From jens.meiert at erde3.com  Thu Mar  1 02:53:05 2007
From: jens.meiert at erde3.com (Jens Meiert)
Date: Thu, 01 Mar 2007 11:53:05 +0100
Subject: [HTML5] [html] Elements within "title"?
In-Reply-To: <004c01c75b9b$c9dd23c0$2f01a8c0@Seka>
References: <004c01c75b9b$c9dd23c0$2f01a8c0@Seka>
Message-ID: <20070301105305.255300@gmx.net>

> From my experience, the title is normally rendered in places that
> have very limited constraints. Does the <title> element really need
> to be styled?

It's not about styling - and this element most likely doesn't need to be styled at all - but rather about marking up what we mark up everywhere else in documents.

So does it really make sense to correctly mark up abbreviations in documents except in one of their probably most important elements? Likely there are historical reasons, certainly there are compatibility issues, but surely that is inconsistent anyway.

> I think it was a valid suggestion and a good discussion, but I don't
> see any positive effects from an actual implementation of this.

Maybe this change would only make the spec more consistent and all of us slightly happier, but it won't be "cheap", presumably ...

-- 
Jens Meiert
http://meiert.com/en/



From jens.meiert at erde3.com  Thu Mar  1 07:15:29 2007
From: jens.meiert at erde3.com (Jens Meiert)
Date: Thu, 01 Mar 2007 16:15:29 +0100
Subject: [HTML5] [html] Elements within "title"?
In-Reply-To: <abd6c8010703010707m6f3322e9s6e114e17c0e306b2@mail.gmail.com>
References: <20070227085415.317650@gmx.net> <45E3FE7B.5070902@lachy.id.au>	
	<Pine.GSO.4.64.0702271325580.5105@hopeatilhi.cs.tut.fi>	
	<433C4B7E-A113-4A5B-8354-D39F352C265D@iki.fi>	
	<Pine.GSO.4.64.0702271627580.4057@hopeatilhi.cs.tut.fi>	
	<45E59DCC.4080304@lachy.id.au>	
	<50FCB9CB-A8C9-4427-8796-E2B77EA21259@nickshanks.com>	
	<49C257E2C13F584790B2E302E021B6F912A0C694@winse-msg-01.segroup.winse.corp.microsoft.com>
	<45E6C179.2050306@sebastianmendel.de>
	<abd6c8010703010707m6f3322e9s6e114e17c0e306b2@mail.gmail.com>
Message-ID: <20070301151529.269910@gmx.net>

> There's nothing that prevents you from saying the content model of
> title is limited to:
> 
> sup, sub, em, strong, span, abbr (others I'm forgetting).

Right, and that's basically what is proposed [1] (albeit not with a "comprehensive" content model yet).

I understand that there's basic consensus on this issue, but the question is if that problem/inconsistency could be addressed in any future (X)HTML spec? (By the way, this somewhat tempts me to cross-post all the time.)


[1] http://meiert.com/en/blog/20070226/html-semantics-of-title-element-content/

-- 
Jens Meiert
http://meiert.com/en/



From jens.meiert at erde3.com  Thu Mar  1 07:15:29 2007
From: jens.meiert at erde3.com (Jens Meiert)
Date: Thu, 01 Mar 2007 16:15:29 +0100
Subject: [HTML5] [html] Elements within "title"?
In-Reply-To: <abd6c8010703010707m6f3322e9s6e114e17c0e306b2@mail.gmail.com>
References: <20070227085415.317650@gmx.net> <45E3FE7B.5070902@lachy.id.au>	
	<Pine.GSO.4.64.0702271325580.5105@hopeatilhi.cs.tut.fi>	
	<433C4B7E-A113-4A5B-8354-D39F352C265D@iki.fi>	
	<Pine.GSO.4.64.0702271627580.4057@hopeatilhi.cs.tut.fi>	
	<45E59DCC.4080304@lachy.id.au>	
	<50FCB9CB-A8C9-4427-8796-E2B77EA21259@nickshanks.com>	
	<49C257E2C13F584790B2E302E021B6F912A0C694@winse-msg-01.segroup.winse.corp.microsoft.com>
	<45E6C179.2050306@sebastianmendel.de>
	<abd6c8010703010707m6f3322e9s6e114e17c0e306b2@mail.gmail.com>
Message-ID: <20070301151529.269910@gmx.net>


> There's nothing that prevents you from saying the content model of
> title is limited to:
> 
> sup, sub, em, strong, span, abbr (others I'm forgetting).

Right, and that's basically what is proposed [1] (albeit not with a "comprehensive" content model yet).

I understand that there's basic consensus on this issue, but the question is if that problem/inconsistency could be addressed in any future (X)HTML spec? (By the way, this somewhat tempts me to cross-post all the time.)


[1] http://meiert.com/en/blog/20070226/html-semantics-of-title-element-content/

-- 
Jens Meiert
http://meiert.com/en/




From webmaster at keryx.se  Sun Mar  4 12:30:48 2007
From: webmaster at keryx.se (Keryx Web)
Date: Sun, 04 Mar 2007 21:30:48 +0100
Subject: [HTML5] List of discontinued attributes
Message-ID: <45EB2C78.3030100@keryx.se>

Hi

Has anyone made a list of attributes that were allowed in HTMl 4.01 
transitional, but that has been removed for HTML 5?

Lars Gunther



From lachlan.hunt at lachy.id.au  Sun Mar  4 21:52:15 2007
From: lachlan.hunt at lachy.id.au (Lachlan Hunt)
Date: Mon, 05 Mar 2007 16:52:15 +1100
Subject: [HTML5] [html] Elements within "title"?
In-Reply-To: <981168.64927.qm@web50109.mail.yahoo.com>
References: <981168.64927.qm@web50109.mail.yahoo.com>
Message-ID: <45EBB00F.7050806@lachy.id.au>

James Justin Harrell wrote:
> Why not create an attribute for the title element to allow markup?
> 
> <title allowmarkup="true">H<sub>2</sub>0</title>

In the HTML5 parsing algorithm, that would require the tree construction 
phase to set the content model flag in the tokeniser to RCDATA or PCDATA 
based on the presence of the attribute.  Although it might be possible 
to implement, no-one has yet given a valid use case that requires the 
use of markup in the title, is worth the cost of implementation and 
backwards imcompatibility, and has some real practical benefit for users 
and/or user agents.

