Page 1 of 2

HTML 5 is dead?

Posted: Thu Jan 20, 2011 3:05 pm
by MikeGale
It looks like we're definitely not in Kansas anymore. And maybe we're not even in the universe we thought we knew up till the 18th of January 2011.

Just when the W3C launched the new logo for HTML5 (It looks like a shirt badge for a superhero comic to me. I don't like it.) See http://www.w3.org/html/logo/ for the full glory.

Image

The WHAT WG announces that HTML 5 is no more. It's now HTML and it will be a rolling standard. No snapshots, i.e. hard and fast written specs. Instead we will have a set of moving goalposts, though there's a promise that it will always be backward compatible.

See http://bit.ly/fQGoNs for Ian Hickson's announcement.

I got wind that something like this was in the wind, I thought it might happen years from now.

Gone are the days of saying a page is XHTML 4 transitional. I guess we'll have to say it's now HTML-2011-01-21:12:00:00Z or somesuch if we want to be precise.

I haven't processed whether this can actually happen, what the push back will be and what it means for me, CSE etc. Here's a few stream of consciousness thoughts.

1) To define what you've done you'll need something like the specification above. Presumably there'll be something like daily drops of the standard, so it might be more like HTML (formerly HTML 5) version 6.23.1 with an optional 2011-11-11 11:11:11Z.

2) CSE won't be able to define standards like currently does. An interesting challenge.

3) CSE might need to publish updates more often. Maybe with a update plug-in service, automagically distributed!

4) For those of us who need to use XHTML I don't know what this means. I'm increasingly using code to work with markup (and go beyond the extreme limitations of markup) so I want it well formed and rational. Need to establish if this is a disaster.

5) Will browsers get even more out of synch than they are at present. In fact will they say let's forget it and just do our own thing (after all this standard is no longer a real standard anyway).

6) Where will this leave ventures like the developer only plug ins that IE has introduced (in IE 9) to isolate experimental from in the standard. I thought that was a really sane move, I was hoping it would become common in browsers, now I just don't know what'll happen.

7) Chrome OS looks like it's dying, what impact will this have on that.

8) Some mobile devices, especially Android device it seems, essentially never get an update. Will this create a new ghetto of old and no longer cool/hip web viewing machine. (In months, weeks, days or maybe even find them all launched-as-legacy-devices.)

9) Is this a kickback against the announcement from W3C (with that logo announcement) that HTML5 now means, CSS and a slew of other things.

What's your take?

Re: HTML 5 is dead?

Posted: Wed Jan 26, 2011 7:36 pm
by Albert Wiersch
I've been wanting to reply to this. Thanks for posting it Mike.

*) It was a bit of a shock to me to see that logo... and then to read that HTML5 is now just "HTML".

*) It makes sense in at least one regard - that in reality HTML does seem to be like a rolling standard with different browsers implementing HTML5 to different extents. I kind of like the idea... CSE HTML Validator was never about holding strictly to the version of HTML in the DOCTYPE, it was more about what works in the real world, and in that sense the browsers don't really pay much attention to the version of HTML, except for perhaps adjusting "quirks" mode.

*) I wonder if it would make sense to have a defined version of HTML at the beginning of every one or two year - like "HTML 2011" to define what worked for most browsers at the beginning of that year (so authors could design to that some defined & static specification)... but this might also be too much to handle/maintain.

*) People who want reliability and simplicity could still write to HTML4. I think you really can do a lot with it if you exclude all the "fancy" stuff. Probably good enough for 99%+ of needs - which is mainly to deliver basic content (text & images) and perform basic transactions like banking and shopping.

*) I've thought about a quick updater system for CSE HTML Validator... and it is a possibility for the future.

*) What does scare me a little, is if people start designing stuff that only works in one browser again (like they use to on the "old days" with IE and Netscape), but I think that would obviously be a bad idea now as there are just too many different browsers out there so the developer would be locking a lot of people out.

