468,458 Members | 1,858 Online
Bytes | Developer Community
New Post

Home Posts Topics Members FAQ

Post your question to a community of 468,458 developers. It's quick & easy.

HTML vs XHTML

A couple years ago it seemed like XHTML was the direction of most web
markup, a foregone conclusion. Now I return to the scene and I see many
here recommending that authors stick to HTML, albeit in strict mode.

What happened? Why the shift? What's the problem with XHTML that has
everyone saying it be avoided in most cases? Can I take it that this
group seems to think HTML5 is going to be the future direction of the
web rather than XHTML2?

I've been happily going along marking stuff up as XHTML 1.0 Strict for a
couple years now, so I was just looking for some feedback on this issue.
Thanks,
Tim
Apr 11 '06 #1
28 2338
Timothy Larson <th*********@cox.net> wrote:
A couple years ago it seemed like XHTML was the direction of most web
markup, a foregone conclusion. Now I return to the scene and I see many
here recommending that authors stick to HTML, albeit in strict mode.

What happened? Why the shift? What's the problem with XHTML that has
everyone saying it be avoided in most cases?
Please spare us yet another HTML vs XHTML thread, this issue has been
discussed ad nauseum here. Use the archives to find the arguments and
the urls.
Can I take it that this
group seems to think HTML5 is going to be the future direction of the
web rather than XHTML2?


No-one knows. Safari, Opera and Mozilla are supporting the WhatWG
projects and have started to implement bits of the Web forms 2.0 and Web
Applications 1.0 aka (X)HTML5 proposals. I'm not aware of any current
support for XHTML2 in UAs.

--
Spartanicus
Apr 11 '06 #2

Timothy Larson wrote:
A couple years ago it seemed like XHTML was the direction of most web
markup, a foregone conclusion.
Bad, inappropriately done XHTML.

Mea culpa.
Now I return to the scene and I see many
here recommending that authors stick to HTML, albeit in strict mode.
We learned how to do XHTML correctly. Then many people (although not
myself) decided to reject XHTML completely.

I use almost entirely XHTML for my internal processing, totally for my
mobile device output, and sometimes (as Appendix C) for general web
work. XHTML (C) _does_ work on the widespread web, but you have to
know what you're doing.
What happened? Why the shift?
We realised that "XHTML-not-in-XML" wasn't really much use until there
are XML-aware consumers that can actually benefit from it.
Can I take it that this group seems to think HTML5 is
going to be the future direction of the web
XHTML is problematic and still premature for general publishing.

HTML 5 however is totally bogus and an exercise in one person and their
acolytes throwing their toys out of the pram. Hixie is up on my
dartboard next to the photo of Winer.

I predict (and quote me embarassingly on this in a year or two) that
XHTML is just about set for widespread adoption. It won't be as
"pages", but it will be as page fragments delivered to AJAX clients.
These are finally the widespread clients that really understand XML,
even if their parent pages aren't doing it as XML.

We'll also see a return to browser sniffing. AJAX-based "thickish"
clients for the browsers that can, restricted function compatibility
pages in plain HTML for those that can't.
rather than XHTML2?


I'd like to see a move towards Flash rather than XHTML 2! Now _that_
really is a bogosity.

Apr 11 '06 #3
On Tue, 11 Apr 2006, Timothy Larson wrote:
A couple years ago it seemed like XHTML was the direction of most web
markup, a foregone conclusion. Now I return to the scene and I see many
here recommending that authors stick to HTML, albeit in strict mode.

I've been happily going along marking stuff up as XHTML 1.0 Strict for a
couple years now, so I was just looking for some feedback on this issue.


Anything XHTML 1 can do, HTML 4 can do better.
Well ... at least as well as. There's nothing to gain from XHTML 1.0 -
certainly not from XHTML 1.0 Transitional, which we find so often
on the web, mostly in the form of "div soup".

When you ask about XHTML 1.1, try Internet Exploder 6 on
http://www.unics.uni-hannover.de/nht...notation.xhtml
not to mention Google.

--
The 6th of June is Bill Gates Day.