Your example can be handled using:

   <title>H&#x2082;O</title>

(U+2082 is SUBSCRIPT TWO)

Or simply accept the small limitation and write:

   <title>H2O</title>

-- 
Lachlan Hunt
http://lachy.id.au/


From jens.meiert at erde3.com  Mon Mar  5 04:27:09 2007
From: jens.meiert at erde3.com (Jens Meiert)
Date: Mon, 05 Mar 2007 13:27:09 +0100
Subject: [HTML5] [html] Elements within "title"?
In-Reply-To: <Pine.GSO.4.64.0703050814500.12315@mustatilhi.cs.tut.fi>
References: <981168.64927.qm@web50109.mail.yahoo.com>
	<45EBB00F.7050806@lachy.id.au>
	<Pine.GSO.4.64.0703050814500.12315@mustatilhi.cs.tut.fi>
Message-ID: <20070305122709.204760@gmx.net>

> P.S. A use case for allowing markup:
> 
> <title>The famous formula <i>E</i> = 
> <i>m</i><i>c</i><sup>2</sup></title>
> 
> <title>Government of Canada Site |
>    <span lang="fr">Site du gouvernement du Canada</span></title>

That is great. Though, I don't think we need so many specific use cases since the argumentation must work on an abstract basis as well. The problem is a spec inconsistency, and almost every inline element that adds meaning to a document's content also eguals use cases within "title".

Am I right that Lachlan's ("allowmarkup") and Jukka's ("type") attribute proposals solve the compatibility issues anyway?

-- 
Jens Meiert
http://meiert.com/en/



From webmaster at keryx.se  Thu Mar  8 10:49:16 2007
From: webmaster at keryx.se (Keryx Web)
Date: Thu, 08 Mar 2007 19:49:16 +0100
Subject: [HTML5] xml:base
Message-ID: <45F05AAC.6080809@keryx.se>

Hi again on this list.

On the non-help list there has been a discussion if the <base> element
should be allowed in XHTML. Today the spec says that only the xml:base
attribute is allowed.

Since i got a few things wrong I am moving the discussion to this list.

These are my questions:

1. Is the xml:base attribute allowed in XHTML 1.0?

I thought so, but http://www.w3.org/TR/xhtml1/#strict says it is not in
the sense that if I use it my document is not "strictly conforming".
however one of the main points of using XML is the option of namespaces.
It may not be "strictly conforming" according to the DTD, but it is
strictly according to the philosophy of XML.

The spec also says that one *may* use it with other namespaces:

"3.1.2. Using XHTML with other namespaces

The XHTML namespace may be used with other XML namespaces as per
[XMLNS], although such documents are not strictly conforming XHTML 1.0
documents as defined above. "

To say that it is disallowed would be to strong, IMHO.

2. If I use the <base> element to specify one base url and the attribute
to specify another - admittedly bad design - which one should win over
the other? Testing in FFox the attribute does.

Simon Pieters kindly answered that one could use the together, i.e.
specify base url with the element and a path with the attribute. However
that was not what I was asking about.


Best regards

Lars Gunther




From zcorpan at gmail.com  Thu Mar  8 11:36:52 2007
From: zcorpan at gmail.com (Simon Pieters)
Date: Thu, 08 Mar 2007 20:36:52 +0100
Subject: [HTML5] xml:base
In-Reply-To: <45F05AAC.6080809@keryx.se>
References: <45F05AAC.6080809@keryx.se>
Message-ID: <op.tovx3qzh7a8kvn@hp-a0a83fcd39d2>

On Thu, 08 Mar 2007 19:49:16 +0100, Keryx Web <webmaster at keryx.se> wrote:

> 1. Is the xml:base attribute allowed in XHTML 1.0?
>
> I thought so, but http://www.w3.org/TR/xhtml1/#strict says it is not in
> the sense that if I use it my document is not "strictly conforming".
> however one of the main points of using XML is the option of namespaces.
> It may not be "strictly conforming" according to the DTD, but it is
> strictly according to the philosophy of XML.
>
> The spec also says that one *may* use it with other namespaces:
>
> "3.1.2. Using XHTML with other namespaces
>
> The XHTML namespace may be used with other XML namespaces as per
> [XMLNS], although such documents are not strictly conforming XHTML 1.0
> documents as defined above. "

I think this could be interpreted in two ways.

(1) The XHTML namespace may be used with other namespaces, but then you  
will only conform to Namespaces in XML, not XHTML 1.0.

(2) There is an implied concept of loosely conforming XHTML 1.0 documents.

> 2. If I use the <base> element to specify one base url and the attribute
> to specify another - admittedly bad design - which one should win over
> the other? Testing in FFox the attribute does.

They don't conflict.

<base> specifies the document's base URI. Try  
javascript:alert(document.baseURI).

xml:base specifies the element's base URI. Try  
javascript:alert(document.documentElement.baseURI).

> Simon Pieters kindly answered that one could use the together, i.e.
> specify base url with the element and a path with the attribute. However
> that was not what I was asking about.

The use of relative URIs was just to illustrate in what order they are  
applied. If you specify a full URI with xml:base then it won't be resolved  
against the parent node's base URI.

