Subject: Engine failure on the InfoBahn
Date: August 20, 1994 09:48

---

8:00 Toowong, Australia :: 21 JUL 94

You all may be wondering why I've been out of touch for so long. Well, Murphy the Optimist discovered many laws governing everyday reality, the primary axiom being "Anything that can go wrong will." In my close encounter with Murphy, the PCMCIA II high speed modem in the Toshiba partially broke down sometime after the first day of my trip into the Australian bush. It failed to acquire a dial tone from the telephone line and wouldn't dial out (even with an X3 in the init string). In all other manners it appeared to operate flawlessly, even responding to all known AT modem commands.

Since I'd been experiencing various difficulties connecting on some phone systems already, and seen a friend's modem fail to connect in much the same manner, I couldn't be absolutely certain the lack of a dial tone didn't originate as an incompatibility with the local telecommunications system which In some remote areas remains somewhat backwards. Only three years ago an operator manually switched lines for Canaway Downs (an outback sheep station I recently visited and about which you'll hear more) and neighbouring properties. Remember those ancient images from Pathe newsreels of telephone switchboards?

The breakdown occurred on my first day of travel into some of the most remote country in the industrialized world. For the next two days, believing the local phone systems to be at fault, I got further and further away from technological salvation. I was never certain whether the modem itself was at fault until this week back in Brisbane where a service tech confirmed my worst fears.

Given these experiences, I here define

Jennings' Law of the Technological Frontier

The Mean Time to Failure for any portable technology is defined by the following relationship:

Where M is Mean Time to Failure measured in years; D represents the hours required to transport the device to a reliable service technician using any available transportation; R is the probability of systematic interface failure between the portable device and any local technology to which it must be connected.

The formula M = 1/D², a classic Murphyism, governs situations such as your car running out of gas when the distance to the nearest filling station is maximized. This relationship does not adequately define the conditions of portable technology. For example, the likelihood that your PCMCIA modem will break down increases with distance from civilization and decreases with local telecom system reliability. In this new formula the variable 'R' accounts for two additional factors.

1. The probability that the portable technology's failure mode will be indistinguishable from a system interface failure.
2. The probability that the interface itself will break the portable technology.

The second of these represents not a Murphyism but a very real problem described by systems theory. For example, plugging your 110v appliance into a 220v Australian electrical socket will likely result in a defunct appliance. The failure occurs at the interface, breaking the device.

On the other hand, the first factor is pure Murphy. If your modem fails to connect via the remote microwave station link 3 days into the Australian bush is it due to an interface problem, or is the modem broken? Remoteness limits your troubleshooting options; until you connect it to a system known to be reliable you can't be sure whether the modem has broken or simply failed to interface properly with the foreign system. Or as Murphy would say, the more difficult it is to determine whether the device is broken, the more likely the device will appear to fail.

The Murphy gotcha here: "What do you do?" Do you go to the expense of shipping the modem to a service center only to find that the modem works fine but couldn't connect properly to the local interface? Or do you continue traipsing about the Outback hoping to find an oasis of reliability in Australian Telecom's rural/bush technology when, in fact, the modem is broken?

Neither of these options appealed greatly to me, but being, like Murphy, an optimist, I took the second.

11:51 Kingscliff, Australia :: 1 AUG 94

Well, I'd have been better off to be pessimistic.

In addition to learning that the modem was indeed broken, I also discovered how expensive communications technology can be here. A PCMCIA Type II 14.4 Fax Modem runs in the neighbourhood of $700 Australian, about $720 CDN. The same device can be had in Canada for less than $300 CDN.

Had the cost been lower I'd have simply replaced the modem. Once reconnected, pursuing the Vancouver retailer of the broken modem for satisfaction could easily and inexpensively have been achieved through email. Instead, several lengthy phone calls to Vancouver were necessary and I have been required to pay the costs of FedExing the defunct modem to North America and shipping the replacement back to Oz. Further, no replacement would be shipped until the broken modem had been received and the warranty verified by the manufacturer.

That process began two weeks ago and I have not yet received the replacement. <sigh>

1:17 Toowong, Queensland :: 20 AUG 94

Replacing the modem proved to be where Murphy really went to town. It would take 4 weeks + a day for the replacement to arrive. This delay was due primarily to a mixture of tragedy and incompetence in equal parts at the Canadian retailer. After assuring me the replacement would be shipped in the afternoon, it sat on Rene's desk for one week while he recovered from his lunchtime car accident. His coworkers didn't even apologize a week later when I called. Initially, they insisted it had been sent! An apologetic Rene was back at work a week after that. "It's on it's way!" he assured me. "When was it sent?" I asked. "Ummm. This morning." was the reply.

At least he apologized.

When a courier assures you a package crossing the Equator and both tropics will arrive "Noon Friday", assume Monday morning. That's how it happened both there and back. I sat around the apartment on three Fridays and three Mondays before it arrived on the third Monday at 8:45. The knocks on the door didn't wake me up so the modem wasn't actually delivered until Tuesday, at Noon, even though I was up at 8AM waiting for it.

So it was Tuesday, August 16 that I finally inserted the long-awaited device into the PCMCIA slot, booted the computer, started my mail client . . . and failed to connect.

WHAT????

Small problem. PCMCIA modems, of course, require software configuration and that software comes on a setup disk. No such disk came in the envelope. I know where mine is-the one that came with the defunct modem-in a box at the bottom of a pile of boxes in a storage locker. No one distributes Megahertz modems or setup disks in Australia and the technical service numbers in North America are useless here because Australian Telecom doesn't generate the proper tones to activate their touch-tone menus and the lines offer no fall through case to get a human receptionist.

Sigh.

Salvation came in the form of a 1200 baud external modem my brother-in-law borrowed from work. It only took about 2 hours to download the setup disk at about $16 US an hour ($8/hr to connect to CompuServe via Australia's FALNET and another $8/hr, I think, for downloading software from Compuserve's Megahertz forum.)

Finally, IT WORKS!!!!

I learned a few lessons from this ordeal:

1. Prevention: I'm trying to track down a telephone line filter that will hopefully reduce line spikes and foreign signal abnormalities that could damage the modem.
2. Modem Replacement:

   1. If you really, really must stay connected, bite the bullet and purchase a new modem locally.
   2. Beg, borrow or rent a temporary modem replacement.
   3. Order one from home: the cheaper prices will offset the shipping cost
   4. Wait until you get home.

   Don't be tempted by the carrot dangling in your face 'cause Murphy's holding the stick. Even a used low-baud external can keep you operational and these can often be gotten cheaply with some legwork. Should you elect to have a replacement shipped out and you're more than a few time zones away from all your suppliers, you should probably consider it very important to stay connected; beg, borrow or rent a temporary replacement. eMail and FAX are much more cost-effective methods for communicating with the people 'handling' your problems than telephone, particularly if you're forced to make those calls in the middle of the night. Save the phone calls for when you're really steamed.

3. Dealing With Suppliers: The reliability of industry people seems inversely proportional to your ability to walk in their door and hassle them. Follow up on every promise rather than waiting a week for your dreams to come true. And if something feels wrong, or someone tells you they 'think' something's going to happen, press the agent until they've verified their beliefs. Essentially, treat them as if you're the technical service rep and they're the dumb user, be courteous but firm and make no assumptions as to the depth and breadth of their knowledge.

Patrick. -- Responses Sought --