Apr 11 '06 #4
On Tue, 11 Apr 2006, Andreas Prilop wrote:

[the usual about HTML versus XHTML. It really would be better if the
original poster would get up to speed by reading what's already been
said, and then ask questions from there, rather than giving the
impression of posing the question "cold". IMHO and YMMV, anyway.]
When you ask about XHTML 1.1, try Internet Exploder 6 on
http://www.unics.uni-hannover.de/nht...notation.xhtml
Ive got IE6 sort-of tamed for that now. It prompts me for whether I
want to save or open the file. When I choose "Open", IE fires-up
Mozilla to open a disk copy of the file:

file:///C:/Documents%20and%20Settings/myusername/Local%20Settings/Temporary%20Internet%20Files/Content.IE5/8LYNC9IB/ruby-annotation%5B1%5D.xhtml

I suppose it would be nicer if IE could tell my www browser which real
URL to open, instead of opening a local file copy. Fortunately, I
don't normally use IE as if it were a www browser (it's the right
thing to use for applying the frequent security fixes to MS software,
on the other hand).
not to mention Google.


I hope the hon Usenauts are going to understand that you're talking
about how Google indexes your cited XHTML page (not about how IE
accesses Google's page!).

cheers

--

Nothing good has ever come of putting physicists and computer people in
the same room. - Joseph DeMartino (seen on a.h.b-o-u)
Apr 11 '06 #5
On Tue, 11 Apr 2006, Alan J. Flavell wrote:
file:///C:/Documents%20and%20Settings/myusername/Local%20Settings/
Temporary%20Internet%20Files/Content.IE5/8LYNC9IB/ruby-annotation%5B1%5D.xhtml

^^^^^^^

What does this "ruby-annotation[1].xhtml" mean? Why "[1]"?

--
The 6th of June is Bill Gates Day.

Apr 11 '06 #6
In article <11**********************@i39g2000cwa.googlegroups .com>,
"Andy Dingley" <di*****@codesmiths.com> wrote:

[big snip]

I'd like to see a move towards Flash rather than XHTML 2! Now _that_
really is a bogosity.


Well, if it weren't for Flash, we wouldn't have

http://www.myoldmac.net/webse-e-flash.htm

(Just for fun, be sure to double-click on the Netscape 3.6 icon...)
Apr 11 '06 #7
On Tue, 11 Apr 2006, Andreas Prilop wrote:
On Tue, 11 Apr 2006, Alan J. Flavell wrote:
file:///C:/Documents%20and%20Settings/myusername/Local%20Settings/
Temporary%20Internet%20Files/Content.IE5/8LYNC9IB/ruby-annotation%5B1%5D.xhtml ^^^^^^^

What does this "ruby-annotation[1].xhtml" mean?


It's what IE6 used as a temporary file name. Does it need to "mean"
anything? This is IE we're talking about, remember :-} The one that
makes up rules (even when in violation of mandatory requirements of
the interworking specifications) as it goes along, and changes them
whenever some aspect of their security loopholes becomes too obvious.
Why "[1]"?


Because it was the first temporary file with that name, I guess?

Apr 11 '06 #8
David Stone <no******@domain.invalid> wrote:
In article <11**********************@i39g2000cwa.googlegroups .com>,
"Andy Dingley" <di*****@codesmiths.com> wrote:

[big snip]

I'd like to see a move towards Flash rather than XHTML 2! Now _that_
really is a bogosity.


Well, if it weren't for Flash, we wouldn't have

http://www.myoldmac.net/webse-e-flash.htm

(Just for fun, be sure to double-click on the Netscape 3.6 icon...)


Now that's Art.
--
Andy Templeman <http://www.templeman.org.uk/>
Apr 11 '06 #9
In article <11**********************@i39g2000cwa.googlegroups .com>,
"Andy Dingley" <di*****@codesmiths.com> wrote:
HTML 5 however is totally bogus and an exercise in one person and their
acolytes throwing their toys out of the pram. Hixie is up on my
dartboard next to the photo of Winer.


That's harsh. Why do you think HTML5 is "totally bogus"?

I guess I am one of the acolytes. At least I learned a new word. :-)

--
Henri Sivonen
hs******@iki.fi
http://hsivonen.iki.fi/
Mozilla Web Author FAQ: http://mozilla.org/docs/web-developer/faq.html
Apr 11 '06 #10
Spartanicus wrote:
Please spare us yet another HTML vs XHTML thread, this issue has been
discussed ad nauseum here. Use the archives to find the arguments and
the urls.


