On Jan 3, 2022, at 12:06:53, Joseph Reichman wrote: > > I’m just trying to generate the 4 fields need > To define a variable to Rexx when using assembler > 1) length of fetch value > 2) address of the name of the variable > 3) length of variable name > 4) address of the fetch buffer > "name"? I see too many indirections. What facility needs the length or address of the name? IRXEXCOM and ADDRESS LINKMVS surely don't.
> I 100 % agree with Charles that using the adata is better > It’s just not practical > Alas, if what both you and Peter say are true, no scheme mentioned so far is both practical and correct. Take your pick. > I would have to be aware every time one of these copybooks changes > re-assemble it to get adata > Nobody is going to tell me so > Everyday I would have to check endeavor to see if they copybooks were changed > Or else I would have to read production programs that use the copybooks > Isn't that equally so for the AREAD scheme? Or do you plan to read the copybooks dynamically? You might as well invoke HLASM dynamically. Concerning length of this thread: <https://en.wikipedia.org/wiki/Law_of_triviality>. -- gil