Monday, September 15th, 2008

2022 and why dates aren’t what matter

<blockquote>

Jeremy Keith

“Jeff, I agree with you. What’s important isn’t when a spec is officially “done”, it’s when we can starting using the new shiny stuff. That’s why I was advising against paying any attention to the 2022 date: it’s completely irrelevant to developers like you and me.”

This was in response to Jeff Croft’s piece on HTML 5 and the 2022 date that people found on the recent Hixie interview. I had a feeling people would jump on that, but as others have mentioned and Jeremy (and Simon) say clearly is that a lot of HTML 5 is here or arriving very fast indeed. The browsers are engaged again, because we as developers are very engaged, so we should expect more.

Anne Van Kesteren points out what a W3C recommendation is all about. It is true “recommendation” is a strange word. For some reason that feels like a fluffy thing that people do early on. “Hey, before you get started, I have a few recommendations” rather than something that comes later on in any process :)

Related Content:

Posted by Dion Almaer at 2:05 am
4 Comments

+++--
3.2 rating from 14 votes

4 Comments »

Comments feed TrackBack URI

“Agile” recommendations ^^

Comment by Ben — September 15, 2008

This is such a sad joke. Just goes to show that our new standards overlords are totally out of touch, despite being shining beacons of being in touch with reality as compared to our old standards overlords.
.
To WHAT WG/W3C: This kind of stuff is why people are distrustful of you and consider you irrelevant. Sure, the date is “not significant”; it’s just a message to all of us that the standards process is so bureaucratic and full of unnecessary obstacles as to require a decade and a half to complete a specification.
.
And we’re supposed to start using it before that? Do you guys do web development? Like, ever? When fully half of your work is working around implementation differences, you’re supposed to be encouraged that you can start using “the shiny new stuff” in different parts of differently implemented interpretations of an incomplete spec across a browser landscape that never seems to converge?
.
Moreover, as much as WHAT has given the IE team flack for its duplicitous participation in the standards process, you guys just gave them free reign to do whatever the hell they want with this spec for the next decade or more. When we complain that IE isn’t cooperating, we’ll face the completely moronic retort that they can’t be expected to fully implement an incomplete standard: just like CSS 3.
.
It’s not like we don’t understand, the job in front of you is difficult. Standardizing all the nonsense that existing web practices entail is… a monumental task. But if you can’t make it happen in over a decade, something is very wrong.

Comment by eyelidlessness — September 15, 2008

2022? By then neuro implants equipped with Flash Player 38 will be developed and that’ll render HTML5 useless.

I totally agree with eyelidlessness: It’s a difficult job, but more than ten years?

Comment by mawe — September 16, 2008

The reason that World Wide Web **Consortium*** produces Recommendations is that the W3C _is NOT a standards body_, it is a “consortia” (look up the definition!). The difference is that standards bodies, like ISO, produce standards that can become law; with legal consequences for those who do not follow them. On the other hand, a consortium, like the W3C, can only produce “recommendations” because there is no legal recourse for companies not following a recommendation.

Now, Anne is correct to say that the HTML5 spec should be considered mostly done by 2009 (yes! DONE! AS IN “FINISHED”!). Anne, and the W3C process document, make it really clear that recommendation means two fully interoperable implementations. To have interoperable implementations you need something to test against (a test suite). Test suites that capture all the features of HTML 5 working together will take 11+ years to produce because HTML5 is so complex.

On the other hand, it means that development of HTML6 could begin as early as 2010.

People should get over the date and just enjoy the new features. Leave the process up to the standards people.

Comment by marcosc — September 16, 2008

Leave a comment

You must be logged in to post a comment.