I don't want to start a flamewar. If you happen to recall whenabouts
one of the more interesting/informative discussions occurred, that would
help me wade through all the threads.

Thanks,
Tim
Apr 12 '06 #11
Gazing into my crystal ball I observed David Stone
<no******@domain.invalid> writing in news:no.email-F3AA86.12462811042006
@news1.chem.utoronto.ca:
Just for fun, be sure to double-click on the Netscape 3.6 icon...


Hey, I still have Netscape 3.0 - It's kind of refreshing to what an old war
horse does with new stuff.

--
Adrienne Boswell
Please respond to the group so others can share
http://www.cavalcade-of-coding.info
Apr 12 '06 #12
In article <Xn****************************@69.28.186.121>,
Adrienne Boswell <ar****@yahoo.com> wrote:
Hey, I still have Netscape 3.0 - It's kind of refreshing to what an old war
horse does with new stuff.


I still have Netscape 1.0 on original floppy for Mac, although I am sure
bitrot has destroyed it. I even paid for it and have the goofy 'Netscape
Navigator Reference Guide' which came with it, all in a silver tri-fold
Netscape jacket. I also have 'A Brief Guide to Mosaic Macintosh Edition'
by Michael Fraase, but I have no idea why I have that; although I did
use Mosaic, I didn't buy it. I just put the Mosaic stuff in the silver
jacket with Netscape for safekeeping and enlightenment for future
generations.
Thanks for reminding me of what I have and to search for and find it
again. I generally tend to throw stuff away. I also have an invoice for
NS3 for 49 bucks and a one year Navigator license for $17 dated in
August 1996. By golly, that's going in the silver folder too.

leo

--
<http://web0.greatbasin.net/~leo/>
Apr 12 '06 #13
Timothy Larson <th*********@cox.net> wrote:
Please spare us yet another HTML vs XHTML thread, this issue has been
discussed ad nauseum here. Use the archives to find the arguments and
the urls.


I don't want to start a flamewar. If you happen to recall whenabouts
one of the more interesting/informative discussions occurred, that would
help me wade through all the threads.


It takes a few minutes to browse past threads, spend them.

--
Spartanicus
Apr 12 '06 #14
Spartanicus wrote:
Timothy Larson <th*********@cox.net> wrote:
If you happen to recall whenabouts
one of the more interesting/informative discussions occurred, that would
help me wade through all the threads.


It takes a few minutes to browse past threads, spend them.


If you can't be helpful in response to questions, don't bother to post
replies. Your post certainly didn't do me any good - how did it help you?
Apr 12 '06 #15
Timothy Larson <th*********@cox.net> wrote:
If you happen to recall whenabouts
one of the more interesting/informative discussions occurred, that would
help me wade through all the threads.


It takes a few minutes to browse past threads, spend them.


If you can't be helpful in response to questions, don't bother to post
replies. Your post certainly didn't do me any good - how did it help you?


This is not a help desk, it's a discussion group.

To function properly it is important that people do not restart
discussions that have been thrashed to death, they should find the
answer in the archives.

--
Spartanicus
Apr 12 '06 #16
Andreas Prilop wrote:
What does this "ruby-annotation[1].xhtml" mean? Why "[1]"?


If MSIE downloaded another file with the filename
"ruby-annotation.xhtml", that would then be saved as
"ruby-annotation[2].xhtml". That's all there is to it: avoiding file
name conflicts.

--
Bart.
Apr 13 '06 #17
Henri Sivonen wrote:
That's harsh. Why do you think HTML5 is "totally bogus"?

I guess I am one of the acolytes. At least I learned a new word. :-)


What do you like of HTML5 compared to XHTML?
Apr 13 '06 #18
Spartanicus wrote:
This is not a help desk, it's a discussion group.

To function properly it is important that people do not restart
discussions that have been thrashed to death, they should find the
answer in the archives.


I've been on Usenet since 1992. I don't need your opinion to inform me
what a newsgroup is or is not, or how it should or shouldn't be used to
realize its potential. Maybe you weren't around back in those days, or
maybe years of the "new" internet have made you jaded.

