were does this HTML syntax come from?

[Andy Fish]

Looking at an HTML document recently, I found this syntax which was
completely new to me

<![if !supportAnnotations]>
<style>...</style>
<![endif]>

it looks to be some kind of conditional logic but it's not in a javascript
block, just in the normal HTML markup. none of by browsers took exception to
it

there was no attempt to define or set the variable "supportAnnotations"
anywhere, so I'm not sure if it's referencing some built-in variable or
what.

Can anyone point me at anything which explains this syntax?

TIA

Andy

[Leif K-Brooks]

Andy Fish wrote:

Looking at an HTML document recently, I found this syntax which was
completely new to me

<![if !supportAnnotations]>
<style>...</style>
<![endif]>

It's proprietary garbage invented by Microsoft which will only work in
their so-called Web browser. You shouldn't try to use it on the public
Internet.

[Martin Honnen]

Andy Fish wrote:

Looking at an HTML document recently, I found this syntax which was
completely new to me

<![if !supportAnnotations]>
<style>...</style>
<![endif]>

it looks to be some kind of conditional logic but it's not in a javascript
block, just in the normal HTML markup. none of by browsers took exception to
it

there was no attempt to define or set the variable "supportAnnotations"
anywhere, so I'm not sure if it's referencing some built-in variable or
what.

Can anyone point me at anything which explains this syntax?

It is likely an attempt to use MS IE's conditional comments:
http://msdn.microsoft.com/library/de...omment_ovw.asp

--

Martin Honnen
http://JavaScript.FAQTs.com/

[paul haine]

Leif K-Brooks <eu*****@ecritters.biz> wrote in news:312vh2F35vu4nU1@uni-
berlin.de:

Andy Fish wrote:

Looking at an HTML document recently, I found this syntax which was
completely new to me

<![if !supportAnnotations]>
<style>...</style>
<![endif]>

It's proprietary garbage invented by Microsoft which will only work in
their so-called Web browser. You shouldn't try to use it on the public
Internet.

Using a form of these conditional comments is very useful for filtering
hacks or content to specific versions of Internet Explorer, though:

http://www.quirksmode.org/css/condcom.html

Validates, and all non-IE browsers will ignore it.

--
paul haine
http://joeblade.com/
http://design.joeblade.com/

[Richard]

Andy Fish wrote:

Looking at an HTML document recently, I found this syntax which was
completely new to me <![if !supportAnnotations]>
<style>...</style>
<![endif]>

The ! generally indicates this is a comment.
Looks like maybe the author uses them for his own purposes and fails to
remove them.
There is nothing in standard html that supports <![ ]>.

[Alan J. Flavell]

On Sat, 4 Dec 2004, Richard wrote:

> <![endif]>

The ! generally indicates this is a comment.

Not to anyone who understands HTML, it doesn't.

Don't you recognise a Microsoftism when you see one?

[Jan Roland Eriksson]

On Sat, 4 Dec 2004 09:58:38 -0600, "Richard" <An*******@127.001> wrote:

Andy Fish wrote:

Looking at an HTML document recently, I found this syntax which was
completely new to me <![if !supportAnnotations]>
<style>...</style>
<![endif]>

The ! generally indicates this is a comment.
BEEP, ALARM, BEEP (i.e. that is not correct)

The '<!' char sequence is a "Markup Declaration Open" mark as per the
SGML Concrete Reference Syntax and it's as correct in HTML as it is in
better applications of SGML.
There is nothing in standard html that supports <![ ]>.

That is correct, but probably you do not understand the real reason why.

SGML (and thus HTML) defines '<!' + '[' as the start of a "Marked
Section" and a marked section opener must be followed by a valid "Status
Keyword". A correct marked section would be as in following example.

<![ IGNORE [Anything in here will be ignored by an SGML parser]]>

Looking back at the example markup that started this thread it can be
clearly stated that the word 'if' is not a correct SGML status keyword
and thus the whole example becomes totally invalid from an SGML (HTML)
standpoint.

If it "works" for some level of "works" in today's browsers that is all
because of existing buggy behavior in those browsers.

This setup OTOH...

<!--[if !supportAnnotations]>
<style>...</style>
<![endif]-->

....would make for the construct to be a valid comment for all browsers,
except for MSIE which happens to be programmed to process certain
contents of SGML comments as a sort of a "private programming language".

For the one who really wants to bother his mind with e.g. style sheet
designs that gives similar rendering on all browsers, including MSIE,
the MS "Conditional Comment" thingy could be used to set up the specific
error correction styling that IE needs to behave decent visavi already
decent browsers.

--
Rex

[Alan J. Flavell]

On Wed, 1 Dec 2004, paul haine wrote:

<![if !supportAnnotations]>
<style>...</style>
<![endif]>

Using a form of these conditional comments

Petitio principii. The above are not "comments", not in HTML, not in
SGML.
is very useful for filtering
hacks or content to specific versions of Internet Explorer, though:

http://www.quirksmode.org/css/condcom.html
Fairynuff, but the cited URL refers to

|| <!--[if IE]>
Validates,
Don't you see the difference?
and all non-IE browsers will ignore it.

The latter form is a comment, right.

The cited form, if it happens to validate, is sheer happenstance.
Lord knows what an SGML browser would make of it.

Where's Arjun Ray when we need him? ;-)

[Jan Roland Eriksson]

On Sun, 5 Dec 2004 22:59:18 +0000, "Alan J. Flavell"
<fl*****@ph.gla.ac.uk> wrote:

[...]

Where's Arjun Ray when we need him? ;-)

Good question, and I would be the first to really applaud his
reappearance here.

From what I know, in concentrated form...

