#36 Code 128 check failure / warnings


hi Agner-

Thank you for sending this! I have been able to re-create the failure, and
it is obviously a bug (those assertions should never fire). Would you mind
opening a bug tracker?
you can also attach your images there if you like...

I will certainly get this fixed.

Thanks again,

On 03/07/2010 02:48 PM, Agner wrote:
> > Hi, i used e-mail instead of form to be able to attach 2 files, not that
> > info is confidential.
> >
> > windows 0.10 ver.
> > windows vista,
> >
> > Program runs without problems
> >
> > attached file gives 4 warnings, and reads.
> > CODE-128:05000002131271101
> > CODE-128:05000002131270101
> > correct except the letter/101, i have a regular barcode scanner that
> > reads it correct.
> >
> > tiff is scanned at 300dpi (document scanner max) and barcodes are
> > labels, comming from 3.party and cant be made better/different. But
> > seems like its a checknumber problem rather than a quality problem.
> >
> > Adding -q only removes CODE-128: not the warnings, and adding >log.txt
> > only puts the code in the log.txt file not warnings so not much of a
> > log, i had to screendump to show errorcodes.
> >
> > A workaround is add >code.txt to commandline, ignore warnings, and
> > remove last 3 digits. But that seems to be inviting problems, thus
> > making it only a typing helper, not a sure identification process. And
> > would be easier to just capture output, not having it to write a file to
> > read.
> >
> > attached is a tiff, cropped to keep size down on mail while still
> > 300dpi, and a jpg screendump of the warnings.
> >
> > best regards
> >
> > Agner
> > Denmark


  • Nobody/Anonymous

    cropped, gives warnings, and no letter check

  • spadix

    spadix - 2010-03-12
    • labels: --> Scanning/Decoding
    • milestone: --> version_0.10
    • priority: 5 --> 6
    • assigned_to: nobody --> spadix
    • status: open --> open-fixed
  • spadix

    spadix - 2010-03-12

    ok, I have cleaned up these assertions (pushed to hg, rev d31d75b149cb).

    > correct except the letter/101

    Unfortunately, it seems that the symbol does not contain the data you are expecting. I have verified that zbar is correctly decoding the results you reported:

    I cannot explain why the human readable interpretation is truncated or where the letter comes from; this representation is not part of the Code 128 spec, so is either something specific to your application, or possibly the symbol is incorrectly encoded.

    > But that seems to be inviting problems, thus making it only a typing helper, not a sure identification process.

    The decode is correct and as reliable as Code 128 can be (which is quite good compared to other linear symbologies). How you use the data contained in the barcode is up to your application.

    It is worth noting that these warnings should not affect the reliability of the results; you may continue to use the current version until the fix is available in the next release.

    thanks again for reporting this!


Log in to post a comment.

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:

JavaScript is required for this form.

No, thanks