PE> The reliable observer who couldn't even
PE> remember he had seen it more than once?

RS> Peoples recollection of what happened almost THREE YEARS ago can indeed
RS> leave something to be desired. 

It's a significant thing.  And the wording of the claim says a 
hell of a lot too.

RS> Thats why it can be handy to keep the
RS> messages they wrote AT THE TIME and when you do, its STILL the COMPELLING
RS> evidence that the double send problem is in the Spirit, COZ THERE AINT
RS> NO OTHER MODEM PRESENT EXCEPT A SPIRIT. It really is that basic.

RS> No amount of your proclaiming on datascopes will change that.

YES Rod, *IF* that was what Dave Hatch really saw, then YES, there
is a fault in the Spirit.  A fault that occurs so rarely I've never
seen it myself, despite doing Spirit->Spirit calls for 2 years.
Yes, there is indeed a rare bug.  So rare that it isn't even worth
fixing.  The Maestro bug is COMPLETELY different.  It is reproducible
by anyone, happens every time.

PE> Nevermind the fact that it was a vastly different problem anyway,

RS> Crap, its the classic Spirit double send problem, PROVEN by pausing
RS> the sending system to prove it really is the double send problem.

No, it's A double-sending problem.  One of FOUR different double
sending problems seen with the Spirit.  One of which was hardware
and solved a long time ago.  One which occurs EXTREMELY rarely, if
at all, and isn't reproducible so no investigation can be done on
it anyway.  Then we have the MNP problem and the V42 problem.

RS> The ONLY difference is that the frequency is certainly lower
RS> with a pair of spirits than say a Spirit/Maestro, but thats
RS> not surprising at all give that a tiny change in the timing in
RS> the code behind the Spirit makes it go away FOREVER to a Maestro.

It doesn't actually.  Whenever a Maestro connected to my Spirit,
the Spirit flashed like mad.  EVERY time.  Up until the point
that Zmodem started going.  Until then, all the session negotiation
showed up as MR flashing.  EVERY time.  

PE> one which can't be reproduced by anyone except him,

RS> You dont even know whether anyone ever tried to Paul.

I tried for 2 years, Rod.  2 calls a day no less.  Crikey, we
haven't even checked whether that really was a Spirit on the
other end, maybe he just thought it was, because the caller had
told him some time ago that it was, nevermind he had a Maestro
loaner.

Fairly irrelevant though, even if we were SURE that the problem
had occurred on that VERY RARE occasion, it has NOTHING to do
with the problem we are trying to solve, which is completely
different.  As completely different as the hardware MR-flashing
fault.

PE> compared to one that can be reproduced by anyone anytime.

RS> That just shows that the frequency between
RS> a pair of Spirits is much lower. No news.

Shows that they're completely different too.

RS> Its ZERO with a modified Binkley.

No it isn't.  Only in a Zmodem (not Sealink/Xmodem) transfer can
you see the problem CIRCUMVENTED for that file, until you get some
back-traffic.

RS> we DO have COMPELLING evidence that the fault MUST be in
RS> the Spirit BECAUSE there aint no other fucking modem present.

Yes, for TWO of the FOUR double-send problems, we know undoubtedly
that they are the fault of the Spirit.  One because MBE had isolated
the defective hardware, the other because there were (apparently)
only Spirits communicating.  We don't have a clue about the other
two.

PE> Yeah, and if it you could reproduce that problem at WILL,

RS> Thats just one type of fault Paul. Yes, faults which CANT be reproduced
RS> at will are certainly much more difficult to resolve where the fault is,

And one that can be reproduced at will, compared to one that can't,
is a COMPLETELY DIFFERENT PROBLEM.

RS> BUT if you know what you are doing, you can resolve those TOO. And you

And you don't.  If you can't see the difference between a 0% problem
and a 100% problem, can't even notice a TINY little bit of difference
between those two, I don't think I'll be able to explain it.

RS> DONT do that by sweeping the vital evidence under the carpet. The
RS> COMPELLING evidence is that its seen WITH NOTHING BUT SPIRITS PRESENT.
RS> So it cant possibly be in any modem except a Spirit. COMPELLING EVIDENCE.

Yes, compelling evidence for the 0% fault.  Says ABSOLUTELY NOTHING
about the 100% fault.

PE> bearing in mind that the Maestro/Spirit problem
PE> can be reproduced at WILL ANYTIME, 100%.

RS> Proves sweet fuck all about anything much, all that proves is that
RS> the occurrence frequency varys. And goes to ZERO in that collection
RS> of hardware with a modified Binkley behind the Spirit.

