Re: ISSUE-54: doctype-legacy-compat

On Tue, Jan 20, 2009 at 1:22 PM, Karl Dubost wrote:
>
> 1. Which are the tools which modify their behaviour according to a versioned
> doctype?
> 2. Which issues will these tools have with no versioned doctype ?
>
>
> # Browsers
>
>  This is case is pretty much set and had been ad nauseam discussed.
>  <!DOCTYPE html> triggers the standard mode.
>
> # Authoring tools
>
> # Converters
>
> # Tidying tools

All these tools should use the "current" version of the spec (current
here meaning either "last" or "most widely deployed" –which we expect
to be the same–, at the time the tool is released, of course), of
course with "adjustments" for the deployed "code base" (workarounds,
etc.).
I'd expect the latest version of an authoring tool, XXX-to-HTML
converter or tidying tool today to output markup that is understood by
today's UAs.

I mean, there's no point in producing HTML 3.2 today, unless you
target yesterday's UAs. In that case, just use a tool from yesterday
(actually, you could also use some "profile" of HTML 4 with a tool
from today, but given that HTML5 removes <font>, you won't be able to
produce such a document using a tool of tomorrow).

Overall, the whole thing is which UA(s) you're targetting, not which
HTML spec version. If you only target Internet Explorer (in an
intranet), there's no reason you wouldn't use <marquee> and <bgsound>,
or onmouseenter/onmouseleave, or <?namespace>s and HTC components.

As for HTML-to-XXX converters, they're not much different than browsers.

HTML-to-HTML converters fall (imo) in the "tidying tools" category,
they should produce HTMLcurrent (eventually with workarounds for known
UA bugs); and they might very well take the source HTML spec version
as a parameter: when HTML8 will be out, you could run HTMLTidy with
"treat input as HTML5 and output HTML8".
Same for authoring tool when it comes to opening an existing file (vs.
what it outputs).

> # Semantic parsers (if any)

Are they any different from browsers?
I mean, if a document is labelled as HTML 4.01, would such a parser
ignore an <article> or <dialog>?



-- 
Thomas Broyer

Received on Tuesday, 20 January 2009 22:59:05 UTC