I'll accept the blame for epubs that don't validate and try to explain what's happening.
To load and save files in DOC, DOCX, ODT and EPUB formats Buzzword uses a third party library called Aspose.Words for .NET. I lead the team that develops Aspose.Words. Please note, that I do not represent Adobe, this post is my personal views only.
As far as I can tell epubs do not validate now only because Aspose.Words exports "XHTML transitional", whereas the EPUB validator requires "XHTML strict". Commonly occuring HTML things like outputting "td.bgcolor" cause validation errors. I simple struggle to find the explanation for EPUB to require strict XHTML and as you can see you can read such EPUBs on most readers correctly.
The good thing is that we are now working to make Aspose.Words produce EPUB that validates and you should see that in one of the next Buzzword releases when they upgrade to use newer Aspose.Words. Don't forget that the whole Buzzword is still labeled as "Beta".
|