CONNECTION BLUES
Despite Deepak's remarks on the subject, I'm not really 'Jack the Modem
Killer' though it must have seemed that way recently. I did go through four of them in just under a month.
We've had true functionality problems with our dialup access since we
instituted a dedicated line 5 years ago. Our line would not support a
connection faster than 28k and, though the connection was stable, it simply wasn't fast enough to do us right. Years of fruitless calls to the phone company elicited many, many tests of the line and much sympathy from the various SWB representatives with whom we had contact (they were uniformly nice and cooperative) but nothing more.
Through this time I developed a reputation as the most patient of men. This due to the fact that I was doing downloads taking as long as a week at 28k. I saw no alternative, so I simply let the machine churn when I needed a big download.
Finally, Deepak offered to become involved. I have no idea what passed
between him and SWB, but we got a new line, though previously this had been declared impossible by the phone company, due to lack of unused circuits.
Speed went up; often as high as 40k. So, of course, some contractor
unrelated to the phone company promptly cut the line. As I was online when this happened, our trusty Zoom #2949L external serial modem bit the big one and will now no longer connect at more than about 14k. No problem, right? Toss in a new modem and go. That's exactly what I did.
We're somewhat poor, right now, so I used a Pine/ESS PCI unit I had on hand. The dreaded 'winmodem'. Ahh, but it worked. The line repair had evidently not affected things badly as we were still getting on at fine speeds. The connection was not as stable, though. And it soon became apparent the modem was at fault. Dumping the provided drivers, I obtained a set of generic drivers from the ESS site and stability improved, somewhat, though not as much as I wished.
Friday last, the modem started really dying. Connection speeds went way down and disconnects became hourly occurrances. My frustration knew no bounds. And, being as our local wholesalers were closed for the holiday and it was the end of the month, meaning we were anxiously awaiting our next checks, I couldn't just dash out and get a new modem. That's why I missed the show, last Sunday. Couldn't even stay connected long enough to partake in the chat, as in the early part of the show, the modem died completely.
Checks have come now. We have a new modem and a backup. Both are proven to work just fine. I'm a happy camper again. Apologies to my wife for a week of unrelenting grumpiness continue apace and will for some time.
Because of all the folderol, this last week, I've gotten to thinking about how reliable our computer equipment has been. We are not up to date, here on the Ottershouse LAN. The computer upon which I type this is almost 10 years old and hasn't really ever missed a beat (my trusty Debian GNU/Linux 486 which is our archive server and which we run Seti on, also). My Windows 2000 Pro box (upon which I do the most work and spend the most time) is over 5 years old. Though it has gotten many updates, it's still only a 500 MHz machine. The mainboard is the original one, and still churning away, despite years of running overclocked. Scanners, printers and other peripherals have given us the most problems; not the computers themselves.
Still, when I consider the mind-boggling complexity of the computers
we use, I am astounded at their reliability. So reliable have they become that we are able to take it for granted and so become especially nonplussed when they do fail. I find that amazing.
SCO LAWSUIT AGAINST IBM
No, I'm not going to expound on the lawsuit, much less comment on who may be in the right or in the wrong. I just wish to say that it's the best soap-opera available since the Watergate Congressional hearings. The claims and counter-claims, charges and counter-charges started out being extreme and as they get more complex and grandiose, it's a gas to watch. I suspect the outcome, whatever it ends up being, will be anti-climax to the run-up, but it could be momentous, instead. Meanwhile, it's comedy and suspense. Something to distract me from the day to day.
OPEN SOURCE MACHINE DEVELOPMENT?
By now, everyone knows the values (and hazards, for that matter) of the Open Source development model for software. I have begun to wonder whether such an approach would work for a mechanical device. I guess I had best explain.
I use a wheelchair, powered by my arms. That is the mobility device which suits me best. My wife, on the other hand, needs a powered scooter of some type, which she doesn't have and which we cannot afford. My experiences in the field of equipping vehicles for disabled folks and selling mobility related devices has led me to some serious conclusions about such things.
Wheelchairs, scooters, lifts and devices like that are grossly over-priced. This is because most of them are paid for by the government, through Medicare, the Veteran's Administration and similar programs. The 'wisdom' of the various legislatures involved over the years dictates there be NO price competition on these items. The programs involved will simply pay up, regardless of cost. A vendor of such items and related services is unlikely to be reined in on matters of price, barring fraud or something really gross. So, they charge a lot more than the goods and services warrant. When one approaches a company to purchase such a device or service, they are charged these same high prices, to avoid the appearance of gouging the government.
Now, it's true one can buy a seemingly capable three wheeled scooter for well under $1000 (US), but inspection shows these machines are not really all that good a deal, looking more like shills to drive customers to higher priced models than actually usable and durable appliances. My own close inspection of a few of these convinced me I wouldn't send an enemy riding one, much less someone I loved. (Though I do admit I have not seen them all, by a long shot.)
I was trained as a machine designer, in my youth. But I lack the necessary design skills to do up a safe, reliable and efficient electric motor control and related systems. It's sweepstakes odds Otter and I will never be able to afford a good scooter for her. So, I got to wondering if an open design and development program over the Internet could come up with a good, serviceable design. If the result were as I envision it, the scooter would be able to be built at home, from off-the-shelf components, with a minimum of welding , machining or other equipment intensive processes. It may even be possible to do it with none of that, as the number of possible components available for retail sale is so huge, a winning combination might be found without 'purpose building' much of anything except for those parts needed to adapt the basic model to someone's specific needs.
The mark of success in such a project would be that a significant number of them would be built. Were the resulting design good enough, perhaps someone would choose to offer kits or assemble them commercially and something like a BSD type license on the design would make this possible.
What do you think? Is it feasible? Want to help? Let me know what you think about this, or anything else I write here or say on the air, please.
Jack
© 2003 Jack Imsdahl
© 2002 - 2004 by On Computers and the Videotex Services Coalition.