RS> you were doing the SAME thing, mindlessly binning useful evidence
RS> which was all that was available and proclaiming that there was
RS> just one valid way to consider the issue being considered, using
RS> some silly idea about the only valid way to test possibilitys.

RS> Soorree, the real world aint like that Paul, at times you have
RS> to use the evidence thats less than ideal, coz its all you've
RS> fucking got. And whatever you proclaim, its quite feasible to
RS> get useful information from the less than ideal evidence.

RS> That is possible right now on the question of the quality of your line.
RS> We have the Courier/Courier stats when your line was called by Russ from
RS> FNQ from that immaculate line. We have the comparison with the data on the
RS> other calls he made to a Courier, and can deduce some useful information
RS> on the state of your line compared with say the state of Bills line.

RS> You can also check the Courier/Courier stats by
RS> making sure they do correlate with CPS figures so
RS> there cant be anything do drastic wrong with them.

RS> You can proclaim that the one valid way to measure lines is to hire
RS> test equipment and use it on the lines being looked at if you like,
RS> pity the real world is that that that AINT the only way. Tho it would
RS> certainly be nice to do that if the cost wasnt a consideration.

PE> I didn't say that.

RS> You did basically, proclaiming that the USR/USR stats are completely
RS> useless and of no value whatever. Which essentially means that that
RS> other approach with specialised test equipment is the only thing you
RS> think is a viable way to measure a specific line.

PE> In fact, the M34F may well be a way of telling
PE> line quality, that has yet to be determined.

RS> Which essentially means that you are currently proclaiming that
RS> there aint no way to characterise a line except test equipment.

PE> Not at all.

Fraid so, if you just bin USR/USR figures because there is a rather obvious
discrepancy with USR/nonUSR figures, there aint no way to decide if the
M34F/M34F figures are any more meaningful than the USR/USR figures. without
using some test equipment on a particular line to see what it actually is,
and then comparing that data with the M34F/M34F stats and the USR/USR stats
to see which gets closest to what the test equipment found.

PE> If the results of the M34F are consistent, it could well be a good tool.

Presumably you mean if the M34F figures dont change depending on
what modem is on the other end of the line. Yes, that approach has
some theoretical validity, BUT we ALREADY know that the Rx level
the M34F claims to have seen with one particular AT command is dud
coz it says 50dbm all the time. Clearly the M34F has problems too.

PE> You would normally expect a tool to do what it is designed to do.

Yes, and when you can see that it claims an RX level of 50dbm
ALWAYS, when it cant possibly actually be always that value,
we do in fact KNOW that we have seen a problem which is even
worse than we have seen with th USR/nonUSR figures.

In those circumstances, when the USR/USR figures do make considerable
sense when you can see how they vary with KNOWN bad lines, they may
well be more useful if you cant justify the expense of the test equipment.
@EOT:

---