Arjun had his company office on the south end of Manhattan.
When the twin towers fell down, his point of business became isolated
south of the street that the US authorities decided to use as a
"terrorist" border line (and that included total electronic isolation).

Suffice to say that Arjun lost all contact with his present, and future
potential, clients because of this, and he had to close down his
operation.

He managed to moved some of the more important records over to his own
apartment located straight east and over the water from his Manhattan
office, but basically; since the authorities did not even give
permission for him to visit his own office, he was stumped.

I have no accurate knowledge of his whereabouts today; last record I
could find on Usenet is in rec.pets.cats.*

(Arjun loves Cats as "mans best friend" and so do I, so we do have that
in common at least)

--
Rex

[Michael Winter]

On Sun, 5 Dec 2004 22:59:18 +0000, Alan J. Flavell <fl*****@ph.gla.ac.uk>
wrote:

On Wed, 1 Dec 2004, paul haine wrote:

[snip]

http://www.quirksmode.org/css/condcom.html

Fairynuff, but the cited URL refers to

|| <!--[if IE]>

However, what the article fails to mention is that there is a difference
between

<![...]> and <!--[...]>

As the article states, conditional comments are only understood by IE5+.
The first form of comment allows earlier IE versions to parse the contents
of the "comment" (Microsoft use the term, "downlevel-revealed"), whilst
the second form does not ("downlevel-hidden"). So, which form you use may
be of some importance.

See
<URL:http://msdn.microsoft.com/workshop/author/dhtml/overview/ccomment_ovw.asp>.

Opera users: You may need to either disable Javascript or spoof as IE
before viewing the document above. Microsoft seem intent on redirecting to
their tree-based navigation system which is incompatible with Opera.

[snip]

Mike

--
Michael Winter
Replace ".invalid" with ".uk" to reply by e-mail.

[Alan J. Flavell]

On Mon, 6 Dec 2004, Michael Winter wrote:

On Sun, 5 Dec 2004 22:59:18 +0000, Alan J. Flavell <fl*****@ph.gla.ac.uk>
wrote:

On Wed, 1 Dec 2004, paul haine wrote:
[snip]

http://www.quirksmode.org/css/condcom.html

Fairynuff, but the cited URL refers to

|| <!--[if IE]>

However, what the article fails to mention is that there is a difference
between

<![...]> and <!--[...]>

Quite! Which was indeed the point that I was trying to bring out.
The first form of comment
There's only one form of comment there. The other is something else.
So, which form you use may be of some importance.
I don't see any particular objection to using the comment (by those
who want to treat IE differently than a WWW browser, I mean) - not
even in a WWW context.

The other one is a different matter, though.
See
<URL:http://msdn.microsoft.com/workshop/author/dhtml/overview/ccomment_ovw.asp>.

Sure. Then see what an SGML parser makes of the other one.

MS (ignorantly|mischievously) -calling- it a comment does not
miraculously make it into one.

[Michael Winter]

On Mon, 6 Dec 2004 13:36:04 +0000, Alan J. Flavell <fl*****@ph.gla.ac.uk>
wrote:

On Mon, 6 Dec 2004, Michael Winter wrote:
[snip]

However, what the article fails to mention is that there is a
difference between

<![...]> and <!--[...]>

Quite! Which was indeed the point that I was trying to bring out.

I was referring to the functional difference, but I am aware of the fact
that <![ is not in any way an actual comment delimiter. It was certainly
not my intention to imply otherwise.

The first form of comment

There's only one form of comment there. The other is something else.

Instances of 'comment' in my previous post should have been quoted in a
gesture of acknowledgement. In the end, only one was.

[snip]
MS (ignorantly|mischievously) -calling- it a comment does not
miraculously make it into one.

Absolutely! Again, I was simply drawing attention to the functional
difference that Microsoft applies to the two forms, and that the second
form above is not necessarily an equivalent replacement of the first.

Mike

--
Michael Winter
Replace ".invalid" with ".uk" to reply by e-mail.

[Arjun Ray]

On Wed, 01 Dec 2004 16:03:48 +0000, paul haine wrote:

Using a form of these conditional comments is very useful for filtering
hacks or content to specific versions of Internet Explorer, though:

Indeed. Here is an example:

http://www.nyct.net/~aray/

(with thanks to Alan Flavell for suggesting an "improvement").

--
The day Microsoft makes something that doesn't suck is probably the
day they start making vacuum cleaners. -- Ernst Jan Plugge

[Arjun Ray]

On Mon, 06 Dec 2004 02:10:56 +0100, Jan Roland Eriksson wrote:

I have no accurate knowledge of his whereabouts today;
To the consternation and despair of many, I'm still around, in the same
place. The usual e-ddresses still work, too (except the q2.net one).
last record I could find on Usenet is in rec.pets.cats.*

Yes, I follow those groups regularly.

[Jan Roland Eriksson]

On Sun, 12 Dec 2004 05:03:23 GMT, "Arjun Ray" <ar**@nmds.com.invalid>
wrote:

On Mon, 06 Dec 2004 02:10:56 +0100, Jan Roland Eriksson wrote:

I have no accurate knowledge of his whereabouts today;

To the consternation and despair of many, I'm still around, in the same
place...

Welcome back; you have been missed :-)

--
Rex

[Nick Kew]

In article <pa****************************@nmds.com.invalid >,
"Arjun Ray" <ar**@nmds.com.invalid> writes:

I have no accurate knowledge of his whereabouts today;

To the consternation and despair of many, I'm still around, in the same
place. The usual e-ddresses still work, too (except the q2.net one).

Yay! Welcome back to the madhouse. Can I look forward to further
enlightement of a technical nature?

--
Nick Kew

Nick's manifesto: http://www.htmlhelp.com/~nick/