W3C home > Mailing lists > Public > www-svg@w3.org > March 2009

Re: Element Whitelisting

From: James Graham <jgraham@opera.com>
Date: Wed, 25 Mar 2009 10:57:11 +0100
Message-ID: <49C9FFF7.4080409@opera.com>
To: Rob Sayre <rsayre@mozilla.com>
CC: Doug Schepers <schepers@w3.org>, Sam Ruby <rubys@intertwingly.net>, public-html@w3.org, www-svg <www-svg@w3.org>
Rob Sayre wrote:
> On 3/25/09 3:30 AM, Doug Schepers wrote:
>> In particular, I was personally disappointed that the request to call 
>> out unquoted attributes and case-folding as parse errors was not 
>> incorporated; 
> 
> It's incorrect for the HTML spec to call them parse errors if they can 
> be interoperably parsed by HTML parsers. It's not a good use of 
> Mozilla's resources to spend time debating interoperable errors.

On the basis that things labeled parse errors generally lead to 
unexpected, albeit interoperable, parsing, they are bad for authors. It 
is part of our design principles that we consider the needs of authors 
(above implementors, even) as part of the language design. If you feel 
that it is not a good use of your/your employer's time to debate these 
points, I guess you can just ignore them.

> There is probably a role for a lint tool here, but that requires 
> interaction with users more than standards group involvement.

IMHO such abdication of responsibility would be doing users of the 
language a great disservice.


Oh and on the more technical issue, it seems like the issue of 
non-XML-like SVG is only a problem in the case where the SVG is hand 
authored, uses non-XML-like features, and then later someone wants to 
copy it into an authoring tool that does not support the SVG-in-HTML 
syntax. This seems like a situation that will be rare to begin with 
(since the impression I get is that most SVG is authored in a tool) and 
will disappear over time (as SVG editors start to support SVG in 
text/html). On the other hand, the inconsistency between SVG and 
anything else in text/html is something that would always have to be 
called out explicitly in any tutorial, that would make the platform look 
confusing and inconsistent, and that anyone used to HTML would probably 
get wrong anyway, it doesn't seem like making these cases a parse error 
would buy us much.
Received on Wednesday, 25 March 2009 09:57:55 GMT

This archive was generated by hypermail 2.3.1 : Friday, 8 March 2013 15:54:41 GMT