Re: Version indicators (was Re: [public-webapps] Comment on Widget URI (7))

On Dec 30, 2009, at 17:54 , Karl Dubost wrote:
>> - this document is feasibly valid for specification versions 1 through 8 through the application of the ignore-unknown rule to features (and be able to list which features would which version);
>> - this document is strictly valid from version 9 on.
> 
> What would it solve for the person writing a document?
> 
> There are two main usages for validators:
> 	1. Checking if you author correctly.
> 	2. Checking if the document respect an intended semantics.

It would tell you what version you're authoring against. That really shouldn't be something that authors should have to care about — if anything it's too complex. There are very few people out there who could look at an HTML document and tell you correctly whether it's 5, 4, 3.2, etc.

>> I think that this is probably best addressed through a must-understand flag. Sometimes I may use a new element and if you ignore it I don't care.
> 
> Sometimes spec writers change the semantics of an element, which makes it harder to create a stable environment.

Well, spec writers may indeed do many silly things. That's one reason why we talk about writing up architectural advice :)

> Versioning has two very different consequences if we are talking about authoring or consuming.

Certainly, but I fail to see how version indicators help with any of the authoring-related issues: there are too many cases in which authors will get version indicators wrong for it to be a useful tool.

-- 
Robin Berjon - http://berjon.com/

Received on Monday, 11 January 2010 16:19:52 UTC