HTH,
-- 
Simon Pieters


From lachlan.hunt at lachy.id.au  Thu Mar  8 17:41:05 2007
From: lachlan.hunt at lachy.id.au (Lachlan Hunt)
Date: Fri, 09 Mar 2007 12:41:05 +1100
Subject: [HTML5] [html] Elements within "title"?
In-Reply-To: <op.towcl0zvsmjzpq@acer3010.parc.com>
References: <20070227085415.317650@gmx.net> <45E3FE7B.5070902@lachy.id.au>
	<op.towcl0zvsmjzpq@acer3010.parc.com>
Message-ID: <45F0BB31.5010809@lachy.id.au>

Steven Pemberton wrote:
> The title element gives metadata about the document. In XHTML2's unified 
> treatment of metadata, the title element is considered as a shorthand for
> 
>     <meta property="title">Your title text here</meta>
> 
> But the meta element *can* contain other elements, so if you need to 
> have a title with elements, then that is the place to put it:

If that's the case, then why doesn't XHTML2 just make the content model 
of <title> and <meta> the same?  Why confuse authors by requiring that 
they use a completely different element if they want to include markup 
within it?

Where in the XHTML2 spec does it actually define the UA conformance 
requirements for treating <title> as shorthand for that <meta> element, 
or vice versa?

It states [1]:

| The title of a document is metadata about the document, and so a title
| like <title>About W3C</title> is equivalent to <meta about=""
| property="title">About W3C</meta>.

But neither that statement, nor any others I could find, include any 
conformance criteria.

Which takes precedence if both are specified, and they differ?  In fact, 
which takes precedence if an author inclueds 2 or more title elements?

Besides, allowing markup in titles at all seems to ignore the fact that 
most uses for title are limited to plain text, and thus not useful.

[1] http://www.w3.org/TR/xhtml2/mod-document.html#sec_7.3.

-- 
Lachlan Hunt
http://lachy.id.au/


From ian at hixie.ch  Thu Mar 15 07:53:04 2007
From: ian at hixie.ch (Ian Hickson)
Date: Thu, 15 Mar 2007 14:53:04 +0000 (UTC)
Subject: [HTML5] [whatwg] Scripting and WF2 repetition model?
In-Reply-To: <Pine.LNX.4.64.0703150929010.5334@localhost>
References: <Pine.LNX.4.64.0703141842180.5334@localhost>
	<Pine.LNX.4.62.0703141855240.15496@dhalsim.dreamhost.com>
	<Pine.LNX.4.64.0703150929010.5334@localhost>
Message-ID: <Pine.LNX.4.62.0703151446430.26307@dhalsim.dreamhost.com>

On Thu, 15 Mar 2007, Dave Raggett wrote:
> 
> Thanks for the helpful reply. Unfortunately, Opera 9 (Linux) reports a 
> type error on calling getElementsByName("row1.quantity").

Oh, try invoking that on the document instead of the element, that should 
do it. My bad.


> By using alert and innerHTML, I see that row1 has the following content:
> 
> <TD><INPUT name="row1.item" type="text" title="product name"></TD>
> <TD><INPUT name="row1.quantity" type="number" title="number purchased"></TD>
> <TD><INPUT name="row1.unitprice" type="number" title="price per unit"></TD>
> 
> A work around would be to use getElementsByTagName on the input elements 
> for the row, as the position of the fields of interest within the row is 
> known in advance. That works when it comes to getting the requisite 
> fields for quantity and unitprice, but mysteriously their value property 
> is always blank despite evidence to the contrary on the browser 
> rendering of the page.

Sounds like a bug in Opera.


> I have to admit to a certain amount of puzzlement over the WF2 choice of 
> name syntax for repeating fields. It would have been a lot simpler to 
> avoid the period, so that the name was consistent with ECMAScript, HTML4 
> and DOM1.

There is no name syntax for repeating fields in WF2. You can use whatever 
syntax you like. All that's going on is that the template name is being 
prefilled. For example, you could do:

   <input name="row-number-[row]" value="this is the row number #[row]">

...and it'll still do the substitution.

I agree that better JS shortcuts could make life easier. It's not clear 
how to provide those while still providing real DOM support. (XForms, 
e.g., gets around this by not providing real DOM support.)


> It also seems that DOM1 only defines getElementsByName on the document 
> object and not on every element. So it is not a big suprise when it 
> didn't work in Opera on the tr (row) element.

Indeed. Change it to the document and my script should work, in theory, 
notwithstanding any bugs.

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


From jens.meiert at erde3.com  Thu Mar  1 02:53:05 2007
From: jens.meiert at erde3.com (Jens Meiert)
Date: Thu, 01 Mar 2007 11:53:05 +0100
Subject: [HTML5] [html] Elements within "title"?
In-Reply-To: <004c01c75b9b$c9dd23c0$2f01a8c0@Seka>
References: <004c01c75b9b$c9dd23c0$2f01a8c0@Seka>
Message-ID: <20070301105305.255300@gmx.net>

> From my experience, the title is normally rendered in places that
> have very limited constraints. Does the <title> element really need
> to be styled?

It's not about styling - and this element most likely doesn't need to be styled at all - but rather about marking up what we mark up everywhere else in documents.

So does it really make sense to correctly mark up abbreviations in documents except in one of their probably most important elements? Likely there are historical reasons, certainly there are compatibility issues, but surely that is inconsistent anyway.

> I think it was a valid suggestion and a good discussion, but I don't
> see any positive effects from an actual implementation of this.

Maybe this change would only make the spec more consistent and all of us slightly happier, but it won't be "cheap", presumably ...

-- 
Jens Meiert
http://meiert.com/en/



From jens.meiert at erde3.com  Thu Mar  1 07:15:29 2007
From: jens.meiert at erde3.com (Jens Meiert)
Date: Thu, 01 Mar 2007 16:15:29 +0100
Subject: [HTML5] [html] Elements within "title"?
In-Reply-To: <abd6c8010703010707m6f3322e9s6e114e17c0e306b2@mail.gmail.com>
References: <20070227085415.317650@gmx.net> <45E3FE7B.5070902@lachy.id.au>	
	<Pine.GSO.4.64.0702271325580.5105@hopeatilhi.cs.tut.fi>	
	<433C4B7E-A113-4A5B-8354-D39F352C265D@iki.fi>	
	<Pine.GSO.4.64.0702271627580.4057@hopeatilhi.cs.tut.fi>	
	<45E59DCC.4080304@lachy.id.au>	
	<50FCB9CB-A8C9-4427-8796-E2B77EA21259@nickshanks.com>	
	<49C257E2C13F584790B2E302E021B6F912A0C694@winse-msg-01.segroup.winse.corp.microsoft.com>
	<45E6C179.2050306@sebastianmendel.de>
	<abd6c8010703010707m6f3322e9s6e114e17c0e306b2@mail.gmail.com>
Message-ID: <20070301151529.269910@gmx.net>

> There's nothing that prevents you from saying the content model of
> title is limited to:
> 
> sup, sub, em, strong, span, abbr (others I'm forgetting).

