> Isn't the real solution to simply remove SLA from AST as an option and for
> AST to simply use PAL routines all the time (and mangled function names).
Yes - we could do that I guess. I think that Rodney's initial idea was
that a clear dependency on SLALIB may actually be a selling point for AST,
but I suspect that has not turned out to be the case.
Thre is also the slight danger that other thirdparties may have developed
their own c wrappers which may similarly fall foul of names clashes with
the new c wrappers distributed with slalib.
David
|