PE> The latter hasn't been proven by a long shot. 

BG> That's your opinion Paul, but I know several BBS operators (running 
BG> Couriers) who would disagree with you.

PE> The ones that haven't done adequate testing.

BG> More complete and utter crap.  Dave was so happy with the way his Courier 
BG> performs on his board that he bought another four for use at his work as 

And Gandhi was so pleased with the taste of his piss that he kept
on drinking it.  So what?  The fact remains that David didn't do
adequate testing, he just tested it on his own pristine lines, just
like you did at CHH.  That's why it was ME that first discovered
the problem with Supra, problem calling Rockwells, problem with ATZ.
And they are three that you have been able to verify yourself.

BG> internet modems, and they've been extremely reliable.  You may have noticed 
BG> that most who buy Couriers turn into raving zealots too.  I wonder why?

For the same reason as I was having a big argument with someone who
bought a $3000 PEP modem, and was calling V32bis the "latest fad",
PEP is heaps better.  When you spend 3 times the price on something
that is at best marginally better, of course you want to defend
yourself from looking the fool you are.

PE> It could well be.  The M34F is a real-world modem. 

BG> So?  Every modem available for public sale is a "real world" modem.

That is correct.

PE> Inability to connect to it is a worry for the owner of both modems.

BG> Sure, but it doesn't mean that they're BOTH fucked, does it?

It means AT LEAST one is fucked, and that they are BOTH *deficient*,
not fucked, *deficient*.

BG> It's true though that the days of everybody's 2400bps modems connecting 
BG> reliably appear to have gone.

PE> I couldn't connect to Paragon's 2400 bps modems reliably, or at
PE> least I didn't get an MNP connection.  Not just from my home line
PE> either.  AND I reported the problem to the sysop, and they said
PE> that others had reported it too.

BG> One bad example says nothing useful about the reliability of V.22bis 
BG> though.

You said "everybody's 2400bps modems connecting reliably".  That's
just a Bill Grimsley wind-out-arse.

BG> However, given that the various protocol specs are set in stone by the 
BG> ITU-T, one would reasonably expect that all modems should be able to 
BG> connect with each other without problems, but this is just not the case.  
BG> Unfortunately, no modem is 100% perfect these days.

PE> ANY of the modems we have available, MAY be 100% conformant to the
PE> spec though.  We have no way of knowing whether all the problems
PE> for (e.g.) the USR are actually the other modem's fault. 

BG> True, but we already know that the first 19 revisions of the Rockwell V.FC 
BG> and V.34 chipsets were buggy as hell, so the odds are that the current one 

Bill, the day when you could say "we know" anything are long gone.
Gone the day you decided to claim that all Rockwell's violated 
the V34 spec because they only did 16-state trellis coding.  Turned
out you just MADE THAT UP.  You didn't have a fucking clue what the
standard said.

BG> isn't perfect either.  What's more, no amount of new controller code will 
BG> fix a hard-coded chipset bug, so if it's dodgy, you're stuck with it.

If I've got such a rare bug that has taken 1 year to show up, I'm
not going to be losing much sleep over that.  I am far more
concerned with the things that AREN'T controller code, and also
AREN'T being fixed, even though I HAVE got flashrom capability for
that.

BG> At least with generic DSP technology, it is possible to fix ANY buggy 
BG> controller or datapump code with a user-downloadable replacement.

Yeah, yeah, heard it all before with the Spirit.  I was even silly
enough to believe it at the time.

PE> However, we can say that all of the modems have deficient code, as, even
PE> if they are 100% conformant, they need to be more than that to make
PE> up for other modem's inadequacies.

BG> Nope, that's a silly comment.  It's not reasonable to suggest that if modem 
BG> A follows the ITU specs to the letter, yet can't reliably connect with 
BG> modem B because it doesn't, that modem A is in some way deficient.  The 

It is reasonable to do exactly that.

BG> whole point of the ITU recommendations is connectivity between different 
BG> brands.  

Yes, that's why modem negotiation failure is an indication of one of
the modems being FUCKED.  However, as the owner of the other modem, I
STILL want to be able to communicate to the modem at fault.  Especially
if I happen to be trying to FREQ a file from that board and it keeps
hanging up and it's an O/S call etc etc.  No thanks, I want a workaround
in my modem, EVEN if it is the other modem at fault.  And no-one has
established that it's the other modem's fault anyway.  BFN.  Paul.
@EOT:

---