Bad PDF from new versions of Simple Scan

need to confirm this on version 2.31.91 backport on lucid. the saved PDFs are just blank.

this is still true in 2.32.0-0ubuntu1 backported to lucid. although i doubt it, but can someone please confirm that it isn't present in maverick.

I use simple-scan 2.32.0 under Maverik 64bit.
The PDF Output creates empty PDF Files with a file-size of 50.4 kb.

There is nothing in the .xsession-errors log, and no console output. This makes it hard for me to provide you with useful data to solve this Bug.

In the meanwhile i print the scanned documents as PDF. <- A convenient work around.

Confirmed it in 2.31.91 compiled from sources (I'm using archlinux). Just blank pages in the saved pdf documents.

I can confirm such bug in 2.32.0-0ubuntu1 version on Maverick.

What application are you using to view the output PDFs? Could you please attach a PDF which is showing as white, and provide a screenshot of simple-scan showing what it should contain.

I can confirm this bug. When using text mode, simple-scan shows the scanned page correctly, but when saving as PDF, evince renders it completely empty.

I attach a test scan, with my Canon LiDE 30. This worked fine in lucid and early maverick, it might be correlated to the switching to 4 gray levels?

$ evince /tmp/test.pdf
Error (12878): Unknown operator ',760010'
Error (12893): Unknown operator ',200012'
Error (12904): Too few (2) args to 'cm' operator
Error (12878): Unknown operator ',760010'
Error (12893): Unknown operator ',200012'
Error (12904): Too few (2) args to 'cm' operator
Error (12878): Unknown operator ',760010'
Error (12893): Unknown operator ',200012'
Error (12904): Too few (2) args to 'cm' operator
Error (12878): Unknown operator ',760010'
Error (12893): Unknown operator ',200012'
Error (12904): Too few (2) args to 'cm' operator
Error (12878): Unknown operator ',760010'
Error (12893): Unknown operator ',200012'
Error (12904): Too few (2) args to 'cm' operator

(not sure whether the error messages are related)

The top part is really blank, I scanned a post card from a magazine.

This is a screenshot of simple-scan after scanning the page.

It doesn't depend on the page contents, I tried that with several different pages.

Confirmed, it's a locale problem when writing the floating point transform:

5 0 obj
<<
/Length 45
>>
stream
q
617,760010 0 0 847,200012 0 0 cm
/Im0 Do
Q
endstream
endobj

(i.e. the 617,760010). Workaround would be to use LANG=C simple-scan to run.

Patch that fixes the problem

SRU ack, please upload. Thanks a lot for the fix!

Accepted simple-scan into maverick-proposed, the package will build now and be available in a few hours. Please test and give feedback here. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you in advance!

This bug is okay, it seems, but when I scan, it save me a file which, for an A4 paper for example, isn't entire : (see the attachmant...).

works for even backported to lucid, but i am "only" on de_DE.UTF-8.

@Kévin, does the version 2.32.0-0ubuntu1 currently in maverick already exhibit the same problem or is it introduced by 2.32.0-0ubuntu2 in -proposed ?

Same problem as Kévin. Simple-scan is compiled from source with the applied patch on opensuse 12.2 (x86_64)

Hi, I can confirm the @Kévin's problem.

Saving a scanned document as a PDF file result in a missing top part of the page.

And this is the scanned document saved as a TIFF file (and this is works as expected).

It appears that the bug reported by Kévin comes from the patch itself. The pdf output is OK when simple-scan is compiled from the original source and executed with C locale.

I tried to reproduce the bug with 2.32.0-0ubuntu1 version of simple scan ( actual maverick version, I never had it (neither in photo or text mode). It seems to come from the patch I agree.

Fixed patch - it was reusing the same buffer for the width and height so it would always make the output size square. I happened to test it with a square scan!

Uploaded with fixed patch as simple-scan 2.32.0-0ubuntu3, please accept into proposed.

2.32.0-0ubuntu3 accepted into maverick-proposed, please test.

Works fine now - thanks!

Kévin, Damien, zmoog, can you try the latest version and confirm it fixes the problem for you too with no regression ? Thanks in advance.

That's good, it works fine!!

backported 2.32.0-0ubuntu3 still works fine. thumbs up for moving to -updates.

That's great! Thank you for testing and confirming.
marking as verification-done.

Yep! It works.. thanks..

zmoog
 Il giorno 14/ott/2010 16.21, "Jean-Baptiste Lallement" <
<email address hidden>> ha scritto:
> Kévin, Damien, zmoog, can you try the latest version and confirm it
> fixes the problem for you too with no regression ? Thanks in advance.
>
> --
> Bad PDF from new versions of Simple Scan
> https://bugs.launchpad.net/bugs/644734
> You received this bug notification because you are a direct subscriber
> of the bug.

This bug was fixed in the package simple-scan - 2.32.0-0ubuntu3

---------------
simple-scan (2.32.0-0ubuntu3) maverick-proposed; urgency=low

  * debian/patches/03-blank_pdf.patch:
    - Fix error that caused the output PDF to be square

simple-scan (2.32.0-0ubuntu2) maverick-proposed; urgency=low

  * debian/patches/03_render_bug.patch:
    - Fix crash in page rendering (LP: #645901)
  * debian/patches/04_blank_pdf.patch
    - Fix invalid PDF files generate in locales that use ',' for a decimal point
      (LP: #644734)
 -- Robert Ancell <email address hidden> Wed, 13 Oct 2010 09:27:05 +1100

Copied to natty as well.

Confirming resolution OK

Fixed in simple-scan 2.32.0.1

I have 2.32.0-0ubuntu4 in Maverick but I still cannot get a working PDF.

After I have scanned as text and saved as PDF this is what I get:

thonov@thonov-ubuntu:~/Documents$ file Scanned\ Document.pdf
Scanned Document.pdf: PDF document, version 1.3
thonov@thonov-ubuntu:~/Documents$ ls -la Scanned\ Document.pdf
-rw-r--r-- 1 thonov thonov 118 2011-01-05 13:47 Scanned Document.pdf

Only 118 bytes..?

One more thing, If I do save as then Scanned Document.pdf is pre-filled as filename. However, if I change it to .jpg I get a working JPEG-file.