Right, and that's basically what is proposed [1] (albeit not with a "comprehensive" content model yet).

I understand that there's basic consensus on this issue, but the question is if that problem/inconsistency could be addressed in any future (X)HTML spec? (By the way, this somewhat tempts me to cross-post all the time.)


[1] http://meiert.com/en/blog/20070226/html-semantics-of-title-element-content/

-- 
Jens Meiert
http://meiert.com/en/



From jens.meiert at erde3.com  Thu Mar  1 07:15:29 2007
From: jens.meiert at erde3.com (Jens Meiert)
Date: Thu, 01 Mar 2007 16:15:29 +0100
Subject: [HTML5] [html] Elements within "title"?
In-Reply-To: <abd6c8010703010707m6f3322e9s6e114e17c0e306b2@mail.gmail.com>
References: <20070227085415.317650@gmx.net> <45E3FE7B.5070902@lachy.id.au>	
	<Pine.GSO.4.64.0702271325580.5105@hopeatilhi.cs.tut.fi>	
	<433C4B7E-A113-4A5B-8354-D39F352C265D@iki.fi>	
	<Pine.GSO.4.64.0702271627580.4057@hopeatilhi.cs.tut.fi>	
	<45E59DCC.4080304@lachy.id.au>	
	<50FCB9CB-A8C9-4427-8796-E2B77EA21259@nickshanks.com>	
	<49C257E2C13F584790B2E302E021B6F912A0C694@winse-msg-01.segroup.winse.corp.microsoft.com>
	<45E6C179.2050306@sebastianmendel.de>
	<abd6c8010703010707m6f3322e9s6e114e17c0e306b2@mail.gmail.com>
Message-ID: <20070301151529.269910@gmx.net>


> There's nothing that prevents you from saying the content model of
> title is limited to:
> 
> sup, sub, em, strong, span, abbr (others I'm forgetting).

Right, and that's basically what is proposed [1] (albeit not with a "comprehensive" content model yet).

I understand that there's basic consensus on this issue, but the question is if that problem/inconsistency could be addressed in any future (X)HTML spec? (By the way, this somewhat tempts me to cross-post all the time.)


[1] http://meiert.com/en/blog/20070226/html-semantics-of-title-element-content/

-- 
Jens Meiert
http://meiert.com/en/




From webmaster at keryx.se  Sun Mar  4 12:30:48 2007
From: webmaster at keryx.se (Keryx Web)
Date: Sun, 04 Mar 2007 21:30:48 +0100
Subject: [HTML5] List of discontinued attributes
Message-ID: <45EB2C78.3030100@keryx.se>

Hi

Has anyone made a list of attributes that were allowed in HTMl 4.01 
transitional, but that has been removed for HTML 5?

Lars Gunther



From lachlan.hunt at lachy.id.au  Sun Mar  4 21:52:15 2007
From: lachlan.hunt at lachy.id.au (Lachlan Hunt)
Date: Mon, 05 Mar 2007 16:52:15 +1100
Subject: [HTML5] [html] Elements within "title"?
In-Reply-To: <981168.64927.qm@web50109.mail.yahoo.com>
References: <981168.64927.qm@web50109.mail.yahoo.com>
Message-ID: <45EBB00F.7050806@lachy.id.au>

James Justin Harrell wrote:
> Why not create an attribute for the title element to allow markup?
> 
> <title allowmarkup="true">H<sub>2</sub>0</title>

In the HTML5 parsing algorithm, that would require the tree construction 
phase to set the content model flag in the tokeniser to RCDATA or PCDATA 
based on the presence of the attribute.  Although it might be possible 
to implement, no-one has yet given a valid use case that requires the 
use of markup in the title, is worth the cost of implementation and 
backwards imcompatibility, and has some real practical benefit for users 
and/or user agents.

Your example can be handled using:

   <title>H&#x2082;O</title>

(U+2082 is SUBSCRIPT TWO)

Or simply accept the small limitation and write:

   <title>H2O</title>

-- 
Lachlan Hunt
http://lachy.id.au/


From jens.meiert at erde3.com  Mon Mar  5 04:27:09 2007
From: jens.meiert at erde3.com (Jens Meiert)
Date: Mon, 05 Mar 2007 13:27:09 +0100
Subject: [HTML5] [html] Elements within "title"?
In-Reply-To: <Pine.GSO.4.64.0703050814500.12315@mustatilhi.cs.tut.fi>
References: <981168.64927.qm@web50109.mail.yahoo.com>
	<45EBB00F.7050806@lachy.id.au>
	<Pine.GSO.4.64.0703050814500.12315@mustatilhi.cs.tut.fi>
Message-ID: <20070305122709.204760@gmx.net>

> P.S. A use case for allowing markup:
> 
> <title>The famous formula <i>E</i> = 
> <i>m</i><i>c</i><sup>2</sup></title>
> 
> <title>Government of Canada Site |
>    <span lang="fr">Site du gouvernement du Canada</span></title>

That is great. Though, I don't think we need so many specific use cases since the argumentation must work on an abstract basis as well. The problem is a spec inconsistency, and almost every inline element that adds meaning to a document's content also eguals use cases within "title".

Am I right that Lachlan's ("allowmarkup") and Jukka's ("type") attribute proposals solve the compatibility issues anyway?

