Earlier versions suck against just about anything. Earlier today I was helping
out a customer with a new (purchased less than 2 months ago) Compaq with one of
these modems in it with revision 2.1.2.107. It usually took multiple tries to
connect, disconnected frequently, and had awful throughput while it was
connected. Dialing our Max6K did just about as bad (it usually took fewer
tries to connect, but disconnected more often). Dialing into a bank of regular
Sportster 33.6K analog modems was better, but still failed to negotiate about 1
out of 3 tries and had AWFUL throughput. As if the online behavior of this
modem wasn't bad enough, it even created problems when not in use. The computer
would often reboot when the phone rang.
So, I went to the customer's house and downloaded and installed the latest
upgrade from Compaq's web site (2.1.2.134 was the newest available from them).
This removed Compaq's very innovative "reboot-when-the-phone-rings" feature
(Darn, I always wanted to be able to run Scandisk remotely;). Now the modem
connects to the PM3's on the first try, and with reasonable throughput. It
usually negotiates 49,333 or 48K then drops down to 44K or so within the first
few minutes. Seems stable after that. Connects fine to the Sportster 33.6's
as well. But it still fails negotiation 3 out of 4 times with the Max. Yep,
you heard it right. This piece-of-junk modem on a poor phone line gets along
better with the PM3's running 3.8.2 than with a new Max6K. Same modem, same
phone line, PRI's on same exchange. Maybe a newer revision may have done
better with the Max, but it's clear enough that Lucent isn't the only one
having difficulty making modem code that's compatible with a variety of
bad modems. Is it even possible to support all this junk and still adhere to
standards closely enough to allow good modems to work properly?
The decision makers at Compaq (and HP, and Acer, etc.) should be deeply
ashamed of having shipped this garbage to their customers. And unfortunately,
it's us, the ISP's, who are having to bear much of the burden of their bad
business practices by providing technical support to help their customers deal
with the defective products they've been sold. Tech support staffers don't
work for free. And to top it all off, the customers often blame us, and think
that we're just trying to pass the buck if we tell them the truth about the
product that they've been sold. Sure it may be solvable after the fact by a
firmware upgrade, but it was defective when it shipped, and not even solvable
at the time for those machines that shipped a couple of months ago. Some of
these are the same vendors who were shipping PCTel junk before 56K arrived on
the scene, and LT Winmodems back when its modem code was so awful. Now that
the LT is working well, they've abandoned it in favor of something else that
doesn't work. It's almost as if these companies have made it an official
policy to ship defective modems whenever one is available on the market.
-- Peace, William Kitchen -- bill@iglobal.net - 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/>