Javascript and Vista

[Dr J R Stockton]

In comp.lang.javascript message <[email protected]>, Sun,

[...] Thomas 'PointedEars' Lahn [...] posted:

Since there is no working user agent that requires those pseudo-comments in
order to prevent the code from being displayed
...
you can safely remove them in old *HTML* documents and omit them in
new ones.

False logic. User agents are not the only thing which may be required
to process such documents.

You are mistaken. Apparently you don't know what a(n HTML) user agent is:

,-<http://www.w3.org/TR/REC-html40/conform.html#didx-user_agent>
|
| HTML user agent
| An HTML user agent is any device that interprets HTML documents.
| User agents include visual browsers (text-only and graphical),
| non-visual browsers (audio, Braille), search robots, proxies, etc.
|
| [...]

That covers users. But it is not only users and their agents that may
process documents.

Moreover, you did not, it seems, consider the definition, immediately
above that one, of "user"; it includes the word "rendered".

You should also consider, above that, the definition of "authoring
tool", noting that it does not match the normal English meaning of those
words. I refer you, for a non-computing example, to Asimov; you will
recognise the right part when you get there.

You need to learn to do better than to regurgitate incompletely-relevant
references; you need to learn to think in the manner of a normal human
being.

 

[Thomas 'PointedEars' Lahn]

[...] Thomas 'PointedEars' Lahn [...] posted:

[...] Thomas 'PointedEars' Lahn [...] posted:
Since there is no working user agent that requires those pseudo-comments in
order to prevent the code from being displayed
...
you can safely remove them in old *HTML* documents and omit them in
new ones.
False logic. User agents are not the only thing which may be required
to process such documents.

You are mistaken. Apparently you don't know what a(n HTML) user agent is:

,-<http://www.w3.org/TR/REC-html40/conform.html#didx-user_agent>
|
| HTML user agent
| An HTML user agent is any device that interprets HTML documents.
| User agents include visual browsers (text-only and graphical),
| non-visual browsers (audio, Braille), search robots, proxies, etc.
|
| [...]

That covers users.

No, it covers "any device that interprets HTML documents". The
specification, and therefore I, could not be more clear about that.


PointedEars

 

[The Magpie]

[snip]

I'm interested in this bit - I tend to comment freely so what is
the now-accepted way to add comments?

It was *never* accepted coding style to use markup comments as
script comments.

Yep - fair enough. I misunderstood your comment and do use comments
the right way. My apologies.

See for example: http://www.spartanicus.utvinternet.ie/no-xhtml.htm
http://hixie.ch/advocacy/xhtml

That would be fine except for the line that reads: "Unfortunately, IE6
does not support application/xhtml+xml (in fact, it does not support
XHTML at all)."

Which, of course, totally buggers things up (see below).

Quite the opposite is true.

http://www.w3.org/2000/07/8378/xhtml/media-types/results

Indeed - and reading that link I note that serving it as HTML works on
every browser except Amaya - which frankly is so unusual a browser I
have never even heard of it. I also note that is shows that serving it
as anything *other* than HTML does indeed bugger up a lot of browsers.

To be honest, that site only confirms that serving it as HTML is the
way to go at the moment - but I would love to see an alternative
provided it actually works!

You would be well-advised to use the Google Groups archives of this
newsgroup before posting; this has been discussed several times
already.

No I wouldn't. I have never used Google Groups and see no reason to
start now. I certainly do not approve of Google's archive and would
not dream of using it - indeed, you will notice that this and most
other messages I post are set to refuse archiving altogether.

 

[Thomas 'PointedEars' Lahn]

That would be fine except for the line that reads: "Unfortunately, IE6
does not support application/xhtml+xml (in fact, it does not support
XHTML at all)."

JFYI: The above are advisories *against* using XHTML.

Which, of course, totally buggers things up (see below).

In fact, they are buggered up long before:

Indeed - and reading that link I note that serving it as HTML works on
every browser except Amaya - which frankly is so unusual a browser I
have never even heard of it. I also note that is shows that serving it
as anything *other* than HTML does indeed bugger up a lot of browsers.

To be honest, that site only confirms that serving it as HTML is the
way to go at the moment -

XHTML can *not* be "served as HTML"; despite all (misguided) efforts of
XHTML 1.0 Appendix C, the two markup languages are inherently incompatible.
(The final opportunity I saw for making them compatible was busted shortly
ago when someone pointed out to me that elements with empty content model
MUST NOT have end tags in HTML.) Incidentally, XHTML Basic/1.1 and
XHTML+MathML SHOULD NOT be served as text/html; I think, I would hope, that
a lesson had been learned there.

The results only confirm that there are many user agents out there that
*may* have sufficient error-correction to display XHTML *1.0* documents
properly in that case anyway. However, I seriously doubt this is a Good
Thing in the long-term.

but I would love to see an alternative provided it actually works!

Here it comes: HTML 4.01. As the advisories point out, as UAs do not look
at the DOCTYPE declaration, when serving XHTML as text/html you lose all
benefits that XHTML has to offer, and even gain disadvantages as compared to
properly served markup.

