a bone to pick

[JJBBDD]

as a strong CF advocate, take this as a friendly gripe. and if someone more
experienced can explain why this 'problem' is actually not a problem, i'd
appreciate it.

the issue, generalized, is that certain cf tags will work differently
depending on which attributes are included. (note: not the attribute value
which can always be dynamically evaluated).

EG 1, <cfquery...> when running queries of queries you cannot include the
datasource attribute.
EG 2 <cfchart...> in some instances it doesn't make sense to include a scaleTo
attribute, but the attribute won't take a null value or empty string or the
like without throwing an error.

there are probably more examples of this, but the problem should be clear
enough: you end up with redundant code having to fork between the tags (one
with the pesky attribute, one without), which leads to extra maintainence and
bugs and less modularity and is pretty much antitheitcal to the spirit of OO
code, a spirit that CF has been embracing with gusto.

that said, i'm wondering what solutions there are or could be to this problem.
perhaps a reserved word that, when an attribute is set to it, tells the
compiler to ignore it. OR, a way to dynamically create tags from strings--
something like a delayed runtime.

thoughts? insults? harangues?

[MikerRoo]

I agree. Objects, including CF tags, should ignore attributes if they are not
used or not appropriate.
Isn't this also the paradigm for HTML?

The thing to do is to [url]http://www.macromedia.com/go/wish/[/url] when you see
something like this.