Semantics in HTML 5?

[brandaggio]

Thought this was really a good article.

[RyanReese]

Well you posted a question

"Semantics in HTML 5?"

That indicates you have a question.

*Waiting*.

[SoulScratch]

I think the idea of having attributes to define the semantics of an element makes sense, and maybe we should cut down on the number of extended elements that are in the HTML 5 proposal/pre-spec as we'll always be looking to add more and more, his article made me realize there's a certain limit and that HTML natively lacks extensibility since it's an application of sgml and not xml/xhtml, where you can define your own elements with a dtd.

Though, I think the real focus and pressure should be put on how to rid the world of IE6 and 7.

[RyanReese]

I put the destruction of IE on the top of the new presidents things to do list.

Heck if I worked for MS I'd destroy their a$$es.

[felgall]

The W3C published a complete solution to the semantics problem over 10 years ago with XHTML but it suffers from the same problem as HTML5 in that IE doesn't support it. All the other problems with extensibility in HTML5 are resolved in XHTML 1.1 though so why bother with HTML5 when it has all the same problem as XHTML1.1 plus the lack of extensibility.

[JsD2K]

I say put our heads down and bash through IE's problems like we have for the past 10 years

[brandaggio]

I philosophically agree, on this topic, with Felgall - moreover I am not a fan of "bolt on" anything - semantics, SEO etc.

It is interesting to see people take a stab at it though.

[felgall]

It is interesting to see people take a stab at it though.

I agree. It is really early days yet. By the time HTML 5 finally reaches candidate status they will probably have added and removed hundreds of different tags trying to figure out which is the right combination. The only certainty is that a lot of the ones that have been added so far will not survive.

[Dan Schulz]

I think the idea of having attributes to define the semantics of an element makes sense, and maybe we should cut down on the number of extended elements that are in the HTML 5 proposal/pre-spec as we'll always be looking to add more and more, his article made me realize there's a certain limit and that HTML natively lacks extensibility since it's an application of sgml and not xml/xhtml, where you can define your own elements with a dtd.

Though, I think the real focus and pressure should be put on how to rid the world of IE6 and 7.

How about finalizing the specification so it can be deployable in the real world before figuring out how to get rid of browsers that don't support proposed standards that are still under development and could change between now and when they finally reach recommendation status?

[felgall]

How about finalizing the specification so it can be deployable in the real world before figuring out how to get rid of browsers that don't support proposed standards that are still under development and could change between now and when they finally reach recommendation status?

I think that there are two totally separate issues getting confused with one another here. The standard that IE6 and 7 don't support which is the reason why people are looking forward to their disappearance is CSS 2.1. That standard was made final quite some time ago and has no connection with the HTML 5 standard which is still at an early draft stage. Any discussion of those issues is irrelevant to this thread.

The tags in HTML 5 could be all completely different by the time it is finally adopted and long before then all of the current browsers will be so far in the past that most people will not even remember their names. After all how many people today remember the browsers WorldWideWeb and Mosaic? As such the support or lack for things currently proposed to be a part of HTML 5 are irrelevant. Any support in current browsers for proposed tags would be more in the nature of an experiment to see whether the inclusion of those tags is going to be worthwhile or not. Hopefully sufficient thought will go into it so that we do not end up with repeats of past mistakes such as frames and the target attribute.