Monday, October 17, 2005

iTunes needs better error reporting

The last two releases of iTunes (and weren't they rapid fire!) seemed to have a problem with my iPod. And I write seemed because it's difficult to tell: all iTunes would tell me was that it couldn't sync my iPod because “An unknown error occurred (–50)”. (This is iTunes on Windows XP. I use iTunes on Windows mostly for historical reasons—I had ripped my CD collection to that machine before I had a Mac.) With the last iTunes in the 5.x line, it would inform me of the error, and then carry on regardless (which is interesting behaviour), so I just ignored it. After updating to 6.0, evidently error number –50 became more serious, as it wouldn't sync at all. So I was forced to do something about it.

I checked the software on the iPod itself—it was the latest version. Being in Windows-mode, where the standard second-line solution to every problem is re-installation, I figured I'd just wipe the iPod and re-install the music from the PC. It worked, but it wasn't very satisfying.

Really bad error reporting seems to be a standard Apple software trait. Why is that?

iPod: Survives fall from 1 metre

Last night I tried very hard to destroy another iPod. Trying to slip it into its very snug neoprene cover while walking (I really shouldn't multitask), I dropped it from a height of about a metre. The impact might have been less severe if it had hit the carpet on which I was standing, but, of course, there was an object on the floor in the middle of the drop-zone: an LCD monitor I had been meaning to move for days. So, the iPod collided with the plastic base of the monitor. It didn't seem like a critical impact, but the iPod was unresponsive when I picked it up. I sighed, several times, while envisaging another trip to the local Apple Centre, and wondering about the insurance implications of destroying something that was a replacement for something I destroyed previously. However, after a web search to remind me how to perform a hard reset, I managed to return it to life. Nothing broken, nothing lost.

Thursday, October 6, 2005

USB mass storage data loss

I have an Apacer Handy Steno HT202 256MB USB flash storage device. (Am I the only person who refuses to call these things "flash drives"?) It had on it a tar archive from which I wanted to extract some data. Observant readers will note that I am using the past tense. Knowing that my Windows XP box would almost certainly be incapable of extracting data from a tar archive, I plugged it into the PowerBook. It came up on the desktop immediately with no fuss. (This, of course, is in stark contrast to the considerable amount of fuss caused by plugging one of these devices into a Windows XP machine. Why would I want a dialog asking me how the device should be dealt with every time I plug it in? Just mount the file system and leave me alone!) I opened a Terminal window and navigated to the device. I asked tar to extract the data.

Now, it was a pretty big archive, probably around 100MB. And being a USB device, it didn't really all happen at once, so I left it running. The PowerBook was not connected to the mains, and, as far as I can see, it went to sleep in the middle of the archive extraction. When I woke the machine back up, the tar process was wedged hard. Control-C didn't work. Control-Z didn't work. kill didn't work. kill -KILL (inexplicably) didn't work. Nothing worked. So I got the Finder to terminate the shell. And as a side effect of something in this chain of events, the USB mass storage device terminated my data. Gone. Well, not quite: MacOS X wouldn't admit to seeing anything in the directory where the archive was, and wouldn't even give a directory listing. Windows XP saw the directory and the archive, but wouldn't read the latter. FreeBSD 5.3 saw everything, but, again, wouldn't read it.

Ultimately this was my fault.
  • The data should have been moved off the USB stick months ago.
  • I should never have been doing the extraction on the device itself.
  • I should have had the PowerBook plugged into the mains if I was doing something important.
However, is it really too much to ask that either MacOS X allows USB data transfer during sleep mode, or it doesn't fall asleep while accessing a mass storage device? Do I have to wear the entirety of the blame here?