It was commonly accepted then that technical NGs such as this one were
places that people could use as a resource to get solutions. This isn't
alt.fan.ricky-martin or something, where we can "discuss" back and forth
with no goal or objective in mind other than the discussion itself.
We're not here solely to discuss the vagaries and philosophies of markup
design without any practical application. Most of us are here because
at one time or another we've run into real-life problems and need
practical solutions. We ask for advice at those times, and those who
have been around a little longer and gotten help in the past themselves
try to offer suggestions. Ofttimes those suggestions amount to "educate
yourself, resources at <url>" but that's OK.

So trying to tell me that I shouldn't be able to expect a constructive
answer to a question is a joke. If you can't further a discussion by
adding to it constructively, don't say anything at all. _That's_ what a
discussion group is - people having meaningful and relevant discourse.
If all you can add is "go search the archives" then close this group
down and replace it with a pointer to Google. News hosts don't maintain
an unlimited history of groups' postings, so it's not always easy to go
back. Maybe a group has an archive, or maybe not. Even now that Google
seems to be fulfilling that role nigh-universally, it's still not a
panacea because searching doesn't guarantee you'll necessarily find the
best resource - in a group like this, there are probably thousands of
posts that mention "html" and "xhtml". I was duly diligent in searching
recent discussions, as appropriate for NG netiquette - my original
question itself alluded to the fact that I'd heard of the issue (which I
couldn't have if I hadn't been) but not the reasoning behind it.

With this in mind, I hardly thinks it's inappropriate to ask the NG at
large to point me in the right direction. If you don't want to help me,
fine, no one's forcing you to spend your valuable time on me. I
certainly don't have the expectation that I'll get an answer, though I
am hopeful. But why spend your time to give a non-answer? That only
wastes everyone's time. If you don't have an answer, don't reply!

Yeah, I could take the time to wade through thousands of posts - I know
how to search, and can certainly use those skills here. Pointing me to
a particular thread (or answering any question from anyone) doesn't do
you any immediate good, that is true. But at some point I may have a
suggestion to contribute toward a dilemma of yours, and that's what
makes NGs like this work. It's a community of knowledgable people
willing to contribute for mutual benefit and edification.

Given that searching can be difficult and NG histories are often
limited, and given that our time is precious and no one likes to have it
wasted in rehashing material, I am more than happy to accept a pointer
to a reference - specifications, documentation, past threads - as an
answer. However, archiving your non-answer is a waste of resources and
contributes absolutely nothing to the collection of knowledge.
Archiving a pointer to existing information results in more people
finding what they need more quickly when they do searches in the future.

Apologies for the long OT rant.

Tim
Apr 13 '06 #19
To further the education of mankind, Timothy Larson <th*********@cox.net>
declaimed:
With this in mind, I hardly thinks it's inappropriate to ask the NG at
large to point me in the right direction. If you don't want to help me,
fine, no one's forcing you to spend your valuable time on me. I
certainly don't have the expectation that I'll get an answer, though I
am hopeful. But why spend your time to give a non-answer? That only
wastes everyone's time. If you don't have an answer, don't reply!


You are forgetting the nature of the question. It was a philosophical
question, not a technical question. You got a philsophical answer.

--
Neredbojias
Infinity can have limits.
Apr 13 '06 #20
Timothy Larson <th*********@cox.net> wrote:
Apologies for the long OT rant.


You'd long have the answers you are looking for had you spent that time
looking at the archives.

But apparently you can't be arsed to check out the first link from
http://groups.google.com/group/comp....=xhtml+vs+html

So here ya go you lazy sod:
http://www.spartanicus.utvinternet.ie/no-xhtml.htm

--
Spartanicus
Apr 13 '06 #21
Henri Sivonen wrote:
That's harsh. Why do you think HTML5 is "totally bogus"?


It's as bogus as RSS 2.0, and for much the same reason. It's
prioritising a personal dislike of some aspect of the existing spec
(which might even be a valid complaint) over the benefits of having a
widely accepted standard.

Standardisation is itself valuable, at least when it comes to web
publishing standards.
As to why HTML 5 is bogus, rather than Hixie's attitude being bogus...

We don't need "HTML 4+1" to improve the rendering of web pages. HTML 5
doesn't even attempt to offer much here anyway.

In fact what we _really_ need is HTML 4 + CSS. 1997 is _long_ overdue!

