>What happens though when the data contains no spaces or line feeds. You see I'm parsing some very long HTML parameters which may or may not contain spaces. I was using Words and Wordnum, until IBM's website starting throwing long parameters at me and my cookies got pretty bug too. I think that MLINE will break up the line according to SET MEMOWIDTH, in which case it will wreck my data. And the max setting of MEMOWIDTH is 1024 (used to be lower)
>Tricky eh!

Not exactly - you roll your own in that case. Say you got a looong string wanting to cut it along "?" marks.

* lcLongString assumed to exist
lcCut=""
do while len lcLongString>0
lcCut=lcCut+left(lcLongString, 128)
lcLongString=subs(lcLongString,129)
for i=1 to words(lcCut,"?")
lcWord=wordnum(lcCut,i,"?")
Do Whatever with lcWord
endfor
lcCut=subs(lcCut, rat("?",lcCut)+1)
endd

Or something along these lines. Of course, this won't be as quick as when we could use MLINE(), but, hey, this is a harder case. OTOH, you can always say
lcLongString=chrtran(lcLongString, "?", " ")
and go the old track, if it surely doesn't contain spaces. Question mark here is taken just a supposed separator, your real separator may be anything else.

>And what this "create cursor mm (m m)". Very lazy indeed! ROFL :-). At least you didn't call the cursor "m" so you could have fun finding m.m

Well, the m.m trick is the only thing that stopped me from doing that - it's real alias would probably be Wnnn (nnn for area number), and that'd require at least two additional lines, so I calculated one extra "m" is cheaper :)