W3C home > Mailing lists > Public > public-appformats@w3.org > September 2006


From: Mark Birbeck <mark.birbeck@x-port.net>
Date: Fri, 1 Sep 2006 17:30:44 +0100
Message-ID: <640dd5060609010930u4a664ed7r6a60d4ec31fdc4b9@mail.gmail.com>
To: "Anne van Kesteren" <annevk@opera.com>
Cc: public-appformats@w3.org, www-forms@w3.org


My point was different--apologies for not explaining it more clearly,
but I was actually responding not to the idea that processing 'stops'
but to the idea that the user doesn't get anything other than the
error message. I'll explain.

In Firefox, Opera and IE, the error messages used when processing an
XML document appear in the same location as the document rendering
would. So your example of "XML parsing failed" would appear directly
in the browser. This means that you get an either/or approach--if the
document is completely ok you get to read it, but if there is any
problem at all, you see nothing but the error message.

In Sidewinder, we have additional status windows that indicate what is
happening (loading document from A... parsing... validating... etc.).
If we see a document that 'claims' to be XHTML, but is not well-formed
or fails validation against the schemas, we can 'fallback' to HTML in
the viewing window, because we have the status windows available to
show error information. A user (most probably a developer or author)
can do whatever they like with this 'additional'
information--including continuing to read the news story that has just
appeared in view. :)

Our conclusion was that for an ordinary user the error messages are
pointless, so you might as well just 'fallback' to HTML rendering. The
error messages help authors and developers, but even then, it's more
useful to show something than nothing--so again a 'fallback' to HTML
is useful.



On 01/09/06, Anne van Kesteren <annevk@opera.com> wrote:
> On Fri, 01 Sep 2006 14:19:34 +0100, Mark Birbeck <mark.birbeck@x-port.net>
> wrote:
> >>   5. "XML parsing failed: syntax error (Line: 8, Character: 0)" in
> >> Opera;
> >
> > That's interesting...does failing to parse properly necessarily have
> > to prevent rendering?
> That's not really the point. The mistake could be on the first line. For
> example, lots of pages have whitespace before the XML Prolog or don't
> declare a namespace or insert some <meta> elements that are not "self
> closing." And whether or not we render half of the page, it would still be
> worse than in Internet Explorer. Not to speak of the differences in
> scripting and such.
> --
> Anne van Kesteren
> <http://annevankesteren.nl/>
> <http://www.opera.com/>

Mark Birbeck
x-port.net Ltd.

e: Mark.Birbeck@x-port.net
t: +44 (0) 20 7689 9232
w: http://www.formsPlayer.com/
b: http://internet-apps.blogspot.com/

Download our XForms processor from
Received on Friday, 1 September 2006 16:31:02 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:50:05 UTC