Macintosh computers have never been error free, much like any other set of mechanical devices. The most recent problems have had to do with the SuperDrives and with IIcx/ci power supplies. Conscientious people have helped pin down the causes of these problems and posted the fixes on Usenet.
The SuperDrives have had problems from the very beginning, although the furor has died down a bit recently. Part of the problem was that people were formatting disks inappropriately. However, the problems still persist. Arthur Hills of the University of Waterloo took a disciplined approach to the problem and found that different disks often had the same bad blocks (checking with SUM TuneUp), and if more than one bad block was found, the separation of the bad blocks was a multiple of 36 (don’t know the units). Needless to say, this sort of regularity with errors is odd. After long communications with the Apple Technical Response Group in Toronto and Kao Didak, the disk manufacturer, the following became clear.
Apple has shipped two different types of floppy drives, an early version with a green controller board and later version with a blue controller board. The green controller board lacks a phase-lock-loop circuit to regulate the speed of the drive. This circuit is important because it guarantees a specific rotational speed within 1%. Without the phase-lock-loop circuit the rotational speed can vary greatly. If bits are written at different speeds than they are read at later, it appears to the computer that the bit has shifted, which can causes cyclic redundancy check (CRC) errors. A similar problem can occur if you take a disk written in one Mac to another whose drive is faster or slower. The fix is of course to replace the drives with new drives with the blue controller board. So if you have problems, try checking your disks with SUM TuneUp and if you can document the problem, you may be able to get your drive replaced. Of course, buying SUM will cost at least $100, so it may be easier to just get a new drive.
The second problem is a bit stranger. Robin Goldstone reports that some IIcx and IIci machines will sometimes fail to turn on with either of the switches. The probability of experiencing startup problems is higher if the system has been turned off overnight, strangely enough. The simple fix is to unplug the power cord from the Mac for more than 10 seconds and then plug it back in. Apple says that if the Mac doesn’t start up properly after the unplugging trick, the power supply should be replaced. If you are having these problems and you have a power supply in one of the following ranges, you should give your friendly local Apple dealer a call and sound pitiful. According to the posting, the serial number ranges are "GE924xxxxxxx through GE953xxxxxxx AND/OR GE001xxxxxxx through GE023xxxxxxx" [sic]. Of course, there may be others as well.
Apple has fixed the problem ("Oh, yeah, we should’ve put a resistor across capacitor C9 and jumper wire on the control board. Details details.") and all the good power supplies now have a white dot adjacent to the Molex output connector, should you happen to know what a Molex output connector looks like. I wouldn’t if it hit me in the nose. Apple is working on a Customer Satisfaction Program, so if you have this problem but don’t have one of the above serial numbers please call your dealer and ask for satisfaction. This is the sort of thing that Apple should replace because it is mind-numbingly frustrating and an opportunity to win some customer loyalty with good service.