, or only the PAD, I intend to use the same scheme as before but to APPEND a tty port id pair i.e. have a root m/c id. as before and add a 'tag'. UNLESS there are demurrals? What about soft pads? The following is that list as extracted from the /usr/lib/x25/directory.local file on ITS63A. Note that some of these numbers are no longer in use (e.g. the convex - which may return!) and the testemas number is cooked up. I have separated out the fields in the NSAP as per Adam's document with a space for clarity - this space is not part of the NSAP! x=X.29; F=NIFTP; M=NIFTP-MAIL. --------------------------- Adam's Note ----------------------------- (GH: Items '*'ed have been commented on above) Subject: NSAP structure to be used for Pink Book etc. From: adam Date: 2 Jul 1987 12:09:52-GMT digits meaning contents 2 for AFI BCD encoded ISO-DCC 38 3 UK 826 4 ac 1100 2 uk.ac format id 00 * 8 ed 15099001 issued by B.Gilmore 3 dept 001 = Comp.Serv. 002 = Comp.Sci. 003 = EE ........etc 4 m/c 0001 its63a 0002 its63b 0003 itspna 0004 chapel * 2 context/service 01 X29/TS29 02 FTP 03 FTP.MAIL 04 FTP.NEWS 05 FTP.JTMP 10 TC4 11 TC0/2 20 CLTS Total 28 digits, leaving 4 spare and 8 for accounting. Now all we need to know is what T-selectors and S-selectors look like.