USB Connectors Are Lousy
I finished Chapter 6 yesterday, and have only one more chapter to write. When I was exporting my Visio drawing files to .png for submission, I discovered that a figure I drew two weeks ago had gone bad. It went bad in a way I've seen before: It was smaller than similar figures...and it crashed Visio. That happened to another Visio drawing file of mine a couple of years ago, when the front-panel USB port on my Antec 900 case went intermittent. I'm guessing that the data connection opened up while the file was being written, and not everything made it from the Windows filesystem to the Cruzer Micro Skin on which the file was stored. This happened to another couple of files, most of them Word documents. In those cases, I found out before I did my weekly backups. This time I didn't, and the bad file had already propagated to my backups and overwritten the intact copy of the file. I have a hardcopy and can probably redraw it in an hour or so. Mercifully, it's not anywhere near as complex as some of the 79 other figures I've drawn for the book in Visio so far. But man, I don't want to go this way again. The Thermaltake V9 BlacX case has worked very well for me since I first described it in this space. I don't think it's the case's fault, nor the fault of the Antec 900 before it.
It's the USB ports. And I don't know quite what to do about it.
I've had a personal computer since 1979, and have trudged my way through generations of removable storage since then. Here's the list:
1979: 8" floppies
1982: 5 1/4" floppies
1985: Original "cafeteria tray" Bernoulli box. (Tick...tick...tick...)
1991: Original Syquest 44MB cartridge hard drive
1995: SyQuest EZ-Drive cartridge hard drive.
1996: Zip 100
1997: Zip 250
1997: Jaz. Got rid of it almost immediately, went back to Zip 100s and 250s.
2004: Thumb drives
Trouble is, I don't know what, if anything, comes next. The problem isn't with thumb drives themselves, nor with the Flash technology inside them. Granted, MLC makes me nervous, and I'd pay extra for SLC thumb drives if they still exist. (I don't think they do.) The problem is entirely with the crappy connectors on the front panels of modern PCs, and with how badly we've abused them.
The USB interface wasn't designed with removable storage in mind. USB was intended for interfacing to things like printers and scanners, for which you plug a cable into the port on the PC and leave it there. IBM invented the USB thumb drive in 2000, and I bought my first one in 2001. They work beautifully, but plugging and unplugging thumb drives all day can't be good for the thin shim stock from which the port's electrical interface is formed. It may also put stress on the copper traces on the circuit board behind the port socket. However it happens, the damned things die, and when they die, they take files (and cases) with them.
We have to be able to do better than this. The obvious solution is to recess the ports behind a channel that aligns the USB plug and keeps it from putting angular force on the port--i.e., a wiggle suppressor. Alas, there's no standard for the bodies of USB plugs, whether for printers or thumb drives. A recessed port would keep all those rubber duckie-shaped thumb drives from plugging in, and who knows what Federal intervention that would trigger.
There have actually been eSATA thumb drives for several years now, but because eSATA ports don't provide power (big mistake!) you have to run a cable from the thumb drive into a nearby USB port. Furthermore, even if there were eSATA ports on the front panels of tower cases (I have yet to see one) I'm not sure that the physical port would fare much better than a USB port with constant plugging and unplugging over a period of years. The eSATA spec didn't anticipate that either. Both USB and eSATA ports seem to be made of the same flimsy stuff.
Carol made a suggestion that is functional but not elegant: Find a short USB adapter cable with a male Type A on one end and a female Type A on the other end. Leave the male plugged into the desktop, and swap out thumb drives on the female end of the cable. I actually found a 40" specimen in the bottom of my cable snake pit, and it works. Shorter cables of that species may be available. Needless to say, it looks like hell. Cables are cheaper than cases, though.
Now, a clever hardware hacker could build the female end of the cable into some sort of base that sits beside the tower and plugs into the back. I could see that. Can see that. Will see that.
When I do, you'll see it here.