CF>V.32terbo as implemented by USR _is_ *EXACTLY* the same as every other
It is the same as far as the AT&T specs go.  Yes.  But USR went beyond
what AT&T's v.32terbo allowed.
CF>it implied that USR invented the protocol (in fact AT&T doesn't have much 
to
CF>go on becaue V.32terbo is itself is a fairly trivial extension of 
32bis).
I can't argue that.  BUT, they do have one significant claim... They
were the ones that made the 'official' specs that went up to 19k.  So
they had a right to name it, etc.  It may not be very creative or
original, but because they did extend it 19k, they had the right to give
it a name.
CF>If V.32terbo as implemented buy USR were not the same as implemented by 
othe
CF>vendors, the modems would not interoperte at 16.8 and 19.2Kbps. If you 
ant
CF>look at an extension of V.32bis that operated at those signalling rates 
that
CF>_is_ incompatable, take a look at the ZyXEL implementation of 16.8 and
CF>19.2Kbps signalling rates.
Again, you are focusing on the 'official' AT&T 19k v.32terbo.  That's
never been at question.  This is about USR's 21k _extension_.
CF> CB> Not at USR's 21k connect rate it isn't.....  And that was the
CF> CB> point.
CF>Nobody else implemented a proprietary extension, but anybody else who
CF>implemented V.32terbo could link at the signalling rates stipulated by the
CF>protocol specification. I fail to see the incompatability that you claim.
Since you are having comprehension difficulties, one last time I'll try
to put it into simpler terms....
AT&T created v.32terbo that could go to a max of 19k.
USR and many others implemented that 'standard' protocol.
USR created their own extension that could connect at 21k.
USR still called that 21k connection "v.32terbo", just like all the
other 19k v.32terbo implementations everybody else was doing.
But, the v.32terbo protocol, as *defined* by AT&T (the 'creators' of
it), could _only_ go as high as 19k.  No higher.  Under no conditions.
Therefor, that 21k connection could not be "v.32terbo".  It can be
called an extension to v.32terbo (such as v.32terbo+, or "USR's 21k
extension to v.32terbo"), but it can not be called v.32terbo because
the v.32terbo does not allow connections at that rate.
If you took the current v.34 33.6k protocol, and you somehow magically
extended it to work at 44k, even if it's compatable at 33k and below,
the higher speed would not be v.34 because the v.34 has a limit of
33.6k. Nothing higher than that could be considered the actual v.34.
The most you can say is that it's an extension.
--- QScan/PCB v1.19b / 01-0162
---------------