CSS Frameset Borders

Hello,

I am in the process of trying to make some existing web pages comply with
W3C's HTML 4.01 Recommendation ('strict' where possible, 'loose' where
not -- i.e. when using framesets and, hence, the 'target' attribute of the
<a> tag...)

I have so far been unsuccessful in finding a compliant equivalent of the
'border' attribute of the <frameset> tag.

<frameset border="0"....>
<frame frameborder="0"...>

produces frames without borders, but W3C's validator says that there is no
attribute 'border'.

Hello,

I am in the process of trying to make some existing web pages comply with
W3C's HTML 4.01 Recommendation ('strict' where possible, 'loose' where
not -- i.e. when using framesets and, hence, the 'target' attribute of the
<a> tag...)

I have so far been unsuccessful in finding a compliant equivalent of the
'border' attribute of the <frameset> tag.

<frameset border="0"....>
<frame frameborder="0"...>

produces frames without borders, but W3C's validator says that there is no
attribute 'border'.

Attempting to use CSS instead, the following seem to have no effect:

frameset
{
border-width: 0px; -AND/OR-
border-style: none; -AND/OR-
border-style: hidden; -AND/OR-
border-collapse: collapse; -whatever THAT means-
}

but all leave me with a narrow two- or three-pixel (ish) border that can
only be described as tenacious.

Does anyone know how to achieve the equivalent of <frameset border="0"...>
using CSS to produce something W3C compliant?

If it makes any difference, I'm viewing my attempts using IE6.

<frameset>
<frame frameborder="0">
</frameset>

That should do it. Also, you'll probably want to be validating the page
that contains the frameset with the Frameset DTD.

The white 'borders' you're still seeing, is the
framespacing, which can be set to "0" and should be applied
to the framesets, but unfortunately isn't valid under 4.01 :-(

<frameset>
<frame frameborder="0">
</frameset>

That should do it. Also, you'll probably want to be validating the page
that contains the frameset with the Frameset DTD.

Unfortunately, this doesn't work -- I still get that two- or three-pixel
border without a 'border=0' in the <frameset> tag.

The white 'borders' you're still seeing, is the
framespacing, which can be set to "0" and should be applied
to the framesets, but unfortunately isn't valid under 4.01 :-(
As you say, using <frameset framespacing="0"...> works as well as

<frameset border="0"...> but neither of them are 4.01-compliant. So, unless I find an alternative that is compliant, I may as well stick with 'border': it's less typing! ;)

A.

I still have to see the light why validation is so important, what the whole
DOCTYPE circus is good for.

In my view, a site validates if a user can see it in his browser.

Ivo:

In my view, a site validates if a user can see it in his browser.
Is it enough if all users _today_ can see the site in their browsers?

Or do you care about the users (and the browsers) of next year, and the
year after that, as well?

Yes, o very much. That 's exactly why all that "content" out in the open
must be looked after continuously and updated to reflect new standards or
become ...old.
Technology changes and I am convinced the browser of tomorrow will be able
to detect the type of document without a !DOCTYPE. Less is more. I also have
no doubt it will be able to render a font tag even though the standard tells
otherwise.
I am no enemy of standards, but have never understood what this
tag-to-end-all-tags was supposed to *do* except validate. As soon as soon as
browser sniffing was invented, browser spoofing was invented, as soon as
doctypes were invented, quirksmode was invented. I read on another
developer's site: suddenly twice as many browsers to test my pages in! That
's more or less the opposite effect of what a standard is supposed to
achieve, isn't it?
How do you know all users can see the site in their browsers?
A w3c standard is no guarantee. Being indexed by our friend the Googlebot
is, yet another kind of validation.
Yours conformistically,
Ivo
--
Bertilo Wennergren <be******@gmx.net> <http://www.bertilow.com>

"Bertilo Wennergren" <be******@gmx.net> wrote in message
news:c1*************@news.t-online.com...

Ivo:

> In my view, a site validates if a user can see it in his browser.

