Re: [opendocument-users] simple OO.org document goes awry in MS Office2007 w/SP2 - what went wrong?

[Svante Schubert <Svante.Schubert@Sun.COM>]

Alex,

Alex Brown wrote:
> Are you aware that the odftoolkit.org has special-casing code to ignore certain faults in documents produced by OpenOffice.org? (See for example from lines 55 and 79 at http://is.gd/13dM8 ). The full list of such changes is at http://is.gd/13dYp  .
>
> The Office-o-tron validator is written to carry out a "pure" validation without favouring any particular implementation. You can see the source at:
>
> http://code.google.com/p/officeotron/source/browse/#svn/trunk/src/org/probatron/officeotron
>
> Comments on this are welcome!
>
>   
Let me jump in as Michael Brauer, owner of the odftoolkit.org validator 
is not present this week.

ODF validators must provide the same results regarding well-formedness 
and validity, there is no 'pure' validation.
AFAIK there is no test environment for validators either that really 
justifies the word 'proven', but I am sure John simply gave his personal 
suggestion.

Certainly there is no 'favouring' of any particular implementation in 
the validator of odftoolkit.org, otherwise it would be a bug of the 
validator.
The fault you are mentioning is not a validity fault (see 
http://www.w3.org/TR/REC-xml/#dt-doctype), but a dead link, which is 
'confusing' the underlying/wrapped Multi Schema Validator and would 
block the complete validation by looking for a non provided DTD. The 
problem of the missing reference (not an ODT issue) is further not 
concealed, but given out by the validator when choosing the logging 
level 'all'.

I am happy to see that more validators step into the scene, esp. with 
different approaches as yours using Schematron, which gives ODF higher 
reliability and users a choice.

Kind regards,
Svante