OASIS Open Document Format for Office Applications (OpenDocument) TC

Hi Michael,

Well lets discuss this in the next TC call then. 

Wrt. to Olivers proposal the interresting question is what do we expect todays readers to do with it... 

When I read the current ODF spec it seems to me that the document is simply not conformant and thus everything can happen....

Anyway: Was just curious whether we say something somewhere in the spec about invalid values already.

You may wanna take a look at "Rules for handling parsing errors" of the CSS spec:
http://www.w3.org/TR/2007/CR-CSS21-20070719/syndata.html#parsing-errors

It would be great to have something similar in ODF.

E.g. when we would use the CSS rules for handling parsing errors the answer to the backward compatiblity question of Olivers proposal would be very easy:


So it would treat the positions as if none was given.

HTH,

~Florian


>>> Michael Brauer - Sun Germany - ham02 - Hamburg 

Hi Florian,

I haven't set up the agenda for the next call, but we have already a lot 
of topics on the agenda. Some of them are waiting for a discussion for a 
long time already, and you have requested the discussion of several 
proposals, too.

We won't have the time to discuss all of these things in the next call. 
I will keep this in the loop of agenda items, but ask for your 
understanding if this will have to be scheduled for a later call.

One more remark below.

Thank you

Michael



Florian Reuter wrote:
> Hi Michael,
> 
> Well lets discuss this in the next TC call then. 
> 
> Wrt. to Olivers proposal the interresting question is what do we expect todays readers to do with it... 
> 
> When I read the current ODF spec it seems to me that the document is simply not conformant and thus everything can happen....
> 
> Anyway: Was just curious whether we say something somewhere in the spec about invalid values already.
> 
> You may wanna take a look at "Rules for handling parsing errors" of the CSS spec:
> http://www.w3.org/TR/2007/CR-CSS21-20070719/syndata.html#parsing-errors
> 
> It would be great to have something similar in ODF.
> 
> E.g. when we would use the CSS rules for handling parsing errors the answer to the backward compatiblity question of Olivers proposal would be very easy:
> 

If my understanding of Oliver's proposal is correct, then all values are 
valid values. It is only the combination of particular attributes and 
their values that was not permitted. So, stating that user agents must 
ignore a declaration with an illegal value won't help in this case.

-- 
Michael Brauer, Technical Architect Software Engineering
StarOffice/OpenOffice.org
Sun Microsystems GmbH             Nagelsweg 55
D-20097 Hamburg, Germany          michael.brauer@sun.com
http://sun.com/staroffice         +49 40 23646 500
http://blogs.sun.com/GullFOSS

Sitz der Gesellschaft: Sun Microsystems GmbH, Sonnenallee 1,
	   D-85551 Kirchheim-Heimstetten
Amtsgericht Muenchen: HRB 161028
Geschaeftsfuehrer: Thomas Schroeder, Wolfgang Engels, Dr. Roland Boemer
Vorsitzender des Aufsichtsrates: Martin Haering

On Friday 04 July 2008, Florian Reuter wrote:
> > Options:
> > a) Treat it as if the attribute was not there
> > b) Try to preserve the value
> > c) Treat is as if the attribute was there but overwrite the invalid value with a "default" value

I don't see the difference between a and c. In both cases the application will use whatever
default value it normally uses for this attribute -- which is IMHO the only sensible thing to
require from an implementation [more precisely not requiring anything when it comes to the
topic of unhandled attribute values. if they're unhandled, then by definition we can't expect
them to be handled ;)]. b) is a common utopia solution which never solved any problem
nor can be implemented fully.

-- 
David Faure, faure@kde.org, sponsored by Trolltech to work on KDE,
Konqueror (http://www.konqueror.org), and KOffice (http://www.koffice.org).

2008/7/4 David Faure