-=> Quoting Edward Hobson to Richard Town <=-

 > hitherto let on, perhaps you'd like to explain where in
 > either V42 or V42bis
 > does it permit V42bis to be switched off when under V42
 > fallback negotiation action?
 EH> V.42bis is not switched off, it is simply not selectable as a protocol
 EH> which is compatible with MNP error correction.

On USRs

 EH> 1. Why don't Rockwell support the MNP/V.42bis combination in all their
 EH> chipsets - presumably this makes only recent V.42 Rockwell chipsets
 EH> ITU-compliant, by your reckoning?

Can't speak for other marques' implimentation of Rockwell basics but my
Zoom V32bis hardware EC/DC external did.  That was 1992

 EH> 2. With your array of commands, can you select a LAP-M connection with
 EH> MNP5 compression - if not, for what reason?

MNP5 is MNP4 with dumb compression.  MNP5's not divisable so as to provide
a seperate compression only method

 EH> 3. Why would Microcom have defined a method for using V.42bis with MNP
 EH> if it already existed in V.42?

Specs do not define methods, only what should be achieved

 EH> 4. Is this still appropriate to USR_MODEM (!?)

Dunno.  It's your posting.  And AFAIK it's not a "business practice"  :-)
 
 EH> That would be the case only if it were part of V.42, which it isn't.

When I see/hear where in spec V42bis gets specifically switched off under
V42 fallback action then I'll give myself a good birch twig thrashing in
the sauna in penetance :)

 EH> Curiously of course, V.42 recommendation does not *imply* any support
 EH> for compression,

V42bis was accepted as the primary means of data compression in January 1990
by the then CCITT

rgdZ
Richard

--- FMail/386 1.02