One Web to Rule Them All

September 20, 2008 10:21 PM

"I would like to take a broader look at the Web. I would like to consider what the Web can do for society on a scale we have not yet seen. And I would like to enlist your help to get us there." ― Tim Berners-Lee announces the World Wide Web Foundation

"Our success will be measured by how well we foster the creativity of our children. Whether future scientists have the tools to cure diseases. Whether people, in developed and developing economies alike, can distinguish reliable health care information from commercial chaff. Whether the next generation will build systems that support democracy, inform the electorate, and promote accountable debate." ― Berners-Lee's speech before the Knight Foundation announcing the formation of the World Wide Web Foundation.

Here is the concept paper.

posted by netbros (30 comments total)

Our success will be measured by how well we foster the creativity of our children. Whether future scientists have the tools to cure diseases

Because, you know, in the past scientists were unable to cure diseases.

It sounds like what he's after is some kind of global moderation system. I think moderation should be left to individual site owners. I mean, it isn't like there are no places where you can find reliable medical information.

Also, I wonder if TBL is feeling a little left out now that people are moving away from HTML/XML and other w3c technologies to new RIA technologies like flash and silverlight.
posted by delmoi at 10:38 PM on September 20, 2008

Um ... not to poop on this wonderful idea, but aren't literacy and critical thinking skills necessary prerequisites to all this?
posted by ZenMasterThis at 10:43 PM on September 20, 2008

If the web moves to flash as a primary mode of content delivery, I'll gladly go back to printed matter organized by card catalogs. Christ, event the YouTube interface sucks.
posted by kaibutsu at 11:29 PM on September 20, 2008

even, even.
posted by kaibutsu at 11:29 PM on September 20, 2008

"Inventor of Web Announces Creation of New Foundation to Bring the Web to All People"

"The World Wide Web was invented in 1989 by Tim Berners-Lee, who envisioned an open, universal information system connecting all people."

"Almost twenty years after he invented the Web, Tim Berners-Lee is leading the effort to create the World Wide Web Foundation ("Web Foundation")"

"Inventor of World Wide Web Launches New Foundation With Knight As Seed Donor
Sir Tim Berners-Lee, inventor of the World Wide Web, unveiled the World Wide Web Foundation Sunday."

I looked him up on Wikipedia. It's true.
posted by longsleeves at 11:30 PM on September 20, 2008

This guy's still around? Meh. He hasn't contributed anything useful to the world since HTTP, just high-minded fluff like the "semantic web."
posted by wastelands at 11:36 PM on September 20, 2008

TBL has been irrelevant for a long time.

He built a beautiful working system that was everything the 'experts' on hypertext thought was terrible, an anti-Xanadu — then after it is wildly successful for exactly those reasons; he turns into Ted Nelson 2, promoting anti-web bullshit like [XHTML, SOAP, RDF, WSDL] and steering the W3C into a corporate hole in the ground.
posted by blasdelf at 11:37 PM on September 20, 2008

Awww, I guess Tim Berners-Lee is feeling guilty

WELL TOUGH. I WILL NEVER FORGIVE YOU FOR TUBGIRL, OR FOR GOATSE.

NO FORGIVENESS FOR TIM BERNERS-LEE
posted by East Manitoba Regional Junior Kabaddi Champion '94 at 11:38 PM on September 20, 2008 [4 favorites]

I mean, it isn't like there are no places where you can find reliable medical information.

Um ... not to poop on this wonderful idea, but aren't literacy and critical thinking skills necessary prerequisites to all this?

The problem is there are fewer and fewer places to find reliable medical information, and even though one is literate and has the requisite thinking skills , it is no longer enough. Berners-Lee is calling for an ethical solution, not one based on disinformation, personal accomplishment or greed.
posted by sluglicker at 12:12 AM on September 21, 2008 [1 favorite]

Forgiveness?

