this topic was brought up during the PDFDays in Cologne by Frank Easton, HUK Coburg who found that converting/validating with one software vendors tool the results are always fine but mixing the tools there are PDF/A validation errors even today. On the other hand the situation improved a lot by software vendors working together in organizations such as the PDF Association.
" … due to the evolution of the PDF/A spec, documents created earlier were no longer consider valid. … " I think that this is not the primary reason. There wer only minor corrections/clarifications wrt the standards document. The reason being that the software to validate was enhanced over time so documents which validated might no longer validate with newer versions of the same software. But keep in mind that this page and the findings were done in 2009. So the overall situation improved a lot - especially around PDF/A-1b which is now well understood and there should be only minor differences (although the vendors customer might expect there are none - see above). Now to the NPEs in preflight - I agree there should be none. Maruan Am 03.11.2014 um 19:00 schrieb Tilman Hausherr <thaush...@t-online.de>: > I've run preflight on my test files, with the thought that there should not > be any NPE. It is not looking good, there is at least one NPEs issue coming. > > And when that is done, one might check the "Bavaria tests" > http://www.pdflib.com/knowledge-base/pdfa/validation-report/ > > The segment "Evolution of PDF/A validation" is sortof hilarious: due to the > evolution of the PDF/A spec, documents created earlier were no longer > consider valid. > > And there's a familiar name mentioned on the page :-) > > Tilman > >
