On Mon, 2006-04-09 at 15:18 -0400, David A. Wheeler wrote:
> So far I have heard tremendous silence about naming conventions for
> function parameters. Which makes sense, because they have absolutely
> no effect on the correctness of the spec. But for the spec to have a "finished"
> look, we really should pick a single convention for us all to use. It helps readers
> slightly by having a single convention. ANd the current pseudo-convention for
> Uppercase parameter names doesn't look as pretty in formulas.
>
> So, here' s what I propose by "fiat":
> * Types need to be in UpperCamelCase.
> * Parameter names need to be lowerCamelCase.
> Try to use the same name between different functions for the same thing.
> Where practical, make them
> meaningful (indicating MEANING not type), e.g.,
> "basis" for date calculation basis (where 0=360-day year), etc.
> Use "x" for math functions taking one numerical parameter,
> "x" and "y" for math functions taking two numerical parameters.
>
> Anyone have a crisis with this, please let us know.
I think we should not force all math functions to use x and y as there
variable names, but choose appropriate names:
COMBIN(x;y) looks strange it should probably be COMBIN(n;k) since the
arguments are integers. x;y looks like they are reals. Similarly the
second argument of BETADIST should probably be the greek letter alpha
rather than y.
Andreas
--
Andreas J. Guelzow
Pyrenean Shepherds
--
Prof. Dr. Andreas J. Guelzow
Dept. of Mathematical & Computing Sciences
Concordia University College of Alberta
