Re: [CSS3] Generated content: environment variables

fantasai wrote:
> a) Not all CSS applications are web browsers. PrinceXML, for example,
>    enables the use of CSS for publications like books, manuals, and theses.

Yes, but last I checked this is the World Wide Web consortium.

As an analogy to your example, not all DOM applications are web browsers.  Some 
are server-side Java apps.  Catering to them has made the DOM specifications 
much more complicated, more vague, harder to implement in a web browser, and 
generally lower quality than they would have been with a sharper focus.  Let's 
try to avoid that sort of thing here.

> b) In a web browser the user should be allowed to override the author's
>    header styles, but it is quite reasonable for the author to be able
>    to specify different header content and style if the user doesn't
>    particularly care.

There's no way to tell on a per-page basis whether the user cares.  Unless you 
think browsers should ask every time the user prints?

>    For example, the CSS specifications could print the
>    title, page number, url, and official date of publication rather than
>    the date of printing.

I don't see why you need "rather than" instead of "in addition to.

>    Driving directions could specify headers with the
>    the starting address, ending address, site homepage URL, and date of
>    retrieval and leave out the cropped, unreadable, and useless CGI URL.

Again, I'm find with all that if you leave out the "leave out" part.

-Boris

Received on Friday, 2 November 2007 17:07:53 UTC