PE> Sorry, how have I had my nose rubbed in V34?

RS> Raving on along the lines that any CONNECT/NONE absolutely
RS> must be because one or other of the modems is breaking
RS> 'the specs'. The real world is more complicated than that.

PE> Pity that YOU said exactly the same thing yourself,

Pigs arse I did, and you never have managed to grasp
the difference between very complex standards for
hardware and much simpler standards for stuff like C,

PE> "That's the whole point of the ITU spec".
PE> You've said that on a number of occasions,

I have not however EVER said that if you do get a CONNECT 28800/NONE that
there isnt the slightest doubt that one of the modems must be breaching
'the specs'. I have in fact said that there is considerable evidence that
V.42 has some real problems ITSELF, which is a different issue entirely.

I hadnt realised you were having such a problem understanding
the difference between say a standard for a language like C and
the much more complex problem of a layered set of standards for
something like a modem until I saw what you said in USR_Modems.

With something like C, its quite possible to be completely specific and
unambiguous when writing 'the specs' and then you can rather easily determine
if a particular bit of code or a compiler does it correctly or not.

In the case of the combination of V34 and V42 you have a considerable
problem. Yes, V34 does indeed make it quite clear that there are a set
of acceptable speeds and symbol rates allowed, with particular trellis
patterns etc. But it DOES NOT rigidly specify what line characteristics
must be satisfied for a particular config to be used. AND then you have
the problem that the modems FIRST negotiate that stuff, and THEN proceed
to negotiate the other stuff like the error correction protocol OVER that
session established at the V34 level. 'the specs' DO NOT says that if the
V42 fails to deliver either a LAPM or MNP4 session at the V42 level, that
the modems must redo the negotiation of the speed and symbol rate stuff at
the V34 level, and then try the V42 phase again, and continue to do that
until other than a /NONE connect is achieved. Yes, thats what we want to
see, no 'the specs' dont mandate that behaviour, so when it is not seen,
you cant start screaming that a particularly modem must have flouted some
aspect of 'the specs' and start hunting down the culprit and burn it at the
stake and demand that its manufacturer fix that flouting of 'the specs'

This is the real world of very complex layered hardware standards,
its a HELL of a lot more complicated than a language standard.

RS> Doubt it will ever actually dawn on you that its just a tad
RS> unlikely that you alone in the entire fucking world are right.

PE> That's what happened to the guy that
PE> said the earth revolved around the sun

RS> Yes, but you aint one of those.

PE> Fraid so.

Fraid not, EVERYONE has been telling you that on
your grossly anal performance on 'the specs', even
those who have been involved in writing them.

You havent even grasped the distinction between the rather simpler
world of programming language specs and the real world of hardware ones.
You wont wont be revolutionising the world on standards any time soon.

I've binned your pathetically anal performance which does nothing
more than a rave on doing a 'I said', 'no I didnt', going no where
and making a pathetic spectacle of yourself hysterically screeching.

If you STILL think that you can PC every trivial blemish in what
YOU claim to be 'the specs' like that question of a 3D address in an
otherwise immaculate origin line from a point and not risk getting flushed
out of Fido for being a mindless anal, likely you will never grasp it.
@EOT:

---