-- 
Jens Meiert
http://meiert.com/en/



From webmaster at keryx.se  Thu Mar  8 10:49:16 2007
From: webmaster at keryx.se (Keryx Web)
Date: Thu, 08 Mar 2007 19:49:16 +0100
Subject: [HTML5] xml:base
Message-ID: <45F05AAC.6080809@keryx.se>

Hi again on this list.

On the non-help list there has been a discussion if the <base> element
should be allowed in XHTML. Today the spec says that only the xml:base
attribute is allowed.

Since i got a few things wrong I am moving the discussion to this list.

These are my questions:

1. Is the xml:base attribute allowed in XHTML 1.0?

I thought so, but http://www.w3.org/TR/xhtml1/#strict says it is not in
the sense that if I use it my document is not "strictly conforming".
however one of the main points of using XML is the option of namespaces.
It may not be "strictly conforming" according to the DTD, but it is
strictly according to the philosophy of XML.

The spec also says that one *may* use it with other namespaces:

"3.1.2. Using XHTML with other namespaces

The XHTML namespace may be used with other XML namespaces as per
[XMLNS], although such documents are not strictly conforming XHTML 1.0
documents as defined above. "

To say that it is disallowed would be to strong, IMHO.

2. If I use the <base> element to specify one base url and the attribute
to specify another - admittedly bad design - which one should win over
the other? Testing in FFox the attribute does.

Simon Pieters kindly answered that one could use the together, i.e.
specify base url with the element and a path with the attribute. However
that was not what I was asking about.


Best regards

Lars Gunther




From zcorpan at gmail.com  Thu Mar  8 11:36:52 2007
From: zcorpan at gmail.com (Simon Pieters)
Date: Thu, 08 Mar 2007 20:36:52 +0100
Subject: [HTML5] xml:base
In-Reply-To: <45F05AAC.6080809@keryx.se>
References: <45F05AAC.6080809@keryx.se>
Message-ID: <op.tovx3qzh7a8kvn@hp-a0a83fcd39d2>

On Thu, 08 Mar 2007 19:49:16 +0100, Keryx Web <webmaster at keryx.se> wrote:

> 1. Is the xml:base attribute allowed in XHTML 1.0?
>
> I thought so, but http://www.w3.org/TR/xhtml1/#strict says it is not in
> the sense that if I use it my document is not "strictly conforming".
> however one of the main points of using XML is the option of namespaces.
> It may not be "strictly conforming" according to the DTD, but it is
> strictly according to the philosophy of XML.
>
> The spec also says that one *may* use it with other namespaces:
>
> "3.1.2. Using XHTML with other namespaces
>
> The XHTML namespace may be used with other XML namespaces as per
> [XMLNS], although such documents are not strictly conforming XHTML 1.0
> documents as defined above. "

I think this could be interpreted in two ways.

(1) The XHTML namespace may be used with other namespaces, but then you  
will only conform to Namespaces in XML, not XHTML 1.0.

(2) There is an implied concept of loosely conforming XHTML 1.0 documents.

> 2. If I use the <base> element to specify one base url and the attribute
> to specify another - admittedly bad design - which one should win over
> the other? Testing in FFox the attribute does.

They don't conflict.

<base> specifies the document's base URI. Try  
javascript:alert(document.baseURI).

xml:base specifies the element's base URI. Try  
javascript:alert(document.documentElement.baseURI).

> Simon Pieters kindly answered that one could use the together, i.e.
> specify base url with the element and a path with the attribute. However
> that was not what I was asking about.

The use of relative URIs was just to illustrate in what order they are  
applied. If you specify a full URI with xml:base then it won't be resolved  
against the parent node's base URI.

HTH,
-- 
Simon Pieters


From lachlan.hunt at lachy.id.au  Thu Mar  8 17:41:05 2007
From: lachlan.hunt at lachy.id.au (Lachlan Hunt)
Date: Fri, 09 Mar 2007 12:41:05 +1100
Subject: [HTML5] [html] Elements within "title"?
In-Reply-To: <op.towcl0zvsmjzpq@acer3010.parc.com>
References: <20070227085415.317650@gmx.net> <45E3FE7B.5070902@lachy.id.au>
	<op.towcl0zvsmjzpq@acer3010.parc.com>
Message-ID: <45F0BB31.5010809@lachy.id.au>

Steven Pemberton wrote:
> The title element gives metadata about the document. In XHTML2's unified 
> treatment of metadata, the title element is considered as a shorthand for
> 
>     <meta property="title">Your title text here</meta>
> 
> But the meta element *can* contain other elements, so if you need to 
> have a title with elements, then that is the place to put it:

If that's the case, then why doesn't XHTML2 just make the content model 
of <title> and <meta> the same?  Why confuse authors by requiring that 
they use a completely different element if they want to include markup 
within it?

Where in the XHTML2 spec does it actually define the UA conformance 
requirements for treating <title> as shorthand for that <meta> element, 
or vice versa?

It states [1]:

| The title of a document is metadata about the document, and so a title
| like <title>About W3C</title> is equivalent to <meta about=""
| property="title">About W3C</meta>.

But neither that statement, nor any others I could find, include any 
conformance criteria.

Which takes precedence if both are specified, and they differ?  In fact, 
which takes precedence if an author inclueds 2 or more title elements?

Besides, allowing markup in titles at all seems to ignore the fact that 
most uses for title are limited to plain text, and thus not useful.

[1] http://www.w3.org/TR/xhtml2/mod-document.html#sec_7.3.

-- 
Lachlan Hunt
http://lachy.id.au/


From ian at hixie.ch  Thu Mar 15 07:53:04 2007
From: ian at hixie.ch (Ian Hickson)
Date: Thu, 15 Mar 2007 14:53:04 +0000 (UTC)
Subject: [HTML5] [whatwg] Scripting and WF2 repetition model?
In-Reply-To: <Pine.LNX.4.64.0703150929010.5334@localhost>
References: <Pine.LNX.4.64.0703141842180.5334@localhost>
	<Pine.LNX.4.62.0703141855240.15496@dhalsim.dreamhost.com>
	<Pine.LNX.4.64.0703150929010.5334@localhost>
Message-ID: <Pine.LNX.4.62.0703151446430.26307@dhalsim.dreamhost.com>

On Thu, 15 Mar 2007, Dave Raggett wrote:
> 
> Thanks for the helpful reply. Unfortunately, Opera 9 (Linux) reports a 
> type error on calling getElementsByName("row1.quantity").

