Here's the description of a mistake I did with international characters in OI, and the consequences. With hindsight what I did was a clear mistake. I'm just reporting it here (on both lists) in case I have adviced others to do the same.

One of the things one has to do when one implements international character sets in OI, is to define an international language row (LND_something) in SYSENV. Or usually it's rather a national row - one tailored to the particular language of the data. Lines 7 and 8 of this LND_… row should contain all the upper and lower case letters of the language - paired. The content of lines 7 and 8 is used for the @Upper.Case and @Lower.Case variables, which again are used in various conversions.

But one should be very careful with including delimiters (line 8). The problem is that some processes in OI do case conversions on arrays. So conversions to uppercase may change delimiters into normal characters, e.g @TM=

to Û.

(And sometimes your own programming does the same (i.e, case conversions on arrays). Btw - my doesn't. I've created my own conversion functions and don't use @Upper.Case and @Lower.Case any more.)

When I implemented international character sets and made my language row, I left out only ÿ=@RM (y umlaut) and

View this thread on the Works forum...