Re: HTML 5 is dead?

Posted: Wed Jan 26, 2011 7:57 pm
by MikeGale
Assuming this all happens, the further disintegration of the web is a real issue.

I am noticing more failures to deliver cross browser pages than before. (I'm thinking sites from major companies.)

The IE plan to have only developers install the experimental features looks like a promising way to mange it. You might even get browser manufacturers co-operating to release a batch of features at the same time!

In a world like that, a CSE that keeps up to date with what the browsers actually do, would have additional importance to companies that take the web seriously. (Especially if the page (scroll!) itself doesn't declare which subset of markup it needs.)

Re: HTML 5 is dead?

Posted: Mon Mar 28, 2011 11:35 am
by marywallace66
HTML 5 is just regular html. Nothing more nothing less. The way you make pages show across browsers is to use a cms (content management system) such as wordpress. It makes simple webpages easy and effective across all browser. Editing the style sheet is easy with an online editor such as http://www.easywebcontent.com

Re: HTML 5 is dead?

Posted: Mon Mar 28, 2011 5:38 pm
by MikeGale
Hi Mary,

Thanks for your input
is to use a cms
There are different approaches to the web. I don't subscribe to the cms only approach. I use Wordpress it's true, but I'm constantly aware that the simple interface prevents me doing things that were dead easy a decade ago (if I wrote the code by hand).

I'm not familiar with Easy Web Content but at $10 per month I'm not going to try it. I suspect it's one of the products that will lock it's users into ignorance by hiding the underlying technology.

The web has so much more potential than the version that most people are aware of. Fully flowing design, talking to back end servers, some decent code in the page, can do a lot more than the limited aspirations of most cms systems.

Seen from that perspective HTML 5 has additional potential. Canvas, persistent storage, combined with the new crop of JavaScript engines / frameworks, SVG and a lot else will eventually make it possible to easily do things that are now quite hard.

So while I agree that cms is good for some purposes, there are things that the CMS's I've tried simply can't do. I recommend exploring that higher potential.

I don't agree that HTML 5 is the same as previous versions. It has introduced new features which change what can be easily done. (I feel the advances are too modest, but that's another story.)

Re: HTML 5 is dead?

Posted: Wed Mar 30, 2011 8:53 pm
by ormaaj
Mary it technically right in that all current browsers have reduced their parsing/rendering modes to just HTML5 and XHTML. If you're using some older version of HTML then you're really just using a subset of HTML5 as far as the browser is concerned. This is quite nice because as a side effect you get the HTML5 tag soup parser's standardized treatment of broken pages across browsers regardless of which version the page claims to be using. For that reason HTML5 can't be too ambitious. It's primary goal is to retain backwards compatibility. If you wanted radical changes that would probably be more appropriate for some XHTML2-ish language which we're probably in for a long wait for. The closest thing to a powerful HTML replacement for doing a lot on the client that's here is IMO remote XUL which unfortunately has been disabled by default in FF4. SVG is another candidate but it has a ways to go (1.1 is a bit of a mess while waiting for 1.2/2.0). IE, which was previously the only browser to support SMIL+XHTML no longer supports it. SMIL still works with SVG but I think there's some confusion about it's relationship with CSS3 transitions/transforms/animations. People who want that kind of stuff are probably just saying the hell with it and going either canvas+JS or Flash.

CMS doesn't really have to be limited. I'm in the process of migrating my pages to a uWSGI backend which just acts as a simple templating system (processes xincludes via LXML). I suppose I could make it feed whatever code to clients I wanted. Should give me more power than simple static pages on the server with a bit of SSI. Thing is without a CMS you don't get nice transparent content negotiation. In order to make a maintainable website using XHTML it's pretty much necessary to use a CMS along with middleware for HTML5 serialization depending on what's going on on the server.

Re: HTML 5 is dead?

