PDF test files (Was: License review in progress)

[Jukka Zitting <ju...@gmail.com>]

Hi,

Another open licensing issue I've come up with is the set of test PDF
documents in pdfbox/trunk/test. Quite a few of those documents seem to
come from various places and there's no indication whether the people
who submitted them had the rights to allow redistribution of the
documents under a permissive open source license. I would assume that
most of the test documents were just used to illustrate particular
issues with little or no consideration of them being later distributed
as part of PDFBox.

Assuming this understanding is correct, we need to figure out what to
do with this test suite. Having a comprehensive test suite with
real-world documents is a great asset, but also a licensing issue. For
example, does Premera Blue Cross from Seattle, WA consent to us
redistributing one of their forms (see input/c21-5916.pdf) under ALv2?
Most likely they couldn't care less, but we need to be prepared if
they do. If we don't have permission from the original authors of the
documents, then we can't distribute them in Apache PDFBox.

The basic option is to simply drop all the test documents for which we
don't have a trail to the required license. That satisfies Apache
policies, but is hardly a sound decision from a quality assurance
point of view. A better option would be to find or create acceptable
replacements for all the troublesome test documents. We could also
play some games with keeping the test documents in a "Tests for
PDFBox" project outside Apache, but I'd rather avoid that if possible.

BR,

Jukka Zitting