Is it enough if all users _today_ can see the site in their browsers? Or do you care about the users (and the browsers) of next year, and the
year after that, as well? Yes, o very much. That 's exactly why all that "content" out in the open
must be looked after continuously and updated to reflect new standards or
become ...old.
Would it be better to make the content future safe without the need for
continuous maintenance - as far as possible? If you make sure you're
code is valid, there's much less risk that you'll have to redo the code
when new and better browsers appear. It's not fail-safe, but it's a very
good start.
Technology changes and I am convinced the browser of tomorrow will be able
to detect the type of document without a !DOCTYPE. Less is more. I also have
no doubt it will be able to render a font tag even though the standard tells
otherwise.
Probably. But that does not go for each and every mistake that makes you
code invalid. It's has happend before that pages that used to work in
browsers, stopped working when a new and more standards compliant
version of a browser came out.
I am no enemy of standards, but have never understood what this
tag-to-end-all-tags was supposed to *do* except validate. As soon as soon as
browser sniffing was invented, browser spoofing was invented, as soon as
doctypes were invented, quirksmode was invented. I read on another
developer's site: suddenly twice as many browsers to test my pages in! That
's more or less the opposite effect of what a standard is supposed to
achieve, isn't it?
I think you're mixing a lot of different things in the same pot in a
very confused way.

Don't sniff. Use the latest (and strictest) HTML version that works in
major browsers. Make sur you trigger standards mode. Validate. Test in
the major browsers (including Lynx). If anything gives problems,
simplify. Stay valid. What more do you want?
How do you know all users can see the site in their browsers?

A w3c standard is no guarantee.

Ivo:

"Bertilo Wennergren" <be******@gmx.net> wrote in message
news:c1*************@news.t-online.com...

Ivo:

Technology changes and I am convinced the browser of tomorrow will be
able to detect the type of document without a !DOCTYPE. Less is more.
I also have no doubt it will be able to render a font tag even though
the standard tells otherwise.

Probably. But that does not go for each and every mistake that makes
you code invalid. It's has happend before that pages that used to work in
browsers, stopped working when a new and more standards compliant
version of a browser came out.

Mistakes in the code are made. Some mistakes make a page not pass 'the
validation test', some mistakes leave a page blank for the human eye, some
mistakes cause a search engine spider to misinterpret a site, and the common
goal is to make those mistakes overlap eachother more and more in our
understanding, defining and engineering of an Internet technology as
undemanding as possible, where even the most messy markup can be accessed
and interpreted.

I am no enemy of standards, but have never understood what this
tag-to-end-all-tags was supposed to *do* except validate.

(...) Use the latest (and strictest) HTML version that works in
major browsers. Make sur you trigger standards mode. Validate. Test in
the major browsers (including Lynx). If anything gives problems,
simplify. Stay valid. What more do you want?
I want to be able to fill my HTML with embeds, bgcolors and -like the OP-
frameborders, and other stuff that is simply ignored by browsers unless it
understands it and when I am happy with the result, seeing that the page
gets visited by both bots and humans, be able to say that my code is valid.
My point is merely I find the Validation thing often overrated in
discussions here as it has restricted our freedom more than anything else
and it will never replace that final test: the real validation does not take
place until a human sees the page. If it is a ten year old document, the
human (and a good browser) will be prepared for some old-fashioned wording.
Ivo
--
Bertilo Wennergren <be******@gmx.net> <http://www.bertilow.com>

"Bertilo Wennergren"

Use the latest (and strictest) HTML version that works in
major browsers. Make sur you trigger standards mode. Validate. Test in
the major browsers (including Lynx). If anything gives problems,
simplify. Stay valid. What more do you want?

I want to be able to fill my HTML with embeds, bgcolors and -like the OP-
frameborders, and other stuff that is simply ignored by browsers unless it
understands it and when I am happy with the result, seeing that the page
gets visited by both bots and humans, be able to say that my code is valid.
Then make a carefully crafted custom DTD that includes the additional
stuff you need, and use the validator to control that you've correctly
followed the rules you've set up for yourself.
My point is merely I find the Validation thing often overrated in
discussions here as it has restricted our freedom more than anything else
and it will never replace that final test: the real validation does not take
place until a human sees the page.

Ivo:

My point is merely I find the Validation thing often overrated in
discussions here as it has restricted our freedom more than anything else
and it will never replace that final test: the real validation does not take
place until a human sees the page.

Of course validation can't replace testing in actual browsers, and with
actual humans. I don't think anyone seriously advocates anything else.