🚀 go-pugleaf

On Mon, 23 Apr 2001 00:30:23 +0200, "Alan J. Flavell"
<flavell@mail.cern.ch> wrote:

>On Sun, 22 Apr 2001, Dr John Stockton wrote:
>
>> >Your <form> with no action also upsets a genuine HTML validator.
>> >As well as the #rrggbb value which is lacking quotes.  But you could
>> >easily have determined that for yourself.
>>
>> Given that my <form> are needed (in AIUI at least one browser), and that
>> no action is actually required, what do you suggest for an empty action?
>
>When you say "no action is actually required" I presume this means
>"the form isn't supposed to work when javascript is disabled".
>Speaking in general terms I'd have to counsel against that approach,
>but it's your web page after all. 

Indeed and writing the form tags with script would overcome the
problem, it is a page about javascript, and the form is there for a
demonstration, so I'm certainly not against it's use, but writing it
out would seem harmless.

>[1]Validators of course can't properly parse an HTML document that's
>meant to be partly generated at run-time from Javascript, so there are
>bound to be additional anomalies if you're doing that.

DOM2 browsers, (or innerHTML extension browsers) could of course
easily provide an add-in which provided validation of just such
processed pages at any stage, indeed some suggestions I made recently
to Nick about page valet would've allowed this (well with the addition
of accepting the 'page' to be validated as POST data, rather than a
url - A wholly client-side solution may be better though as it could
more economically validate the page at each step of script execution.

Jim.