[comp.mail.uucp] Info on grade and seqno in Shere/Shostname startup sequence

jeh@simpact.com (09/23/89)

In Nutshell Books' _Managing uucp and Usenet_, in the description of the systems
file, it says that BSD 4.3 allows a _grade_ subfield within the _schedule_ 
field.  

How is this communicated to the system being called?  My guess is that it
shows up as another option (along with - Qseq and - xdebuglevel) in the 
master's Shostname response to the slave's Shere=hostname message -- true?
Does anybody know the exact format?  

Presumably, a slave not running BSD 4.3 simply ignores this option?  

Also, speaking of Qseq:  When, EXACTLY, in the startup sequence is the
sequence number incremented?  The Nutshell book says it's for every call
*attempt*, but this seems unlikely; if I call a neighbor and get a busy signal,
and subsequently call and get through, it doesn't seem reasonable that I should
have bumped my seq number for the neighbor, as the neighbor won't know that
I called and won't have bumped theirs.  It must be done when some level of
handshaking is done -- but when?  When the slave sees 
<DLE>Shostname - Qseq<NUL> , and the master sees ROK?  Or when a protocol
is agreed upon?  Or when the protocol is started?  Or when the call terminates
normally?  

	--- Jamie Hanrahan, Simpact Associates, San Diego CA
Chair, VMSnet [DECUS uucp] and Internals Working Groups, DECUS VAX Systems SIG 
Internet:  jeh@simpact.com, or if that fails, jeh@crash.cts.com
Uucp:  ...{crash,scubed,decwrl}!simpact!jeh

budd@bu-cs.BU.EDU (Philip Budne) (09/28/89)

In article <677.251ac6ac@simpact.com> jeh@simpact.com writes:
>How is this communicated to the system being called?  My guess is that it
>shows up as another option (along with - Qseq and - xdebuglevel) in the 
>master's Shostname response to the slave's Shere=hostname message -- true?
>Does anybody know the exact format?  

Yes.  Experimentation shows that 4.3 UUCP puts TWO options on the
Shere= line;

	-pX
and
	-vgrade=X

I think the lore is that -vgrade= is (largely) historical.

	Phil Budne, Boston University

honey@terminator.cc.umich.edu (Peter Honeyman) (10/01/89)

Phil Budne writes:
>I think the lore is that -vgrade= is (largely) historical.

honey danber uses -vgrade= --  rick was being nice to at&t.  (at&t did
not return the favor.)

using the private sequence number is not recommended -- it has never
been tested ...

	peter