Path: utzoo!utgpu!water!watmath!clyde!att!rutgers!ucsd!ucbvax!decwrl!hplabs!hp-pcd!uoregon!dboyes
From: dboyes@uoregon.uoregon.edu (David Boyes)
Newsgroups: comp.mail.misc
Subject: Re: UREP help sought
Message-ID: <2316@uoregon.uoregon.edu>
Date: 6 Jul 88 09:07:37 GMT
References: <579@n8emr.UUCP> <403@cseg.uucp> <9830@g.ms.uky.edu>
Reply-To: dboyes@drizzle.UUCP (David Boyes)
Organization: University of Oregon, Computer Science, Eugene OR
Lines: 139
In article <9830@g.ms.uky.edu> david@ms.uky.edu (David Herron -- One of the vertebrae) writes:
>In article <403@cseg.uucp> lag@cseg.uucp (L. Adrian Griffis) writes:
>>[discussion of BITNET maps]
There are lineprinter maps and a SAS/GRAPH program available that
draws a very nice map available from LISTSERV@BITNIC.BITNET. They used
to be available from NETSERV, but they've finally scrapped that dog
and gone to LISTSERV. Send a mail message to LISTSERV@BITNIC.BITNET
containing the word INDEX for a list of available files.
>There's another table that's more interesting. But it's very large
>and I don't know of any documentation describing its internal format.
It's a very simple format. Entries consist of:
:.
where tagname is the field name and value is the obvious. Entries
start with :nick. or :node. entries. The other tags are defined in the
registration documentation you filled out when your node joined
BITNET. See the description of the NAMEFIND command in the VM/SP CMS
Command and Macro Reference Guide for a more detailed description.
>Not that it's very hard to read anyway, but writing a program to
>decode things out of it would be slightly hard. Anyway, it's got
>a name like BITEARN NAMES or some such, and is available through
>your country-wide NETSERVer. (For the USA, it's the one at BITNIC).
Here's a fragment of REXX code that pulls stuff out of that file:
/* we now have all of the information available on the given node in */
/* the string nodeinfo. We use GETTAG to get the value of a specific */
/* tag from this long string of stuff. If the tag does not exist, */
/* we return a null result. */
call GETTAG nodeinfo, ':node.'
node_id = result
call GETTAG nodeinfo, ':site.'
site_id = result
call GETTAG nodeinfo, ':system.'
system_id = result
call GETTAG nodeinfo, ':machine.'
machine_id = result
call GETTAG nodeinfo, ':bitdirector.'
bitdir_id = result
call GETTAG nodeinfo, ':addr.'
address = result
/* display results of the search in display panel */
'CLEAR'
say 'The information you requested for site' node_id 'is: '
say 'Node name :' node_id
say 'Full name of site :' site_id
say 'Site Address :' address
say 'Operating System :' system_id
say 'Machine Type :' machine_id
if bitdir_id = '' then do
bitdir_id = 'No Bitnet director defined for this machine.'
end
say 'BITNET Director :' bitdir_id
end
else do
say 'Node' what_to_find 'was not found in the nodes file. '
end
/* The following routine is used to seperate a value from the */
/* origin string and remove tag identifiers. Handles multiword */
/* tag values... */
GETTAG:
procedure
string = arg(1)
search_pattern = arg(2)
start_position = index(string,search_pattern,1)
if start_position = 0 then do
return ''
end
temp = start_position + 1
next_tag = index(string,':',temp)
if next_tag = 0 then do
next_tag = length(string) + 1
end
start = start_position + length(search_pattern)
size = ((next_tag - 1) - start)
result = substr(string, start, size)
return result
Relatively inefficient, but I wrote it in about 2.5 minutes. Fairly
straightforward.
>>[discussion of POSTMAST id]
>THis is almost correct. It's not "normally" installed. Again, there
>is a tag in that more interesting table which says who to send postmaster
>type mail to.
Not quite. IF the site uses the Crosswell MAILER (like most VM BITNET
sites), then the POSTMAST ID is defined in the distributed MAILER
PROFILE configuration file. The line is in the specials: section at
the bottom of the file and can be directed to ANY id -- MAILER handles
queries to POSTMAST and BITINFO by looking at those entries first.
Also, at most sites, POSTMAST is a human being; VMS and other OS types
often assign this id to automated software, causing even more confusion.
>You have to be able to look it up in that table in order
>to know who to send mail to. *THERE*ISN'T*A*STANDARD*NAME*TO*SEND*TO*!
Cf. above. Also, you tHe user don't need access to MAILER PROFILE. All
you really Need to know is which DVM to send outgoing mail to and what
types of headers it expects. It's the mailer's business to get it there.
>Well, nor is there a standard format for mail either. What a wonderful
>network they have :-).
Don't throw too many stones. Usenet isn't all that great either.
There are 2 formats for BITNET mail:
a) IBM NOTE format. IBM NOTEs are always sent class A, Netdata
format. Does not allow domain addresses; simple
user@node form in headers. Commonly used by non-IBM
systems as least common denominator.
b) MAIL format. Almost always class M, Punch format, filename
of sending userid, filetype of MAIL. 80 byte card
images.
You can figure out which to use by scanning the file XMAILER NAMES; if
you find the destination host there or the destination domain in
DOMAIN names, use format B and send to the address in the tables.
Otherwise, use format A. MAILER already does this for you, though.
>David Herron -- The E-Mail guy
--
David Boyes | Internet: dboyes@drizzle.cs.uoregon.edu | (503) 686-4394
| BITNET: dboyes@uoregon |
DECnet mail addresses -- just say no.