David Crayford:
>If the code doesn't have to be supported by a human then you might be
right.
>If it does then nope! Code that is generated by a program is usually the
>output of a compiler that generates object code that we generally don't
>care about until we need to debug it. If a program generates REXX code then
>it should generate good REXX code that doesn't have telescoping function
arguments.
As a theological discussion, I find this interesting. Our case is like this
(and isn't Rexx, though it could be):
rc = somefunction(tuplecount, elementcount,
thing1, data1, length1,
thing2, data2, length2,
thing3, data3, length3,
thingn, datan, lengthn)
The tuplecount is the number of these triplets, and the elementcount is the
number of things in each datan. You can have up to 100 of these tuples.
Does that offend your sensibilities? I'm honestly curious-I generally agree
100% with your statement, but this feels like a maybe-exception.
If it helps, note that in this instance, each of the tuples are
different-that is, it's not like they could be combined, as they're pointing
to different data types of different lengths. And there is also an API that
handles essentially one item:
rc = somefunction(thing, data, length)
Obviously the same idea as the first one with tuplecount and elementcount
both set to 1.
I've been doing this for almost 40 years, but am not a Computer Science
graduate by any means, so I'm always interested in the theology behind this
stuff, and where it does and doesn't apply.
Cheers,
.phsiii
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