I certainly have sympathy for the view that "HTML 4.01 is all you need,
not XHTML". However this doesn't justify HTML 5, nor is it a long-term
viewpoint. XHTML is currently problematic because of transitory issues
about browser support, not because XHTML is itself flawed. Long-term
there's also a developing value to taking the XML path.

What we do need new is something akin to XForms. "Web applications" are
the interesting part of web development at present - for static
presentation we're still barely scratching what CSS has done for years.
AJAX is hot, as are all manner of "sophisticated but still thin"
client technologies. What these happen to have in common is the use of
XmlHttpRequest or similar as a component and these _simple_,
_lightweight_ components are all based around XML, not SGML.

So if we move towards "Web 2.0" at all, it'll be through XML, not SGML,
parsing models. Yes, we're not there yet.
As to sheer unbridled bogosity, then look at the dichotomy between
HTML5 and XHTML5. If HTML5 is a "solution" to the non-problem of XHTML
being in XML, then the very last thing we need is to split the format
further.

XHTML works. It works today, barring the trivia of SHORTTAG, and
assuming a level of developer competence that's sadly unknown outside
this newsgroup. What's broken and needs fixing is the level of
training, skill and competence out at large, not the spec itself.

If you need a way to indicate HTML/SGML or purist XML processing being
appropriate, then use content-type. That's what it's for. We don't need
a HTML5 that looks either big-endian or little-endian depending on
which way up you hold it. Yes, that's very _clever_, but it's no damn
use for anything.

Remember what happened in 2000 when XHTML Appendix-Bogus started to
appear. Now imagine what a stinking mess HTML 5 would look like, when
passed through the hands of a pack of fashion-victim clowns with
dreamweaver.

Apr 13 '06 #22
VK
> A couple years ago it seemed like XHTML was the direction of most web
markup, a foregone conclusion. Now I return to the scene and I see many
here recommending that authors stick to HTML, albeit in strict mode.

What happened?


Nothing terrible for the humanity. :-) Just one of standard proposals
appeared not needed / not usable for the world community. It happened
before - it may happen in the future.

I point you to the present:
<http://www.geocities.com/schools_ring/test.xml> This hugely simplified
sample gives the rough idea of what the business world will deal in the
next 5 year: XML+XSLT+CSS.

Before going further I would like to remind you the situation at the
moment you left: the main idea was to separate the content from the
layout to make them completely separate independently carried units. At
some point in the evolution it got separated onto two branches. The one
led by W3C was targeted to XHTML. Another one led by Microsoft was
targeted to XSLT. It is important that originally both directions was
fully blessed and controlled by W3C (which must be one of reasons of
their crying right now :-)

XHTML branch appeared to be a dead one. No one ever needed it. For
corporate customers it is still too much of data/layout soup which is
not flexible enough for transformations and platform porting. And for
regular John Doe's it was too complex to learn to make an online
family album - without any evident benefits of these new learnings.
Mostly unknown to anyone, it was evolving in a relatively narrow circle
of supporters. As any isolated branch it became its degradation. The
original *practical* idea of content and layout separation got
substituted by abstract semantical researches: what layout fits better
to a certain content. And the last year or two this degradation failed
on the new level: now it was even more abstract considerations how to
fit all eternity of semantical contexts into a couple of dozens of
official HTML tags. Truthfully this task seems to me even more
challenging then even the Newspeak Development Group task (see Orwell,
"1984") - though completely useless.

While this branch was slow evolving to its imminent end denoted by the
article quoted and linked here
http://groups.google.com/group/comp....47562c9c4af6da,
at the same time another branch was developing wery quickly -
especially after IE 5.5 release. After many complains and screaming
Microsoft brought its XSLT model in accordance with W3C specs, so
starting IE 6.0 they are the same.
The next major step will be done with the release of Windows Vista and
the new Office. Finally it will be the long prepared model there the
Data is totally abstracted from the Format. There will be no more HTML
pages, Word documents, Excel sheets or Access tables. It will be only
Data. But each particular request of this data will trig an avalanche
aka chain reaction of connections between totally separate elements to
represent the data in the needed way. It is going to be this way -
does anyone likes it or not - it is not important anymore.
Respectively all UA's makers have to take an urgent decision: are
they going to continue to participate in the race or they will drop it.
I know at least two UA who already took the positive decision: IE (of
course :-) and Firefox. Safari 2.x seems inclining to the same. The
others either still thinking or simple not aware.

