Dilwyn Jones wrote: > OK, I goofed on that one. > > I assumed (wrongly) that because JM allowed 6 parameters > on a JM ROM and appeared to return a vaguely right number I assumed it > worked. Except it just "pretends" to work and of course you can't tell > accurately from an E number where in the year it was until you do > detailed checks with storing DATE result then using DATE$ to see what > it makes of it - no parameter checks on JM it would seem. Using DATE > with 6 parameters on a JM ROM returns the same result as DATE with no > parameters. Or with any number of parameters! At least JS is kind > enough to give err_bp when you try :-( > > So I have been taught a lesson, not to let APPARENT results fool me.
There were a few things that didn't work [properly] on the JM that when the JS came around, to avoid the problems, it just didn't allow them. eg upto JM, I seem to remember some sort of variable was allowed somewhere (was it integers on SELect? have the info lurking here somewhere), but I think something wasn't quite right, so JS just won't accept it in the first place I fell foul of something in a program I wrote for my JM that worked, but when I gave it to a friend to try on his JS, it didn't work - the JS just didn't accept one of the lines. At a guess either the DATE(yr,...) was intended to be, but it was never written, and so the JS errors; or it was a parameter check that was missing upto JM that was finally fixed in JS. _______________________________________________ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