No it doesn't.  It occurs every time, even with a modified Binkley.
I can see my MR light flash at the beginning of the session.
Never once seen it on a Spirit calling though.  Which shows that
they're VERY DIFFERENT problems.

RS> All that proves is that the Spirit double send bug is
RS> only visible with a particular combination of timing.

And proves that it's a very different problem.  0 vs 100.

PE> You're trying to compare a symptom that occurs
PE> 100% of the time, to one that occurs 0% of the time,

RS> Nope, I am saying that IF ITS SEEN BETWEEN A PAIR OF SPIRITS, THERE
RS> IS NO OTHER MODEM PRESENT SO IT MUST BE IN THE FUCKING SPIRIT PAUL.

Yes, on that VERY DIFFERENT PROBLEM, it is UNDOUBTEDLY the Spirit
at fault.

RS> Thats clearly beyond your logic capabilitys. Its true ANYWAY.

Not beyond my logic capabilities.  Far beyond yours to see that a
problem that occurs 100%, could well have a different cause to
one that occurs 0% of the time.

PE> and you expect me to make a leap of faith that says
PE> yep, they're the exact same problem, no doubt about it,

RS> Thats right, that pausing of the sender fixing the double send, which
RS> is itself an extremely specific fault symptom, PROVES they are the same.

Proves that it was a double-send problem.  And the frequency
and reproducibility PROVES that they're COMPLETELY DIFFERENT
problems.  Not that you'll ever understand that.  I would love
to see you bet your house on the fact that the Maestro doesn't
violate protocol at all, which is what causes the 100% problem,
and the 0% problem is caused by a fault in the Spirit.

PE> no possibility the the supposed Spirit-Spirit double-send could possibly
PE> be a DIFFERENT, RARE bug in the Spirit, if indeed it ever did happen.

RS> Its theoretically possible that there is ALSO a bug in the Maestro,

As theoretically possible as there is one in the Spirit, in fact.

RS> but its RIGOROUSLY PROVEN that there is one in the Spirit. And even

Oh yes, it is RIGOROUSLY PROVEN that there are TWO that are in the
Spirit.  There's Dave Hatch's mysterious one which isn't even worth
worrying about, and then there's the hardware one, which has already
been fixed.

RS> with a Spirit/Maestro pair, its rather unlikely to be in the Maestro

It's highly likely to be in the Maestro.

RS> when we have PROVEN the Spirit has a double send bug, and a tiny timing

We have PROVEN that IF Dave Hatch did indeed see that, then YES,
we have proven that there is a VERY DIFFERENT bug in the Spirit.
FAR different to the Spirit/Maestro problem.

RS> change in the Bink behind the Spirit eliminates it completely TOO.

No it doesn't.  You probably don't know this because you can't
see what I can see.  When I dial a Maestro board with my Spirit,
all I have to do is send a message to the sysop and watch the
MR light flash like buggery.  NO timing sensitivity there, that's
ME typing, not Binkley.  On the other hand when I dial a Spirit,
it's absolutely perfect.  When I dial another brand, absolutely
perfect too.

PE> Combined with the "I've seen it more than once now that I know what to
PE> look for" completely dodgy claim, you may understand why I don't think
PE> a problem who's symptoms are VASTLY, nay INCREDIBLY VASTLY different
PE> (100% vs 0%) are not conclusive proof by any stretch of the imagination.

RS> You are welcome to proclaim that the earth is flat too Paul. Pity

There is not the evidence for the earth being flat.  There is the
evidence for there being TWO completely different problems, ASSUMING
there was actually the second one seen.

RS> Not relevant, I am saying WE HAVE COMPELLING EVIDENCE THAT THE SPIRIT
RS> HAS A DOUBLE SEND BUG COZ ITS BEEN SEEN BETWEEN A PAIR OF SPIRITS.
       ^^^
       
With "A" being the operative word.  Not "THE", which is the only
double-sending problem worth bothering about.       

PE> And hey Rod, did you forget that I saw double-sending between a pair
PE> of Maestro's?  5 times actually, which is more than Dave Hatch's, so
PE> there's the ultimate "proof" that it is CLEARLY a Maestro bug.

RS> Fraid not, all that proves, if it was proven as carefully
RS> as Dave did on the pausing, that the Maestro has one TOO.

Oh yeah, it was done just as carefully.  Can't remember all the
details though, e.g. what ROM levels both of the Maestro's were
at.  BFN.  Paul.
@EOT:

---