> I've recently been developing some applications# using cgi to invoke
> some legacy software on our server and convert the output into html.
> Part of this has involved writing/customizing a "GUI" for the software
> using html-forms. A few extra capabilities would make this a whole lot
> easier: namely
> * multiple clicks on an ismap image, eg to outline an area;
> I've checked the brief description of html v3 on
> and it appears that the first two items are already on the books in some
> called INPUT TYPE=range, INPUT TYPE=scribble,
> (though I tried putting a "range" controller in a form and
> it wasn't recognised by our version of Netscape).
FYI, <INPUT TYPE=scribble> is supported by at least one version of NCSA
Mosaic for X (version 2.6, I think; I can't seem to get it to work with
2.7b2 although that might be due to the bizarre X terminal I'm using).
It's pretty neat: you can do freehand drawing in an input box, and the
browser packages up the line segments you drew and POSTs them to the
server (or something like that).
There's a good interactive demo of this at:
(this requires a SCRIBBLE-supporting browser like Mosaic for X).
Some other suggestions for FORM enhancements:
I'd like to see an (optional) REQUIRED attribute added to the INPUT
element, which would mean "this field must be filled in by the user".
The only processing required by clients would be to make sure the value
is non-NULL before POSTing it.
It might also be useful to be able to specify TYPE=NUMERIC for input
fields, which would require numeric content for that field. Or,
possibly, TYPE=INTEGER and TYPE=FLOAT, where FLOAT allows decimals
and INTEGER doesn't. Of course, this gets messy due to differing
characters used for decimal points, but maybe this can be handled
with LANG, e.g. <INPUT NAME="field1" LANG="fr" TYPE=FLOAT>?
-- Gerald Oskoboiny <email@example.com> http://ugweb.cs.ualberta.ca/~gerald/