Posted: Mon Apr 04, 2011 7:16 pm
by MikeGale
I'm not sure exactly what uWSGI is. My quick research suggested a web server / server helper originally designed to publish Python to web, but now capable of more. If that is the case it looks like a tool for what I think of as the smart web. That is content that taps into the unique potential of the medium in a meaningful way. Not sure how a CMS fits into the data flow or what it does. (I assume that uWSGI doesn't do it all.) One of my concerns with the CMS systems that I was talking about, is that they end up virtually guaranteeing content that does not use the potential. (Text and images only, no specialised use of code, discourage or prevent original thought outside the box...). Is there some succinct reference putting CMS into your context?

Have browsers done away almost entirely with looking at doctype?
have reduced their parsing/rendering modes to just HTML5 and XHTML
(Browser promiscuity in accepting rubbish, has had impact on the rise of incompetent pages and sucking compute cycles out of computers all over the Internet.)

There's an interesting post about why it would be a good idea to stop HTML 5 roll-out. http://bit.ly/g7RHqJ I don't think it's stoppable, but the points are worth thinking about

Re: HTML 5 is dead?

Posted: Tue Apr 05, 2011 8:27 am
by Albert Wiersch
MikeGale wrote:There's an interesting post about why it would be a good idea to stop HTML 5 roll-out. http://bit.ly/g7RHqJ I don't think it's stoppable, but the points are worth thinking about
Thanks Mike... that was indeed interesting. It reminds me of politics and how hard it can be to get things fixed especially when they have become established and expected.

Fixing security issues is a huge task because it would probably require a lot of changes that break things... but I think that it would be worthwhile, especially in this day and age when any technology that can be abused for illicit gain will be.

Re: HTML 5 is dead?

Posted: Tue Apr 05, 2011 5:31 pm
by MikeGale
It reminds me of politics
In some ways it was a lot easier in the earlier days to just get things done. Little or no politics. There were not so many people involved or even interested and those involved were technically savvy.

Like:
1) The Xerox PARC guys lived up to a reputed motto "the best way to predict the future is to invent it". (OO programming, TCP-IP, GUI's etc. etc.)
2) Brendan Eich is reputed to have created the original Javascript in about 10 days.
3) I've heard the original VBScript implementation was done over a weekend.
4) The original XMLHTTP was created a very long time ago and remains much the same. I imagine it was done quietly and efficiently, though I don't have any insights. Astonishingly good idea.

Now we have a lot of people interested. A lot of the involved parties are clueless, but have political clout. Net result smart decisions are hard to impossible as the group reduces the "centre of IQ and competence" (akin to centre of gravity).

I'm coming more and more to the conclusion that the web can hold the answer. With it we can build more competent and productive groups that would never get together in the world of bricks and mortar.

That might take us back to something like the early days but better.

Re: HTML 5 is dead?

Posted: Wed Apr 06, 2011 9:57 am
by Albert Wiersch
MikeGale wrote:Now we have a lot of people interested. A lot of the involved parties are clueless, but have political clout.
Good point. There's a lot of interest because the web is so big now, which means there are a lot of "special interests"... and you probably know what happens when you have a lot of "special interests" involved. :wink:

I think there is still opportunity for new tech to be developed and catch on, but once it really catches on, then a lot of people come in and want their say.

I agree that the web can bring people together from around the world that can make for very productive groups... but the productivity completely depends on the number of people (too many can reduce productivity), their quality (the "wrong" people can reduce productivity), and their interests (compatible interests would be much more productive I'd say).

Re: HTML 5 is dead?

Posted: Wed Apr 06, 2011 6:06 pm
by MikeGale
In the bigger scheme we're still on topic here, I think
I agree that the web can bring people together from around the world that can make for very productive groups... but the productivity completely depends on the number of people (too many can reduce productivity), their quality (the "wrong" people can reduce productivity), and their interests (compatible interests would be much more productive I'd say).
It's an old issue going back to mainframe days. We need to solve it. I'm sure we will, and those who still know of a world before this all happened have experiences and insight that can help. Some random points:

1) Really tight and effective operational groups are small. Say numbers below 8. (8 seems to be a poison number for a committee so the G8 may be doomed!) SAS stick of 4, Jobs+Wozniak come to mind.

