[odf-discuss] Three kinds of tests

[Hanssens Bart]

Daniel Carrera wrote:

> I don't know whether the fact that it is converted from .doc
> is good or bad. My opinion (for what it's worth) is that it is an 
> important dimension because in the real world a lot of ODF documents 
> will start out by being converted from .doc.

I've been testing doc/xls/ppt to the last 6 weeks or so for the Belgian
government, so I think I might answer this one :-)


There are two aspects: quality of conversion from format X to ODF (this
should not part of the ODF conformance tests) and using the resulting
ODF documents as a starting point for real life tests. 

They provide very good input for real life tests (although it shouldn't
matter if the document is generated by a converter or written "natively"
in ODF), as these documents tend to be dirty: often reused / re-edited,
superfluous mark-up inserted by converter tools etc.
But I wouldn't use them for type 1 tests (atomic / unit tests)


Best regards,

Bart