OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

office-accessibility message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [office-accessibility] Re: ODF Test Suite


Hi Dave,

You note below "The point Rich and Peter picked up on, my styles don't 
work in your implementation. Again is that testable?"  Perhaps I'm 
misunderstanding you, but I wonder if you are referring to the bugs we 
discovered in SO/OOo with our guidelines document, and thinking that 
problem arose from Sun and IBM using completely different ODF editors.

I believe the document corruption was caused entirely on my system, by a 
bug in my copy of OOo (and specifically by the copy of OOo on my laptop, 
was was a nightly development build).  I made a bunch of style changes 
when we moved from 12 point to 14 point font for everything, and in the 
process somehow left vestigial style references in the table of contents 
to a style no longer in the document.  OOo shouldn't have let that 
happen.  Then, after further editing, the document got into a state that 
OOo was unable to open.  That too shouldn't have happened.

But I believe this had nothing to do with ODF parsing compatibility 
between two different ODF implementations.


Peter
> On 19/10/06, Pete Brunet <brunet@us.ibm.com> wrote:
>
>
>>  ODF Test Suite - 
>> http://netmoc.cpe.ucf.edu/Projects/OpenDocument/TestSuite.html
>
> Thanks Pete. That looks more like it.
>
> I'm a little disturbed with the way the testing is being done.
> Again it seems to focus on ODF as a display capability, rather than
> as a document format.
> Here is some XML, this is what it should look like.
> No import or export capability - but how to test & what's the spec 
> actually
> require?
> The point Rich and Peter picked up on, my styles don't work in your
> implementation. Again is that testable?
>
> Interoperability isn't mentioned (unless visible differences are found 
> I guess).
> It's using two implementations.
> It's wholly subjective - does this test output look like this?
>
> I'd like to be able to run a suite of tests and see a binary output,
> pass / fail.
> If fail, why.
>
> Am I being too simplistic?
>
>
> regards
>
>



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]