Oh, try invoking that on the document instead of the element, that should 
do it. My bad.


> By using alert and innerHTML, I see that row1 has the following content:
> 
> <TD><INPUT name="row1.item" type="text" title="product name"></TD>
> <TD><INPUT name="row1.quantity" type="number" title="number purchased"></TD>
> <TD><INPUT name="row1.unitprice" type="number" title="price per unit"></TD>
> 
> A work around would be to use getElementsByTagName on the input elements 
> for the row, as the position of the fields of interest within the row is 
> known in advance. That works when it comes to getting the requisite 
> fields for quantity and unitprice, but mysteriously their value property 
> is always blank despite evidence to the contrary on the browser 
> rendering of the page.

Sounds like a bug in Opera.


> I have to admit to a certain amount of puzzlement over the WF2 choice of 
> name syntax for repeating fields. It would have been a lot simpler to 
> avoid the period, so that the name was consistent with ECMAScript, HTML4 
> and DOM1.

There is no name syntax for repeating fields in WF2. You can use whatever 
syntax you like. All that's going on is that the template name is being 
prefilled. For example, you could do:

   <input name="row-number-[row]" value="this is the row number #[row]">

...and it'll still do the substitution.

I agree that better JS shortcuts could make life easier. It's not clear 
how to provide those while still providing real DOM support. (XForms, 
e.g., gets around this by not providing real DOM support.)


> It also seems that DOM1 only defines getElementsByName on the document 
> object and not on every element. So it is not a big suprise when it 
> didn't work in Opera on the tr (row) element.

Indeed. Change it to the document and my script should work, in theory, 
notwithstanding any bugs.

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


From jens.meiert at erde3.com  Thu Mar  1 02:53:05 2007
From: jens.meiert at erde3.com (Jens Meiert)
Date: Thu, 01 Mar 2007 11:53:05 +0100
Subject: [HTML5] [html] Elements within "title"?
In-Reply-To: <004c01c75b9b$c9dd23c0$2f01a8c0@Seka>
References: <004c01c75b9b$c9dd23c0$2f01a8c0@Seka>
Message-ID: <20070301105305.255300@gmx.net>

> From my experience, the title is normally rendered in places that
> have very limited constraints. Does the <title> element really need
> to be styled?

It's not about styling - and this element most likely doesn't need to be styled at all - but rather about marking up what we mark up everywhere else in documents.

So does it really make sense to correctly mark up abbreviations in documents except in one of their probably most important elements? Likely there are historical reasons, certainly there are compatibility issues, but surely that is inconsistent anyway.

> I think it was a valid suggestion and a good discussion, but I don't
> see any positive effects from an actual implementation of this.

Maybe this change would only make the spec more consistent and all of us slightly happier, but it won't be "cheap", presumably ...

-- 
Jens Meiert
http://meiert.com/en/



From jens.meiert at erde3.com  Thu Mar  1 07:15:29 2007
From: jens.meiert at erde3.com (Jens Meiert)
Date: Thu, 01 Mar 2007 16:15:29 +0100
Subject: [HTML5] [html] Elements within "title"?
In-Reply-To: <abd6c8010703010707m6f3322e9s6e114e17c0e306b2@mail.gmail.com>
References: <20070227085415.317650@gmx.net> <45E3FE7B.5070902@lachy.id.au>	
	<Pine.GSO.4.64.0702271325580.5105@hopeatilhi.cs.tut.fi>	
	<433C4B7E-A113-4A5B-8354-D39F352C265D@iki.fi>	
	<Pine.GSO.4.64.0702271627580.4057@hopeatilhi.cs.tut.fi>	
	<45E59DCC.4080304@lachy.id.au>	
	<50FCB9CB-A8C9-4427-8796-E2B77EA21259@nickshanks.com>	
	<49C257E2C13F584790B2E302E021B6F912A0C694@winse-msg-01.segroup.winse.corp.microsoft.com>
	<45E6C179.2050306@sebastianmendel.de>
	<abd6c8010703010707m6f3322e9s6e114e17c0e306b2@mail.gmail.com>
Message-ID: <20070301151529.269910@gmx.net>

> There's nothing that prevents you from saying the content model of
> title is limited to:
> 
> sup, sub, em, strong, span, abbr (others I'm forgetting).

Right, and that's basically what is proposed [1] (albeit not with a "comprehensive" content model yet).

I understand that there's basic consensus on this issue, but the question is if that problem/inconsistency could be addressed in any future (X)HTML spec? (By the way, this somewhat tempts me to cross-post all the time.)


[1] http://meiert.com/en/blog/20070226/html-semantics-of-title-element-content/

-- 
Jens Meiert
http://meiert.com/en/



From jens.meiert at erde3.com  Thu Mar  1 07:15:29 2007
From: jens.meiert at erde3.com (Jens Meiert)
Date: Thu, 01 Mar 2007 16:15:29 +0100
Subject: [HTML5] [html] Elements within "title"?
In-Reply-To: <abd6c8010703010707m6f3322e9s6e114e17c0e306b2@mail.gmail.com>
References: <20070227085415.317650@gmx.net> <45E3FE7B.5070902@lachy.id.au>	
	<Pine.GSO.4.64.0702271325580.5105@hopeatilhi.cs.tut.fi>	
	<433C4B7E-A113-4A5B-8354-D39F352C265D@iki.fi>	
	<Pine.GSO.4.64.0702271627580.4057@hopeatilhi.cs.tut.fi>	
	<45E59DCC.4080304@lachy.id.au>	
	<50FCB9CB-A8C9-4427-8796-E2B77EA21259@nickshanks.com>	
	<49C257E2C13F584790B2E302E021B6F912A0C694@winse-msg-01.segroup.winse.corp.microsoft.com>
	<45E6C179.2050306@sebastianmendel.de>
	<abd6c8010703010707m6f3322e9s6e114e17c0e306b2@mail.gmail.com>
Message-ID: <20070301151529.269910@gmx.net>


> There's nothing that prevents you from saying the content model of
> title is limited to:
> 
> sup, sub, em, strong, span, abbr (others I'm forgetting).

