On Mon, 16 Feb 2004 13:32:13 -0700, Jack Hamilton
>"Chang Y. Chung" <chang_y_chung@HOTMAIL.COM> wrote:
>>(1) in both the case, it is critical to check if study is empty.
>>when it is empty, you get a stransge value for outfile, also -- I
>>this means that study is a required parameter. Consider making it
>I used to use positional parameters, and now I almost never do if
>there's more than one parameter. Named parameters give a clue what the
>parameters are trying to do.
You are absolutely correct. There is no reason not to stick to named
parameters only -- I agree with you.
Having said that, let me try to defend the opinion, if I may. I just
thought, in some cases, the code reads better if you make required
parameters as positional parameters. For example, if you are writing a
macro called %sortBy(by=, options=), I just thought it would be nicer to
be able to write:
instead of the following which is kinds of redundant,
Besides, you can do %sortBy(by=Vara), if you want to, even when the "by"
is a positional parameter.
The key is that if a macro parameter is required, then the parameter value
should be such a critical one that the macro name itself should give
intuitive enough clue about what is required. And if so, then we may get a
code that is, in fact, more easily understood by humans.
But, again, I guess what is "intuitive" depends, so it may be better to
stick to a simpler (and better) coding style. On top of it, I have no
doubt about your experience and wisdom. So, ... I have no problem writing
macros without positional parameters :-)