No I wouldn't. I have never used Google Groups and see no reason to
start now. I certainly do not approve of Google's archive and would
not dream of using it - indeed, you will notice that this and most
other messages I post are set to refuse archiving altogether.

The point of my advice was not for you to use Google Groups (for posting)
but to get informed about previous discussions before you ask such obvious
questions. You may use your private Usenet archive or other archives for
that of course, but Google Groups' archive (which includes the former decent
DejaNews archive) would probably provide more material for research. And
JFTR: That is actually one of the small number of benefits over conventional
Usenet access that makes me not to dismiss it entirely. In that sense, I
have to disagree with your setting the X-No-Archive header: It does not hurt
Google at all, it hurts people who want to do extensive Usenet research but
can't afford to have their own Usenet archive. Solutions you may provide in
your postings will hardly be found.


PointedEars

 

[Dr J R Stockton]

In comp.lang.javascript message <[email protected]>, Mon,

[...] Thomas 'PointedEars' Lahn [...] posted:

Dr J R Stockton wrote:
[...] Thomas 'PointedEars' Lahn [...] posted:
Since there is no working user agent that requires those pseudo-
comments in
order to prevent the code from being displayed
...
you can safely remove them in old *HTML* documents and omit them in
new ones.
False logic. User agents are not the only thing which may be required
to process such documents.
You are mistaken. Apparently you don't know what a(n HTML) user agent is:

,-<http://www.w3.org/TR/REC-html40/conform.html#didx-user_agent>
|
| HTML user agent
| An HTML user agent is any device that interprets HTML documents.
| User agents include visual browsers (text-only and graphical),
| non-visual browsers (audio, Braille), search robots, proxies, etc.
|
| [...]

That covers users.

No, it covers "any device that interprets HTML documents". The
specification, and therefore I, could not be more clear about that.

But "interpret" is not the came as "process". Your understanding of
English is not as good as you like to think.

 

[The Magpie]

XHTML can *not* be "served as HTML"; despite all (misguided) efforts of
XHTML 1.0 Appendix C, the two markup languages are inherently incompatible.

I agree it is not preferrable, but it does appear to actually work.

(The final opportunity I saw for making them compatible was busted shortly
ago when someone pointed out to me that elements with empty content model
MUST NOT have end tags in HTML.)

True - and this would most certainly prevent HTML being served as
XHTML. However, serving XHTML as HTML appears to work fine with empty
content model tags that do have a closing tag.

I really do get puzzled by this issue and would very much like it to
be resolved - as I am sure we all do! It really is time we either had
a working version of XHTML that genuinely works *or* a decent update
of HTML 4.x that addresses the issues relevant to modern web pages.

The results only confirm that there are many user agents out there that
*may* have sufficient error-correction to display XHTML *1.0* documents
properly in that case anyway. However, I seriously doubt this is a Good
Thing in the long-term.
Agreed.

Here it comes: HTML 4.01. As the advisories point out, as UAs do not look
at the DOCTYPE declaration, when serving XHTML as text/html you lose all
benefits that XHTML has to offer, and even gain disadvantages as compared to
properly served markup.

I absolutely agree - though I use XHTML for development reasons rather
than anything else. It is simply easier to correct and validate. Like
yourself though, I admit that nine times out of ten I will start a
page by coding it in HTML 4.01 and convert to XHTML later - and often
back again to HTML 4.01 before I put the page live. Really, that's the
reason I'd love to see the current shambles sorted out.

 

[Thomas 'PointedEars' Lahn]

I agree it is not preferrable, but it does appear to actually work.

It *seems* to work. Jumping to conclusions is the main problem with this
approach. Some tests in a handful of user agents (if that was even done)
mean nothing.

True - and this would most certainly prevent HTML being served as
XHTML. However, serving XHTML as HTML appears to work fine with empty
content model tags that do have a closing tag.

As I said, that relies on *erroneous* *error-correction*. `<br />' equals

I really do get puzzled by this issue and would very much like it to
be resolved - as I am sure we all do! It really is time we either had
a working version of XHTML that genuinely works *or* a decent update
of HTML 4.x that addresses the issues relevant to modern web pages.

Agreed. For example, I would not mind if this particular SHORTTAG feature
was explicitly disabled in the next revision of HTML 4.01.

I absolutely agree - though I use XHTML for development reasons rather
than anything else. It is simply easier to correct and validate.

I doubt that. For example, Eclipse WST have a decent HTML editor that allow
you to build the document just by clicking, and show you almost instantly
where your markup is invalid.

Like yourself though, I admit that nine times out of ten I will start a
page by coding it in HTML 4.01 and convert to XHTML later - and often
back again to HTML 4.01 before I put the page live.

But I don't do that, it would be quite foolish.

Really, that's the reason I'd love to see the current shambles sorted out.

There really are no shambles once you made the informed design decision to
use every language where it is fitted for best. For me, XHTML does not come
in before I'm building a XHTML-specific test case, use MathML client-side,
or develop on Plone (my job requires that). Even SVG is no sufficient
reason to use XHTML, as SVG can be included as external resource in HTML
documents with the `object' element.


F'up2 comp.infosystems.www.authoring.html, where this is on-topic

PointedEars