BG> And who could have said nothing about the EPROM change, but
BG> in the interests of accuracy of testing, told it like it was.

RS> Sure, no one has ever suggested that you are deliberately dishonest
RS> like Bob often is. 

Yeah, I'll give him that.  The zealotry can be carved with a knife,
I could see the blood gushing from his eyes when he posted the 
message about the EPROM fixing the problem.  Zealotry which has
produced some of the wildest claims I've ever seen.  I've heard
reports of wilder claims, e.g. margarine reduces your life 
expectancy by 50 years or somesuch, but I've never witnessed it live.

RS> Clearly tho, when you proclaim on where the fault
RS> lies with a particular undesirable symptom, you are often leaping to
RS> unwarranted conclusions based on an inadequate analysis of the evidence.

Yeah, the conclusion-leaps are just incredibly amazing.  They stand 
out like dogs balls to anyone with more than 3 neurons still firing,
but to Bill, it just reads Zealotry+Zealotry=Conclusion, and wonders
why everyone else is still stuck at the starting post.

BG> Who would have thought that a simple change of ROM code could
BG> have made such a vast difference, especially when the previous
BG> THREE EPROMs were almost identical in performance ?

RS> All that proves is that the deficiency in the code which produced
RS> those handshake failures when the Sportster is used on a line with

Another amazing statement.  A bug will persist in code until found.
If it takes 6 months to find a bug, and they make 4 drops of the
software in 6 months, then 4 EPROMs will show the bug.  God only
knows what goes through Bill's mind when he installs EPROMs.  Just
like a brand new modem, mon!  As independent a test as if I went and
replaced the modem with a Yum Cha Special.

RS> Its actually a graphic demonstration of how careful you have to
RS> be when trying to work out which modem has the problem in field
RS> testing of modems. Its JUST that difficult situation that sorts
RS> out who can do that most difficult fault diagnosis from those
RS> who too readily jump to conclusions based on inadequate evidence.

Yeah, he's never been able to understand how his ENTIRE testing at
CHH and the ENTIRE results of David Drummond's have the potential
of not even finding 1% of the bugs.  *I* don't even find 1% of the
bugs in OS/2.  Not even .1%.  And I've been using it for 3 years or
something.  I've found maybe 5 or so bugs, not even a drop in the
ocean.  That's because *I* don't use OS/2 in as difficult as
circumstances as others.  Crikey, I haven't even got around to 
learning the basics yet.  My wife told me a few months ago about a
program that let you delete a subdirectory and all that was under
it.  I didn't even know it existed.  For a long time I used to do
a zip -0rm xxx if I wanted to delete a whole directory structure.

RS> Who also proclaimed that there must be something fundamentally
RS> fucked with pauls line because he never saw any problems with
RS> V34 modems when testing hundreds of them at CHH, but who ended
RS> up face down in the mud AGAIN, when he discovered he had EVEN
RS> WORSE problems calling CHH himself from his own home line.

BG> Not my fault that my own bad lines weren't obvious
BG> until I started calling others with less-than-perfect
BG> lines, where the problem was compounded.

RS> The point tho is that you didnt have the EVIDENCE that
RS> Pauls line was fucked when you proclaimed that it must be.

He honestly can't even see the point you are making here.  All he
sees is his own zealotry.  His problem is that he doesn't
recognize it.  BFN.  Paul.
@EOT:

---