Discussion:
Running an external JBIG2 encoder if one exists
(too old to reply)
Sean Whitton
2018-05-25 11:03:52 UTC
Permalink
Hello,

An upcoming release of OCRmyPDF, which I maintain in Debian, will call
jbig2 if it can be found on PATH, or gracefully degrade. On Debian,
this won't do anything, since we don't have that package.

I don't have any trained intuitions about patent law, but I assume that
the fact that OCRmyPDF contains code that looks for a jbig2 executable
is not any kind of patent violation?

I note that the US patents on jbig2 encoding have expired.

Please CC me.
--
Sean Whitton
Ian Jackson
2018-05-25 18:06:30 UTC
Permalink
Post by Sean Whitton
An upcoming release of OCRmyPDF, which I maintain in Debian, will call
jbig2 if it can be found on PATH, or gracefully degrade. On Debian,
this won't do anything, since we don't have that package.
I don't have any trained intuitions about patent law, but I assume that
the fact that OCRmyPDF contains code that looks for a jbig2 executable
is not any kind of patent violation?
It seems unlikely to me that this would be a problem.

Not that I think you have done anything wrong, but this is probably a
good time to remind everyone of:
https://www.debian.org/legal/patent
https://www.debian.org/reports/patent-faq

If you want more reassurance than "seems fine" from someone like me, I
guess you should email ***@d.o with the info requested in
/legal/patent.

Ian.
Sean Whitton
2018-05-25 18:33:04 UTC
Permalink
Hello,

Thank you both for your answers.
That said, nobody can sue you over an expired patent, so it sounds
like you shouldn't be particularly worried about those.
I should have said: upstream pointed out that there may be non-US
patents that have not yet expired.
--
Sean Whitton
Florian Weimer
2018-05-26 22:02:30 UTC
Permalink
Post by Sean Whitton
An upcoming release of OCRmyPDF, which I maintain in Debian, will call
jbig2 if it can be found on PATH, or gracefully degrade. On Debian,
this won't do anything, since we don't have that package.
We have JBIG2 decoders, I think. Surely that is sufficient for OCR?
Sean Whitton
2018-05-27 14:39:43 UTC
Permalink
Hello Florian,
Post by Florian Weimer
We have JBIG2 decoders, I think. Surely that is sufficient for OCR?
I don't know what ocrmypdf is doing with jbig, but it's the encoder it
tries to shell out to.
--
Sean Whitton
Loading...