Right, and that's basically what is proposed [1] (albeit not with a "comprehensive" content model yet).

I understand that there's basic consensus on this issue, but the question is if that problem/inconsistency could be addressed in any future (X)HTML spec? (By the way, this somewhat tempts me to cross-post all the time.)


[1] http://meiert.com/en/blog/20070226/html-semantics-of-title-element-content/

-- 
Jens Meiert
http://meiert.com/en/




From webmaster at keryx.se  Sun Mar  4 12:30:48 2007
From: webmaster at keryx.se (Keryx Web)
Date: Sun, 04 Mar 2007 21:30:48 +0100
Subject: [HTML5] List of discontinued attributes
Message-ID: <45EB2C78.3030100@keryx.se>

Hi

Has anyone made a list of attributes that were allowed in HTMl 4.01 
transitional, but that has been removed for HTML 5?

Lars Gunther



From lachlan.hunt at lachy.id.au  Sun Mar  4 21:52:15 2007
From: lachlan.hunt at lachy.id.au (Lachlan Hunt)
Date: Mon, 05 Mar 2007 16:52:15 +1100
Subject: [HTML5] [html] Elements within "title"?
In-Reply-To: <981168.64927.qm@web50109.mail.yahoo.com>
References: <981168.64927.qm@web50109.mail.yahoo.com>
Message-ID: <45EBB00F.7050806@lachy.id.au>

James Justin Harrell wrote:
> Why not create an attribute for the title element to allow markup?
> 
> <title allowmarkup="true">H<sub>2</sub>0</title>

In the HTML5 parsing algorithm, that would require the tree construction 
phase to set the content model flag in the tokeniser to RCDATA or PCDATA 
based on the presence of the attribute.  Although it might be possible 
to implement, no-one has yet given a valid use case that requires the 
use of markup in the title, is worth the cost of implementation and 
backwards imcompatibility, and has some real practical benefit for users 
and/or user agents.

Your example can be handled using:

   <title>H&#x2082;O</title>

(U+2082 is SUBSCRIPT TWO)

Or simply accept the small limitation and write:

   <title>H2O</title>

-- 
Lachlan Hunt
http://lachy.id.au/


From jens.meiert at erde3.com  Mon Mar  5 04:27:09 2007
From: jens.meiert at erde3.com (Jens Meiert)
Date: Mon, 05 Mar 2007 13:27:09 +0100
Subject: [HTML5] [html] Elements within "title"?
In-Reply-To: <Pine.GSO.4.64.0703050814500.12315@mustatilhi.cs.tut.fi>
References: <981168.64927.qm@web50109.mail.yahoo.com>
	<45EBB00F.7050806@lachy.id.au>
	<Pine.GSO.4.64.0703050814500.12315@mustatilhi.cs.tut.fi>
Message-ID: <20070305122709.204760@gmx.net>

> P.S. A use case for allowing markup:
> 
> <title>The famous formula <i>E</i> = 
> <i>m</i><i>c</i><sup>2</sup></title>
> 
> <title>Government of Canada Site |
>    <span lang="fr">Site du gouvernement du Canada</span></title>

That is great. Though, I don't think we need so many specific use cases since the argumentation must work on an abstract basis as well. The problem is a spec inconsistency, and almost every inline element that adds meaning to a document's content also eguals use cases within "title".

Am I right that Lachlan's ("allowmarkup") and Jukka's ("type") attribute proposals solve the compatibility issues anyway?

-- 
Jens Meiert
http://meiert.com/en/



From webmaster at keryx.se  Thu Mar  8 10:49:16 2007
From: webmaster at keryx.se (Keryx Web)
Date: Thu, 08 Mar 2007 19:49:16 +0100
Subject: [HTML5] xml:base
Message-ID: <45F05AAC.6080809@keryx.se>

Hi again on this list.

On the non-help list there has been a discussion if the <base> element
should be allowed in XHTML. Today the spec says that only the xml:base
attribute is allowed.

Since i got a few things wrong I am moving the discussion to this list.

These are my questions:

1. Is the xml:base attribute allowed in XHTML 1.0?

I thought so, but http://www.w3.org/TR/xhtml1/#strict says it is not in
the sense that if I use it my document is not "strictly conforming".
however one of the main points of using XML is the option of namespaces.
It may not be "strictly conforming" according to the DTD, but it is
strictly according to the philosophy of XML.

The spec also says that one *may* use it with other namespaces:

"3.1.2. Using XHTML with other namespaces

The XHTML namespace may be used with other XML namespaces as per
[XMLNS], although such documents are not strictly conforming XHTML 1.0
documents as defined above. "

To say that it is disallowed would be to strong, IMHO.

2. If I use the <base> element to specify one base url and the attribute
to specify another - admittedly bad design - which one should win over
the other? Testing in FFox the attribute does.

Simon Pieters kindly answered that one could use the together, i.e.
specify base url with the element and a path with the attribute. However
that was not what I was asking about.