2) At a push it seems that people can have up to 150 or so "close acquaintances". That tends to be in war or difficult situations. (Dunbar's Number)

3) We already have different ways to discuss issues (disqus, twitter, FB, forums, newspapers, SourceForge, Quora...), ways to split a single discussion into your personal virtual discussion (look at some of the forum controls on /. or Kuro5hin), look at the editorial groups that people like Paul Graham use when writing a post... In my view this sort of thing has a long way to go. (Facebook has a way of grouping things, it's hard to use, a recent recruit to that organisation is, I suspect, currently working hard to somewhat fix that http://j.mp/hnaYES )

4) There are ways to communicate privately on the web. They're just not much used. I think privacy can really help some discussions. (Free of feeling that you're talking in front of the whole audience at Woodstock!)

5) Interests (and attitudes) shift. You sometimes learn stuff then move on. In some situations the ideal is to have a group that is both a little ahead of you and a little behind. You can then learn from some and help others to learn. As time goes on such a group might naturally split. Groups that work like that could be real winners.

That's enough.

Re: HTML 5 is dead?

Posted: Sun Nov 13, 2011 9:09 am
by Lowiepete
So, a headlong rush to convert XHTML strict to HTML5 is not that urgent?

Re: HTML 5 is dead?

Posted: Sun Nov 13, 2011 4:32 pm
by MikeGale
To make that decision I'd make an analysis.

(I've other things to do at present...)

If anybody else has started this already I'd appreciate a link!

Things I'd look at include:
1) My own actual content (not a universal catch all with tags that I'm not going to use!).
2) The browser spectrum I'm currently seeing (in other words, if only 3% of your audience won't get a great result, it's worth moving now, but if that's 25% forget it.)
3) Active testing how the UI works in real use on the major browsers (not just how it looks, but working with web services etc.).
4) Visual checks on a wider range of browsers, including what happens on mobile devices (which might need a separate interface design)

I don't think it's yet a one size fits all decision, though it will get there sometime.

The overarching issue behind all that is the benefit you get from the work.

Re: HTML 5 is dead?

Posted: Mon Nov 14, 2011 4:29 pm
by Albert Wiersch
Lowiepete wrote:So, a headlong rush to convert XHTML strict to HTML5 is not that urgent?
I agree with Mike. Basically, it depends. "It depends" is almost always a great answer. :D

I'd only rush out of XHTML strict if there were a good reason too - like problems with your site because it's XHTML. Perhaps problems caused by using XHTML with certain user agents, including those on mobile browsers. Accessibility may also be a deciding factor as I suspect that it's easier to write accessible HTML5 documents than using older XHTML strict based on HTML 4.01.

If the site is working pretty well, with few or no complaints, then why "rush"? However, I would head towards HTML5 when convenient.

Re: HTML 5 is dead?

Posted: Mon Nov 14, 2011 4:48 pm
by MikeGale
I imagine one of the big motivations is going to be the additional capabilities.

I'd prefer to see more but what is there already will be compelling to many. Especially now that Adobe seems to have thrown in the towel on Flash for mobile. (see http://j.mp/uKvjbm )

SVG, Canvas, video and audio will be compelling for many and the sockets and client side storage for others, etc.

Admittedly the majority of web content ably illustrates that most web site decisions have been made by people who haven't a clue about the power that's available. I shake my head at the number of pages that can't flow and resize, that can't change text size comfortably and a lot more. Cluelessness will prevail I'm sure but at least some will cash in on the possible and use some of the power waiting at our fingertips.

So for those who want to do more with the web, there will be a temptation to jump sooner. (In some cases that will mean two versions of some content, which will be used until HTML5 capable browsers are common enough!)