[whatwg] Web Forms 2.0 Editorial [minor] Section 5

On Wed, 16 Jun 2004, fantasai wrote:
>
> # A form data set is a sequence of control-name, index, current-value triplets
> # constructed from the controls identified in the previous step.
> #
> # The index here is unrelated to the repetition index mentioned earlier.
>
> I suggest calling this index the "control index" to help reduce the confusion.

Good idea.


> # The indices of these two virtual controls are handled separately and could,
> # depending on the values of other controls, end up with different values.
>
> Could you give an example for this?

It's an extreme edge case.

   <input name="test.x">
   <input type="image" name="text">

...the indicies of the image's two virtual controls will be different.
(text.x 1 and text.y 0).


> # The form data set also includes a list of which repetition blocks are
> # involved in the submission.
> #
> # For each control in the form data set, the control and the control's ancestors
> # are examined, up to but not including the first node that is a common ancestor
> # of the control and the form, or is the form itself. For each element so
> # examined, if it is a repetition block that is not an orphan repetition block
> # and whose template does have an ID, and that repetition block has not yet been
> # added to the list of repetition blocks, it is added.
>
> Two problems here:
>    1. This list of repetition blocks is coming out of nowhere, considering the
>       statement that "A form data set *is* a sequence of control-name, index,
>       current-value triplets..."
>    2. You're specifying an algorithm where a description would do.
>
> Replace these two paragraphs with:
>
>    The list of repetition blocks consists of all repetition blocks that include
>    at least one form control from the form control list, in document order.

This is not equivalent to the algorithm given above. Which is why I used
an algorithm.


> And change the beginning of this step from
>
> # A form data set is a sequence of control-name, index, current-value triplets
> # constructed from the controls identified in the previous step.
>
> to
>
>    The form data set consists of a form control list and a repetition block list.
>
>    The form control list is a sequence of control-name, control-index,
>    current-value triplets constructed from the controls identified in the
>    previous step.

Done.


> As for the tables, reverse the rows and columns. It'll make the
> tables much easier to read.

Do you have a script that would do that automatically (in the presence of
colspan/rowspan)? I am not doing this by hand, that's for sure. Also,
given the lengths of the enctype headers, I'm really not sure that that
would actually work well at all.

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'

Received on Wednesday, 23 June 2004 06:18:33 UTC