Hi Leonard,

 >  MvdV> I sat that because you are wrong. One can combine
 >  MvdV> two B channels to achieve 128 kb/s.

 > But that requires making *two* simultaneous calls. To
 > different numbers, which require seperate nodelist entries.

No, it doesn't work like that. For the European ISDN system that is. Maybe
the American system is different in this respect.

It certainly does NOT require two different numbers and separate nodelist
entries. Whether or not a separate call is involved is a matter of
definition. I'd say it is not because the second B channel can be allocated
dynamically during the call. One can set up the call for one B channel and
64K and later switch on a second B channel to get get 128K. One can also
drop it again without breaking the connection when one does not need the
full bandwidth any more.

The usefulness for Fido is limited because the price rate also doubles when
the second B channel is seized.

 > You can do the exact same thing with an obscure brand of
 > modems.

It is not the same. It does not require extra hardware, one interface card
will do it, and it does not require a second pair of wires.

Perhaps it helps if you look at an ISDN line as a 128kB channel that can
either be used in full or split up into two 64k channels that can be used
for separate connections.

Why am I making a point of this? To show that an upper limit must be
specified, otherwise you will run into the the problem demmonstrared: a
programmer not aware of possibilities elsewhere, will assume that xxx is
enough and reserve so many bytes for the value and the software will break
if someone else assumes someting different.

 >  MvdV> How about IP connections? Speeds of up to 100Mb/s
 >  Mvdv> are conceivable.
 >
 > Frankly, the "properties" of IP links are such that if we
 > are going to have them in the same nodelist as analog systems

We have already dropped the concept that the nodelist is for analog systems only.

 > we really need to revise things so that the formats are
 > *different* for the two types of systems, rather than trying
 > to kludge them into the same format.

In that case FTS-5000 should clearly state that the 'baudrate' only applies
the PSTN part.

 > Right now we are in a sort of messed up interim.

Oh, it's a hornet's nest that has been dug up. No use denying it.

Cheers, Michiel

--- InterMail 2.29k