Subject:
|
Re: MSIE5 better but still buggy (was: Re: Why MSIE sucks for the HTML writer)
|
Newsgroups:
|
lugnet.admin.general
|
Date:
|
Mon, 20 Mar 2000 19:31:34 GMT
|
Viewed:
|
1372 times
|
| |
|
|
In lugnet.admin.general, Asher Kobin writes:
> > > Yet another totally braindead MSIE problem is multiple-select lists with
> > > the <SELECT> tag. If none of the <OPTION> items within a <SELECT> have
> > > the SELECTED attribute, MSIE is its infinite wisdom decides to auto-select
> > > the first item for you! Wonderful. So you have to add an empty foofoo
> > > item at the top if you're serving pages to an MSIE browser, if you don't
> > > want anything selected by default. (Perhaps MS has fixed this bug by now.)
> >
> > Still broken in MSIE5! Bummer.
>
> Why do you consider this difference in behavior between the two browsers to
> be "broken" in IE5? Obviously, I can make the reverse argument and say it
> is "broken" in Nav. Neither browser is going to change this behaviour, it
> will break compatibiliy.
Well, maybe you're right about that after all. According to section 8.1.3 of
RFC1866[1] (HTML 2.0), "[t]he initial state has the first option selected,
unless a SELECTED attribute is present on any of the <OPTION> elements."
(This is completely bizarre!)
The HTML 4.0 spec redefines this to be undefined (it's up to the application
whether or not to select the first item in this case).
Thus, it appears that Netscape _went against_ the HTML 2.0 spec when it
implemented the <SELECT> tag, and that Microsoft _adhered_ to the HTML 2.0
spec when it implemented the <SELECT> tag. We could say then that it's
actually the spec itself which is broken, since it provides no way to
guarantee that a multiple-select list box comes up with exactly zero items
preselected. To its credit, in violating the HTML 2.0 spec, Netscape made
the <SELECT> tag work better than it was intended to. It is unfortunate
(IMHO) that Microsoft did not also violate the spec. If they had, then
the HTML 4.0 might now define <SELECT> in a more versatile way than the
crippled way that it does now. Does MSIE give any other way (though perhaps
some other ATTRIBUTE) whereby an HTML author can send a multiple-select list
and guarantee that _zero_ of the list items are preselected? That would be
a nice workaround to the shortcoming in the spec.
--Todd
[1] http://www.ietf.org/rfc/rfc1866.txt
|
|
Message has 2 Replies:
Message is in Reply To:
131 Messages in This Thread:
(Inline display suppressed due to large size. Click Dots below to view.)
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