Fuck that, I'd denounce him as a counter-revolutionary and hold a show trial for his crimes against his creation!
posted by blasdelf at 12:24 AM on September 21, 2008

When it was speculated that HTML5 would reach the status of "Proposed Recommendation" in the year 2022, I knew then they could be safely ignored. I still do standards-based design, mind you, but they're making the UN look agile.
posted by potch at 12:32 AM on September 21, 2008

I was, of course, referring to the W3C.
posted by potch at 12:35 AM on September 21, 2008

Well, if you compare XHTML 2 to HTML 5, it's pretty clear which one offers a vision of the future and which one is frames and background music revisited.
posted by cytherea at 12:38 AM on September 21, 2008

The problem is there are fewer and fewer places to find reliable medical information

Is this true at all? Are you sure this isn't an imagined problem, like so many of TBL's second systems?
posted by blasdelf at 12:42 AM on September 21, 2008

If you want to know the future, imagine a frame stamping on the face of background music. Forever.
posted by East Manitoba Regional Junior Kabaddi Champion '94 at 12:44 AM on September 21, 2008 [4 favorites]

All about alt: With arguments like this, it's no wonder HTML5 will take so long.

In this interview, HTML 5 Editor Ian Hickson discusses his favorite features, the features he thinks might be most contentious, the pain points he expects HTML 5 will address, and much more. He also tells what he would change in the original HTML spec if he could go back in time.
posted by netbros at 12:48 AM on September 21, 2008

Yeah, the W3C has not impressed me in the last few years. I fully understand that they cannot enforce standards - in fact, it's that very recognition which makes me giggle at the semantic web as a concept. I have a hard enough time teaching people the difference between bold and strong, and when to use each, nobody ever uses the cite tag, and we're supposed to get people to annotate what they put on a webpage? Please. XHTML 2 is a joke. CSS would have benefitted by sitting down and talking with graphic designers about what graphic designers actually do. Their specs can be horribly vague and often leave too much up to the discretion of the user-agent. We don't even have columns done in a non-hackish way. Columns, staples of broadsheets and newspapers since when? And let's not forget just ... changing the validator. It's weird when a page judged to be valid XHTML one day is no longer valid XHTML the next, especially if that page hasn't changed. And then there's the horrific CSS validator.

The W3C has basically doomed itself to irrelevance. Even if they came up with good ideas, it'll be quite some time before anyone listens to them again.
posted by adipocere at 12:59 AM on September 21, 2008

cytherea: what makes you think the future isn't in revisiting "frames and background music"?

HTML5 addresses problems the web actually has, and isn't anal-retentive bullshit by/for bureaucrats.
posted by blasdelf at 1:09 AM on September 21, 2008

adicopere: bold and strong are synonyms :)

Oh fuck, don't they want me to mark up that smiley semantically? You know, so it'll be accessible!

The whole concept of External Validation (now there's a semantically loaded phrase) and draconian XML-style error handling is absolutely counter to why the web is successful. Instead of standardizing what happens to tag soup, the W3C pretends that the browser makers are just going to remove error handling all together.
posted by blasdelf at 1:18 AM on September 21, 2008

Bold and strong are not synonyms. The effect to most users is visually the same, but semantically they are not the same.

Bold is a weighting for a typeface. Strong is an emphasis that means that the text is logically of greater weight. In a blind-reader, strong is used with vocal emphasis, whereas bold can (and should) be used to simply make a bit of type more readable or stand out without that text having any particularly more important meaning.

Search engines can (and some do) rate strong text higher for the keywords within; bold does not typically have that. Bold is for graphic designers; strong is used to indicate that you have thirty days to return your product, or that you do not reboot after a reinstall. Note: Google still ranks them the same, because people assume that they are synonyms - it's an excellent case as we so often see in web development of popularity reinforcing errors.

Basically, read it out loud. If you raise your voice or volume of such when you say it, consider a strong tag; if you just want it to look like that, it's time for bold (whether it be a tag or as a style attribute in a span). You can bold a whole paragraph of text to make it stand out, but YELLING OUT THE ENTIRE THING GETS A LITTLE TEDIOUS AFTER THE SECOND SENTENCE.

This may seem like a quibble, but it's basically about as good of an idea as using a cite tag when you're citing a reference. Some people do not care about logical versus physical markup, but accessibility-wise, it's a good idea. For anything which may be machine-parsed in the future (perhaps building your own keyword set for a site - trying to figure out which CSS classes you've designed with a bold weighting correlate to what text is harder than just scanning for strong and em), it is a good idea.