Apr 13 '06 #23
On Thu, 13 Apr 2006, Andy Dingley wrote:
In fact what we _really_ need is HTML 4 + CSS. 1997 is _long_
overdue!
AIUI, now that SGML has the "web SGML Adaptations Annex", it is
technically feasible to write a buttoned-down DTD for something
closely resembling HTML4, which essentially permits the things which
current browsers do with valid HTML, but disallows the things which
SGML didn't previously have the ability to disallow via DTD. I'm
talking primarily about the unbundling of the various features of
SHORTTAG. (Just to be specific, I found this with google:
http://www.is-thought.co.uk/book/sgml-4.htm#Shorttag )

However, I don't expect to see such a cleaned-up HTML DTD coming from
the W3C, seeing that they put their shirts on XML.
I certainly have sympathy for the view that "HTML 4.01 is all you
need, not XHTML". However this doesn't justify HTML 5, nor is it a
long-term viewpoint. XHTML is currently problematic because of
transitory issues about browser support, not because XHTML is itself
flawed.


Unfortunatly, (and on this point I think Hixie's dire warning is
correct, no matter what one's reaction to the rest of his arguments),
by the time that there will be sufficiently widespread *real* support
for XHTML, there will be a vast legacy of XHTML-flavoured tag soup to
be coped with out there, and no browser developer will be able to
resist the demands for tag soup slurpers for rendering it. Thus,
we'll be right back where we were before - with whole menageries of
heuristic fixups, and no developer *daring* to report "hey, this
so-called XHTML is preposterous, no way could one expect to render
this crap".

Apr 13 '06 #24

Alan J. Flavell wrote:
Unfortunatly, (and on this point I think Hixie's dire warning is
correct, no matter what one's reaction to the rest of his arguments),
by the time that there will be sufficiently widespread *real* support
for XHTML, there will be a vast legacy of XHTML-flavoured tag soup to
be coped with out there,


I think this is true, but it's only going to be a problem for _pages_,
not for AJAX fragments or smart thin clients (where XML well-formedness
is forced upon you) and they're where the interesting aspects of web
design is heading.

Something that renders not-too-badly for 3.2 / tables / broken "X" HTML
is going to become a legacy kludge hung on the side of a browser.
No-one really cares. Those who need it wouldn't understand the problem
anyway. Much like EBCDIC support or floppy drives.

Apr 13 '06 #25
VK wrote:
After many complains and screaming
Microsoft brought its XSLT model in accordance with W3C specs,


That's unfair to M$oft. They're evil buggers generally, but for XSLT
they were quick to release something usable even before the ink was dry
on the spec and then they brought it into line with the XSLT mainstream
as soon as possibly practical. For this issue, they behaved well.

Apr 13 '06 #26
In article <11*********************@g10g2000cwb.googlegroups. com>,
"Andy Dingley" <di*****@codesmiths.com> wrote:
Henri Sivonen wrote:
That's harsh. Why do you think HTML5 is "totally bogus"?
It's as bogus as RSS 2.0, and for much the same reason. It's
prioritising a personal dislike of some aspect of the existing spec
(which might even be a valid complaint) over the benefits of having a
widely accepted standard.


RSS 2.0 is bogus, because the datatype of the content of the title and
description elements is ambiguous, which leads to dataloss. HTML5 has no
such bogosity.
We don't need "HTML 4+1" to improve the rendering of web pages. HTML 5
doesn't even attempt to offer much here anyway.
HTML5 is not about rendering. CSS is about rendering. HTML5 is about
adding things that are useful for Web apps while adding some semantic
document-oriented markup on the side.
In fact what we _really_ need is HTML 4 + CSS. 1997 is _long_ overdue!
That combination is in a pretty good shape considering what parts of
HTML 4 are actually implementable.
I certainly have sympathy for the view that "HTML 4.01 is all you need,
not XHTML".
In addition to the HTML5 tag soup serialization, there is the XHTML5
serialization.

