The answer he was looking for was really the *official* reason
why the stupid mystical incantation in necessary in the first
place
The answer to that is that it is *not* necessarily. And that answer
was given, but apparently that answer, no matter its demonstrable
truth, was not the answer he was looking for.
which Laurent correctly gave:
<snip>
You cannot "correctly" give a reason for the necessity of something
that is not necessary.
He had to ask in the newsgroup in the first place because
he wasn't even sure there IS an official reason for it
But he was not interested in listening when he was told that there was
not.
much less where to LOOK FOR IT.
There can be no "where to look" for the reason for the necessity of
something that is not necessary.
So the question WAS warranted
Not if the answer was pre-determined by ill-conceived preconceptions.
(it's not in the FAQ after all).
Difficult to pot the absence of a reason for the necessity of
something into any document. But it any event it is a long way from
being a frequently asked question.
By the way, this so called stupid mystical incantation is
necessary in IE for other things besides iframes.
It is not necessary for IFRAMEs, and (unsurprisingly) also unnecessary
for anything else.
It's necessary for selects, inputs and
textareas.
In what way? IE happily creates input, select and textarea elements
for me using DOM standard element creation code. (<input type="radio">
are the only elements that need special handling as a result of IE's
characteristics, and I still use non-branching code for those).
It may be necessary for other tags as well but I've mostly
found problems with form elements.
Your finding problems with the code you write is not, in itself, a
reason to be attributing issues to the wider world.
Basically, document.createElement
is b0rk in IE.
Based on what evidence? The case of targeting links and form
submission not being as simple as it could be has nothing to do with
the - createElement -, but rather is a consequence of employing -
setAttribute - and/or assigning to the element's - name - property not
having the side effect of associating a name with the corresponding
window object in IE. Thus if anything is broken it is the assignment
to the element's - name - property and/or - setAttribute -. And here
is a reflection of my point about the quality of analysis being so
poor that people fixate on mystical inactions that never do much more
that solve issues by coincidence; you are blaming the - createElement
- method, when with the argument "IFRAME" IE's - createElement -
method does no more and no less than any other - createElement -
method (creates a IFRAME element that is not attached to a document
and has no more than default/implied properties/attributes, and
certainly no name or id).