I'm hacking up a doc to explain a text file format we're using in a customization project, and it in itself is nearly enough of an argument to use XML. Here goes (and I haven't started the address):

The PoS export file will be an ASCII text file of any reasonable (less than a gigabyte) length. Each personโ€™s entry will appear on a separate line.

Each entry in the PoS export file must begin with double quotes. Then, an optional โ€œ%โ€ may appear as a specialty indicator. A space will appear next, regardless of whether a specialty indicator character was written or not. The entryโ€™s last name will then appear, followed immediately by a comma and that comma followed immediately by the peronโ€™s first name. A space will then appear, followed by the personโ€™s middle initial and a second set of double quotes.

After these double quotes, numeric text representing the personโ€™s age in years will appear followed by another comma. Double quotes will appear, followed by a string that will indicate the personโ€™s status, followed by another pair of double quotes. This area is where the โ€œDISโ€ entry for dismissed may appear. Another comma will appear before address information begins.

[more to come here]

All values will appear as capital letters. Lines/entries must not โ€œwrapโ€ for any reason. Lines will be separated by a carriage return and a line feed.


What a mess! Being able to call each an attribute and not worrying about the order would be very nice.

In other news, I'm still wondering when I get to see some 1.6 GHz G5 benchmarks. Today's the last day of "No payments till Jan" at Apple, but I think I'm still waiting until Panther comes out. I keep having to tell myself that I don't need the 12" Powerbook. That's still a neat machine, and nearly the same price (with Superdrive) as the 1.6. One more proc upgrade over the next few months and it might be fast enough to program with. Right? :^) Daggum Apple and their cool and horribly expensive hardware.