Best regards

Lars Gunther




From zcorpan at gmail.com  Thu Mar  8 11:36:52 2007
From: zcorpan at gmail.com (Simon Pieters)
Date: Thu, 08 Mar 2007 20:36:52 +0100
Subject: [HTML5] xml:base
In-Reply-To: <45F05AAC.6080809@keryx.se>
References: <45F05AAC.6080809@keryx.se>
Message-ID: <op.tovx3qzh7a8kvn@hp-a0a83fcd39d2>

On Thu, 08 Mar 2007 19:49:16 +0100, Keryx Web <webmaster at keryx.se> wrote:

> 1. Is the xml:base attribute allowed in XHTML 1.0?
>
> I thought so, but http://www.w3.org/TR/xhtml1/#strict says it is not in
> the sense that if I use it my document is not "strictly conforming".
> however one of the main points of using XML is the option of namespaces.
> It may not be "strictly conforming" according to the DTD, but it is
> strictly according to the philosophy of XML.
>
> The spec also says that one *may* use it with other namespaces:
>
> "3.1.2. Using XHTML with other namespaces
>
> The XHTML namespace may be used with other XML namespaces as per
> [XMLNS], although such documents are not strictly conforming XHTML 1.0
> documents as defined above. "

I think this could be interpreted in two ways.

(1) The XHTML namespace may be used with other namespaces, but then you  
will only conform to Namespaces in XML, not XHTML 1.0.

(2) There is an implied concept of loosely conforming XHTML 1.0 documents.

> 2. If I use the <base> element to specify one base url and the attribute
> to specify another - admittedly bad design - which one should win over
> the other? Testing in FFox the attribute does.

They don't conflict.

<base> specifies the document's base URI. Try  
javascript:alert(document.baseURI).

xml:base specifies the element's base URI. Try  
javascript:alert(document.documentElement.baseURI).

> Simon Pieters kindly answered that one could use the together, i.e.
> specify base url with the element and a path with the attribute. However
> that was not what I was asking about.

The use of relative URIs was just to illustrate in what order they are  
applied. If you specify a full URI with xml:base then it won't be resolved  
against the parent node's base URI.

HTH,
-- 
Simon Pieters


From lachlan.hunt at lachy.id.au  Thu Mar  8 17:41:05 2007
From: lachlan.hunt at lachy.id.au (Lachlan Hunt)
Date: Fri, 09 Mar 2007 12:41:05 +1100
Subject: [HTML5] [html] Elements within "title"?
In-Reply-To: <op.towcl0zvsmjzpq@acer3010.parc.com>
References: <20070227085415.317650@gmx.net> <45E3FE7B.5070902@lachy.id.au>
	<op.towcl0zvsmjzpq@acer3010.parc.com>
Message-ID: <45F0BB31.5010809@lachy.id.au>

Steven Pemberton wrote:
> The title element gives metadata about the document. In XHTML2's unified 
> treatment of metadata, the title element is considered as a shorthand for
> 
>     <meta property="title">Your title text here</meta>
> 
> But the meta element *can* contain other elements, so if you need to 
> have a title with elements, then that is the place to put it:

If that's the case, then why doesn't XHTML2 just make the content model 
of <title> and <meta> the same?  Why confuse authors by requiring that 
they use a completely different element if they want to include markup 
within it?

Where in the XHTML2 spec does it actually define the UA conformance 
requirements for treating <title> as shorthand for that <meta> element, 
or vice versa?

It states [1]:

| The title of a document is metadata about the document, and so a title
| like <title>About W3C</title> is equivalent to <meta about=""
| property="title">About W3C</meta>.

But neither that statement, nor any others I could find, include any 
conformance criteria.

Which takes precedence if both are specified, and they differ?  In fact, 
which takes precedence if an author inclueds 2 or more title elements?

Besides, allowing markup in titles at all seems to ignore the fact that 
most uses for title are limited to plain text, and thus not useful.

[1] http://www.w3.org/TR/xhtml2/mod-document.html#sec_7.3.

-- 
Lachlan Hunt
http://lachy.id.au/


From ian at hixie.ch  Thu Mar 15 07:53:04 2007
From: ian at hixie.ch (Ian Hickson)
Date: Thu, 15 Mar 2007 14:53:04 +0000 (UTC)
Subject: [HTML5] [whatwg] Scripting and WF2 repetition model?
In-Reply-To: <Pine.LNX.4.64.0703150929010.5334@localhost>
References: <Pine.LNX.4.64.0703141842180.5334@localhost>
	<Pine.LNX.4.62.0703141855240.15496@dhalsim.dreamhost.com>
	<Pine.LNX.4.64.0703150929010.5334@localhost>
Message-ID: <Pine.LNX.4.62.0703151446430.26307@dhalsim.dreamhost.com>

On Thu, 15 Mar 2007, Dave Raggett wrote:
> 
> Thanks for the helpful reply. Unfortunately, Opera 9 (Linux) reports a 
> type error on calling getElementsByName("row1.quantity").

Oh, try invoking that on the document instead of the element, that should 
do it. My bad.


> By using alert and innerHTML, I see that row1 has the following content:
> 
> <TD><INPUT name="row1.item" type="text" title="product name"></TD>
> <TD><INPUT name="row1.quantity" type="number" title="number purchased"></TD>
> <TD><INPUT name="row1.unitprice" type="number" title="price per unit"></TD>
> 
> A work around would be to use getElementsByTagName on the input elements 
> for the row, as the position of the fields of interest within the row is 
> known in advance. That works when it comes to getting the requisite 
> fields for quantity and unitprice, but mysteriously their value property 
> is always blank despite evidence to the contrary on the browser 
> rendering of the page.

Sounds like a bug in Opera.


> I have to admit to a certain amount of puzzlement over the WF2 choice of 
> name syntax for repeating fields. It would have been a lot simpler to 
> avoid the period, so that the name was consistent with ECMAScript, HTML4 
> and DOM1.

There is no name syntax for repeating fields in WF2. You can use whatever 
syntax you like. All that's going on is that the template name is being 
prefilled. For example, you could do:

   <input name="row-number-[row]" value="this is the row number #[row]">

...and it'll still do the substitution.

I agree that better JS shortcuts could make life easier. It's not clear 
how to provide those while still providing real DOM support. (XForms, 
e.g., gets around this by not providing real DOM support.)


> It also seems that DOM1 only defines getElementsByName on the document 
> object and not on every element. So it is not a big suprise when it 
> didn't work in Opera on the tr (row) element.

Indeed. Change it to the document and my script should work, in theory, 
notwithstanding any bugs.

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