Physical markup is going to be more and more abstracted out of the body of the text; logical will be staying in there.
posted by adipocere at 2:27 AM on September 21, 2008 [3 favorites]

adipocere: you think I don't know all about that?

You can argue that semantically, B ≠ STRONG and I ≠ EM till you're blue in the face, but that doesn't make it reality. No parser I've seen treats them as anything but synonyms. Markup writers and software generators use the new tags as synonyms of the old ones after being browbeaten by 'semantic web' folks — the B & I button at the bottom of this text box insert STRONG &ampl EM!

I don't know of a single 'semantic' tag that's been widely used in the way it was intended (most are not used at all!). There is no way to have unified semantics when there are a billion authors and all they really care about is presentation. It's just not how the web works. Condolences.
posted by blasdelf at 3:08 AM on September 21, 2008

It's just not how the web works.

Exactly. Every article I've read about the "semantic web" goes on about screen-readers for the blind. Now, not to diss on blind people, but if the only justifiable reason for differentiating between the strong and bold tags is so a few blind people can know whether the text is strong or bold...it's a pretty weak argument. I mean, the 99.9% of people who aren't blind can get by without knowing whether some text is strong or bold, so you have to wonder what all the fuss is about.

And look at i and em. Here I can see where there might be a reason. In some cases, in printed documents, italic text is used to add emphasis. In other cases, it's used, for example, in biological species names, without any emphasis intended. However, once again, we've got by for a few centuries with one font styling (italics) meaning multiple things, so why do we need the web to now over-complicate things?
posted by Jimbob at 4:10 AM on September 21, 2008

I'd also argue that the wide adoption of the ugly mish-mash of technologies that is AJAX has probably killed any hope of a simple, semantic web anyway. You can't unscramble the egg. Unless this functionality is encapsulated in a standard way within HTML instead of essentially being an ugly Javascript hack, there's no hope. I wonder what Facebook sounds like on a screen-reader?
posted by Jimbob at 4:13 AM on September 21, 2008

I wonder what Facebook sounds like on a screen-reader?

angst
posted by Mick at 6:11 AM on September 21, 2008 [4 favorites]

The semantic web is about much more than reading for the blind. It's about letting computers read (and understand) the web for you. You think you'd be able to function without google? The semantic web represents the same kind of leap in potential that the web represents compared to not having the web at all. It''s understandable that people would have a hard time imagining the value; no one was clamoring for the web or even the internet before they existed. And yes, XHTML 2 can handle AJAX, it can semantically mark up Events. And yes, CSS 3 has columns. And speaking of CSS, can you imagine how ugly the web would be today if we weren't using it?

Ugh. What we have here is a failure to imagine.

"This is so fucking American, man: either make something your God and cosmos and then worship it, or else kill it."
posted by cytherea at 8:22 AM on September 21, 2008 [1 favorite]

IMO, any approach to "letting computers read (and understand) the web for you" that involves humans doing more work so that computers have an easier time of it has things completely ass-backwards.

The success of HTML is in large part because almost anyone can write it, without a great deal of machine assistance (i.e a text editor.) Sure, it won't be 'correct' most of the time, and so browsers have become smart enough to do the right thing with this bogus markup in the majority of cases. But for some reason some portion of the engineering community (and I write this as an engineer) has become convinced that it would be easier to make HTML authors more virtuous than it would to make browsers smarter. Good luck with that.

So far, semantic web approaches seem similarly doomed if they continue to rely on metadata. The history of the internet so far seems to suggest that there is more motivation for people to 'game' metadata than there is for them to ensure it's correctness. Without guarantees that the metadata is correct - which would involve an amount of centralized trust that currently seems unfeasible - the whole thing is about as useful as the META tag. Remember that? For now, it seems like metadata-based approaches are only going to work in closed domains that are prepared to do a certain amount of policing.
posted by pascal at 10:07 AM on September 21, 2008 [1 favorite]

HTML was a success because it was forward thinking. Show it to people who haven't used it before, and they're going to scratch their head and react the same way you're reacting to RDF. The same way people initially reacted to CSS. And the reason to get away from tag soup is so you don't have to write the same web pages five different times because every browser decides to interpret HTML differently. Syntax checking is the computer's way of saying that it loves and cares for you.

