IMHO: The slightly slower connect speed appears to be a trade off
in recognizing that not all modems and residential phone lines are
created equal. I'd have to give the Livingston people who worked
on the new modem code an "A" for effort. But, Livingston's effort
is beginning to look like the same ole song. . . Livingston just
can't please all the modems all the time, and neither can any
other vendor.
The dilemma as I see it is that many clueless users much prefer to
see their MickySoft connect speed of 50999 and they could care less
that their modem is retraining every 20 seconds as opposed to a more
stable 45666 connection with little or no retrains.
Out here in GTE & Pac Bell land using ICG provided CT1's, I've had the
best luck so far with running ComOS 3.8, and plan to be backing down
to it once again.
Just my two cents. . .
T.J.
[ No Title, just a low life owner/operator ]
At 03:14 PM 5/11/99 -0500, you wrote:
>"Jason A. Cox" <tech@alpha.familynet.net> wrote:
>>I had one user quit so far because of this issue. I hope that isn't a sign
>>of things to come. Most users tolerate slow connections for a while
>>(suspecting it is their phone line noise, etc. due to the Oklahoma weather),
>>but they are going to figure it out sooner or later!
>>
>>After my first complaint, I scanned the details log (radius) and noticed
>>that several other users that where connecting 49K before are now ending up
>>(stopping) at 31K. Some others are never getting their 33.6K++ connection
>>that they once had.
>>
>>The flip side is that many of the "can't connect" or "disconnecting all the
>>time" users are now happy. I just hope the fix doesn't run off all the
>>rest. =)
>>
>>How can we diagnose this problem and report it to Lucent?
>>
>>Jason A. Cox
>>Family Net, Inc. - jason@familynet.net
>>(918) 663-7134 - FAX (918) 663-7136
>>7935 East 57th Street - Tulsa, OK 74145
>>
>
>We've seen just the opposite - overall higher connection rates vs 3.8.2
>(and we're in your neck of the woods). We participated in the PWM betas
>(of which the result was 3.8.2c2). We worked very closely with Lucent for
>months on all sorts of modem issues - USR 56K v.90 null speedshifts,
>Rockwell HCF madness, Diamond Supra Express KFlex disconnects, Global
>Village no-connect, Motorola SM56 issues, etc.. my guess is what you (and
>the others) might be seeing is that with 3.8.2, while they may connect at
>v.90, they are reneg/retraining down over time, and before they might have
>been disconnected, whereas now they correctly retrain down to v.34 speeds
>rather than disconnecting. Do you log both initial connect speed and
>connect speed on the accounting records as well? This can provide some
>insight. If you are having customers with specific modem issues, the best
>way to track it down is to identify the make/model modem, firmware
>revision, and ensure the problem *is reproducable*, then E-mail or call
>RABU support. They can log in to one of your boxes and run traces to get
>the data they need if the customer is willing to dial in for diagnosis.
>Also, make sure the customer is running the latest firmware on their modem,
>yada yada.
>
>
>---------------------------------------------
>Mike Bacher / mikeb@webzone.net
>Systems Administrator, The WebZone
>Now serving both Tulsa and Oklahoma City!
>Phone: (918) 585-8800 FAX: (918) 587-1526
>Web: http://www.webzone.net
>Supporting the v.90 56k standard
>---------------------------------------------
>-
>To unsubscribe, email 'majordomo@livingston.com' with
>'unsubscribe portmaster-users' in the body of the message.
>Searchable list archive: <URL:http://www.livingston.com/Tech/archive/>
>
>
-
To unsubscribe, email 'majordomo@livingston.com' with
'unsubscribe portmaster-users' in the body of the message.
Searchable list archive: <URL:http://www.livingston.com/Tech/archive/>