Le 2020-03-23 15:00, Alexander Pevzner a écrit :
Hi Thierry,
On 3/19/20 11:54 PM, Thierry HUCAHRD wrote:
I'm currently working on ADF. The idea being to have a complete
support.
The support provided by sane-airscane ADF does not work for Canon IR
ADF scanners.
I've reviewed your ADF implementation with big interest. Nice to meet
my own code in somebody else's project without any reference to its
origins :-)
Hi Alexander,
I've been letting you be aggressive for a while now!
he's an intern who wrote the eSCL backend, he did a great job.
ADF is a working copy.
You examine my code to find the difference that makes the Canon IR
scanners work and at the same time you accuse me of plagiarism.
I agree with Till, community work is good for everyone!
Knowing that escl is defined, we don't invent anything.
Many things are very similar in our code.
I've decided not to answer any more questions unless you're
constructive!
Sincerely
I have a question and a couple of comments.
1) Is it correct, that this line, commented out in my version but
present in your version, is sufficient for Canon IR to work?
<scan:InputSource>%s</scan:InputSource>
2) In your merge request comment you wrote, that your version scans
only the first document from the ADF. From my experience (probably, it
depends on hardware), it scans the entire ADF load, even if software
doesn't fetch ready pages. I bet there will be a sufficient amount of
scanner models that will hang or crash when memory is full - this is
much better to load as much pages, as scanner sends, even if user
doesn't read them.
3) When scan to PDF, duplex ADF sends 2-page PDFs, so it is better to
be prepared to it. I don't have this problem, because I intentionally
don't implement PDF, as I think that benefit of having 1-bit B&W is
low, while added dependencies (and memory overhead) are unreasonably
high.
