Problem with FORMs in draft-ietf-html-spec-04.txt

Paul Burchard (burchard@horizon.math.utah.edu)
Sun, 16 Jul 95 20:10:35 EDT

I just found what I consider to be a serious bug in the FORMs
section of the latest HTML 2.0 draft <draft-ietf-html-spec-04.txt>.
(I apologize for reporting this so late, due to unfortunate
electronic circumstances. Hopefully it's not to late to fix it
before standardization.) Also, a minor typo follows.

1. Multiple Values.

The example in Section 8.2.4 incorrectly handles multiple values.
Without precedent in practice (as far as I know), it folds the
multiple values into one (separating them with a non-syntactically
significant comma, no less). But the intention and reality of FORMs
since 1993 has been to deliver true multiple values. Thus a form
which contains

<LI>Kent <INPUT NAME="city" TYPE=checkbox VALUE="kent">
<LI>Miami <INPUT NAME="city" TYPE=checkbox VALUE="miami">

and supposedly yields:

name=John+Doe&gender=male&family=5&city=kent%2Cmiami&
other=abc%0D%0Adef&nickname=J%26D

should yield, according to long-established existing practice:

name=John+Doe&gender=male&family=5&city=kent&city=miami&
other=abc%0D%0Adef&nickname=J%26D

In Section 8.1.2.3, the same error is hinted at by the wording.
The VALUE of the CHECKBOX field is described as

VALUE
The portion of the value of the field contributed by
this element.

(without any hint as to how different "portions" are put together),
when the correct description would be:

VALUE
The value contributed by this element, if checked.
Multiple checked elements with the same name produce
multiple name/value pairs in the submission.

Although it is not mentioned there, the issue of multiple values
also arises in Section 8.1.3 with <SELECT MULTIPLE>. Again,
multiple name/value pairs should be generated.

2. A Typo.

In Section 8.2.1 there is a spurious `#' char in:

`<a href="http://host/?x=1&#amp;y=2">'

--------------------------------------------------------------------
Paul Burchard <burchard@math.utah.edu>
``I'm still learning how to count backwards from infinity...''
--------------------------------------------------------------------