And only professionals actually write HTML these days. Everyone else uses content managers or blogs or social networking software--applications that could stand to gain enormously from expressing the semantic data, a good deal of which is already contained in their database relations. And it's just not that difficult. Links already have a variable subject and object. The semantic web just adds a variable predicate. But no one actually expects lay people to use it.
posted by cytherea at 1:32 PM on September 21, 2008

And only professionals actually write HTML these days.

Bollocks. I'm not a professional web designer, but I coded up a website for myself yesterday.

But no one actually expects lay people to use it.

And this is why this whole this is so ugly.

The way the web took off with was free, open software for browsing and serving, and some fairly relaxed standards. Anyone could take those tools into their own hands and work with them. Sure, they might create Time Cube, but that was their right.

What you've just described is taking those tools away. Leave them up to the professionals, because everyday people can't be trusted to use them "properly". Let the professionals create the special semantically-enabled software you need to post on the web (and let the professional decide what you're able to do, and what you're allowed to to do with it).

It's like the difference between 'zines and News Corporation.
posted by Jimbob at 2:14 PM on September 21, 2008 [2 favorites]

Honestly, I believe history shows you to be dead wrong regarding lay people's difficulty grokking HTML versus RDF. The adoption rate alone should show you that. (On preview, what Jimbob said.)

Syntax checkers are indeed a web developers friend, but it's a stretch to say that all we have to do is get everyone to use a syntax checker and then we can turn off the old internet. we need to make all of the syntax checkers agree with each other, and all of the browsers agree with them and then get all of the CMSs and blogging systems and Ruby on Rails and other markup generating frameworks to do The Right Thing. For some reason there are only 4 main codelines for reading the internet but thousands of ways of writing for it, perhaps because Postel's law inevitably pushes all the hard work to the receiver.

That 'variable predicate' sounds so easy, but the problem is that it's meaning (unlike that of a link) is not intrinsic to browsing. That [href="http://sarahpalin.org"] means I am linking to Sarah Palin's site is verifiably and obviously true. That [a href="http://sarapalin.org" relation="http://w3c.org/illigitimate_love_child_of"] means I am Sarah Palin's love child is not - even if one trusts the W3C.

The social network example you give falls into the category 'closed domain that is prepared to do some policing' I mentioned before. Facebook knows who your friends are. But for Google (for example) to use that metadata, it must have a trust relationship (and quite likely a commercial one) with Facebook. Effectively this would become a walled garden and a rather less open internet than we have become used to.

I believe you are also conflating 'separation of presentation and content' and 'the semantic web'. You wouldn't be the only one - many people seem to refer to the former as 'semantic markup' for some reason. I do not believe these two things are the same. The former is very convenient for content creators but provably not a requirement for the web to function.
posted by pascal at 2:43 PM on September 21, 2008

Well, no, I wasn't confusing the two; they are not related at all, really. My reply was addressing a few different topics, but basically I contrasting the superficial nature of HTML 5 with the truly revolutionary potential of XHTML 2; and it's in XHTML that you happen to get things like OWL and RDF.

I'm not saying that semantic links will be limited only to things like Facebook, but I suspect that's where we'll see them first--it would certainly be a huge boon to something like Wikipedia, if only to eliminate the horrid and ubiquitous type prefixes found there. Yes, we may never see much hand written OWL, but we don't see much hand written HTML anymore either. Blogs would also be a natural tool. And you really can't say that we'll need to trust the semantic web anymore than we need to trust existing links: consider what comes up when you type in a certain ex-Pennsylvania Senator's name. We already get to choose which phrase we link to what page or anchor. I wouldn't want the semantic web to be any other way. But on the whole, the web's still a valuable resource, I'm not too worried about people playing predicate games.

I don't think it's going to be adopted very rapidly, but the web took some time too. It probably won't get much of a start until someone comes up with the first killer app. But even now, RSS uses RDF.
posted by cytherea at 9:02 PM on September 29, 2008