Dave Westbury writes:
>Thanks. BTW I have one small query about SMS.RESRI error return. QDOS does
not
> return any error (according to docs), but SMSQ can return ERR_OM. The
problem
> is; if there is enough space already on RI stack then SMSQ code doesn't
return
> d0=0. Can I guarantee that if I zero d0 on entry to SMS.RESRI it will only
get
> set to err_om if it occurs? I haven't tried QDOS/Minerva on this yet, I
hope
> they do the same.
Superbasic returns straight back to the parser on an OM error in sb.resri.
Since discovering this by chance some years ago, I stopped testing for it.
That it now returns an error is rather awkward...
I also notice that the value pointers in the name table have all gone
absolute,
ie even SB variable values are stored in some user heap outside the SB.
Havent got round to arrays yet, but all my array-handling toolkits still
seem to work, so perhaps no change there?
Im sure looking forward to "SBASIC, The Definitive Handbook" ;)
Per
