Rod, at 10:56 on Mar 15 1996, you wrote to Bill Grimsley...

BG> This V.34+ chip is either unbelievably better,

RS> Has anyone ever suggested to you that you do
RS> tend to overstate things just a tad at times ? |-)

Since proved pretty conclusively by replacing all four EPROMS individually,
and making several test calls, with stats.  The differences were quite
graphic too.

BG> or there really was a problem with the previous chip.

RS> Or that the previous one wasnt as good on
RS> indifferent lines with a non USR on the other end.

Yeah, seems now that the latest V.34+ EPROM does live up to its claim of
connecting one or two 2400bps steps higher for a given symbol rate.

BG> This new one stick to the lines like shit to a blanket,

RS> You dont KNOW that yet.

After 40-odd calls to Paul's M34F, ALL of which were absolutely perfect in
every respect, I think I DO know that now.

BG> and I'm now calling just about everywhere with no speed
BG> restrictions at all, and connecting at previously unheard of link
BG> rates. With EC as well. Most importantly, calls to both CHH modems
BG> now connect perfectly, even though the lines are still obviously shit.

RS> Corse they will all go bad now you bragged about it |-)

Nope, and I'll continue to brag about it too.  THAT'S how confident I am. :)

RS> Corse, if USR had had the sense to have a flashrom in the Sportster,
RS> you wouldnt have had to fart around so much to get that benefit too |-)

Oh well, the Courier can't be too far away now (he says with trepidation)...

BG> I'm not going to bother collecting stats any more,

RS> I think you should, its the only hard numbers we have, even tho they
RS> leave a lot to be desired when calling a non USR modem even now.

What's the point?  Neither you nor Paul trust the USR stats anyway, so why
bother using them for comparative purposes when all you two will likely do
is use them in a negative fashion?

BG> but have a look at these before I piss them off (and note that
BG> although line levels haven't improved, the connect sure has).

RS> Yeah, noticed the line levels. Odd.

It's incredible, getting 28800/28800/LAPM connects with line levels on the
wrong side of -40dBm.  Who'd have believed it?

RS> I wouldnt be a bit surprised if the fact that the USRs get
RS> brain fade on the line probe to a non USR modem is the reason
RS> they dont do as well to non USR modems as they do to USR modems.

See what I mean?  And you want me to keep recording these stats?  ROFL!

Regards, Bill

--- Msgedsq/2 3.20