I recommend using XHTML inside your server-side system and as a wire
format if you are mixing MathML or SVG with it.
However this doesn't justify HTML 5, nor is it a long-term
viewpoint. XHTML is currently problematic because of transitory issues
about browser support, not because XHTML is itself flawed.
The tag soup serialization is needed in order to provide an update path
for people and institutions who already have tag soupish systems. XHTML2
plus XForms is a "boiling the ocean" solution.
Long-term there's also a developing value to taking the XML path.
There is XHTML5, so that path is not abandoned.
What we do need new is something akin to XForms.
Compared to Web Forms 2.0 and JavaScript XForms promises static analysis
of form field relationships that cannot be done when an imperative
programming language is involved. But why bother with architecture
astronautics if the simpler solution is good enough in practice and fits
better in the reality of the network economics of the browser ecosystem?
"Web applications" are
the interesting part of web development at present - for static
presentation we're still barely scratching what CSS has done for years.
AJAX is hot, as are all manner of "sophisticated but still thin"
client technologies. What these happen to have in common is the use of
XmlHttpRequest or similar as a component and these _simple_,
_lightweight_ components are all based around XML, not SGML.
XMLHttpRequest is about doing stuff with a relatively simple API and an
imperative language. It is the antithesis of static analysis of
functional relationships and layers of abstraction that XForms is
supposed to provide.
So if we move towards "Web 2.0" at all, it'll be through XML, not SGML,
parsing models. Yes, we're not there yet.
HTML5 is not about SGML. It is explicitly jettisoning the fiction of
HTML being an application of SGML. Instead, HTML5 codifies a parsing
model that fits the tag soup reality.
As to sheer unbridled bogosity, then look at the dichotomy between
HTML5 and XHTML5. If HTML5 is a "solution" to the non-problem of XHTML
being in XML, then the very last thing we need is to split the format
further.
The tag soup facet is needed, because of the legacy of tag soupish
systems that need to be incrementally updated and would be too expensive
to re-engineer as proper XML systems.

It's uncool, but the legacy and economic incentives that follow from the
legacy don't go away by simply wishing them away. Web specs need to be
considered in the context of reality.
XHTML works.
Inside a CMS engineered for XML from the ground up, yes.

With IE, Google, WordPress and MovableType, no.

(I am not naming names, but I have caught smart people from both the
WordPress and MovableType camps with their proverbial XML pants down
when they have tried to make the systems designed for tag soup emit
XHTML.)
Remember what happened in 2000 when XHTML Appendix-Bogus started to
appear. Now imagine what a stinking mess HTML 5 would look like, when
passed through the hands of a pack of fashion-victim clowns with
dreamweaver.


At least this time round, the alleged clowns have a realistic parsing
spec to go with it.

--
Henri Sivonen
hs******@iki.fi
http://hsivonen.iki.fi/
Mozilla Web Author FAQ: http://mozilla.org/docs/web-developer/faq.html
Apr 13 '06 #27
In article <7bs%f.2147$ZW3.54@dukeread04>,
Timothy Larson <th*********@cox.net> wrote:
What do you like of HTML5 compared to XHTML?


I like the reality-based approach of providing an update path for tag
soupers without boiling the ocean.

It is ugly and uncool, but specs that aim for success in the real world
cannot ignore network economics and the effect of legacy systems.

--
Henri Sivonen
hs******@iki.fi
http://hsivonen.iki.fi/
Mozilla Web Author FAQ: http://mozilla.org/docs/web-developer/faq.html
Apr 13 '06 #28
Timothy Larson wrote :
Spartanicus wrote:
Timothy Larson <th*********@cox.net> wrote:
If you happen to recall whenabouts one of the more
interesting/informative discussions occurred, that would help me wade
through all the threads.


It takes a few minutes to browse past threads, spend them.


If you can't be helpful in response to questions, don't bother to post
replies. Your post certainly didn't do me any good - how did it help you?


Timothy, I sincerly think Spatanicus answered you what you should have
been answered. I certainly would have answered you to search through
archived posts.

Gérard
--
remove blah to email me
Apr 14 '06 #29

This discussion thread is closed

Replies have been disabled for this discussion.

Similar topics

87 posts views Thread by CMAR | last post: by
47 posts views Thread by Chuck | last post: by
10 posts views Thread by git | last post: by
14 posts views Thread by Stanimir Stamenkov | last post: by
reply views Thread by kmladenovski | last post: by
By using this site, you agree to our Privacy Policy and Terms of Use.