Free Support Forum -

Error calling OCR (Java) on IBM AS/400 (iSeries, IBMi)


We have a trial licence and I’m doing pretty well with it. However I am now trying to do some OCR on a simple TIF file and getting:

WARNING: 1: java.lang.UnsatisfiedLinkError: Platform “os/400-ppc” not supported by class org.bytedeco.onnxruntime.presets.onnxruntim
at org.bytedeco.javacpp.Loader.load(
at org.bytedeco.javacpp.Loader.load(
at ai.onnxruntime.OnnxRuntime.init(
at ai.onnxruntime.OrtEnvironment.(
at com.aspose.ocr.AsposeOCR.edf(Unknown Source
at com.aspose.ocr.AsposeOCR.f(Unknown Source
at com.aspose.ocr.AsposeOCR.RecognizePage(Unknown Source
at com.aspose.ocr.AsposeOCR.RecognizePage(Unknown Source
at com.pacsol.asposeutils.ImageOCR.convert(

I have done some testing on my dev box (x86 Ubuntu 20.04) and now I have run the code on our IBMi running OS 7.3 and Java 1.8 - which gives the problem above. I have done some conversions from docx and tif to PDF okay on the IBMi.

Any clues?

Thank, Mark.


The stack trace shows that the error is related to a dependency i.e. OnnxRuntime. Would you please make sure that all dependencies have been downloaded completely. Also, please share the sample TIFF image with us so that we can test the scenario in our environment and address it accordingly.

Thanks for the quick reply. I can’t upload the tif file because it says I am not authorised. It is just test data that I wrote as a single page in Libre Office, saved as PDF then converted to TIF using Gimp. Also my pom.xml as I am building using maven and the maven-assembly-plugin which I’m certainly no expert at but the other Aspose conversions seem to work okay (eml to pdf, docx to pdf).

My pom.xml could not be uploaded as not authorised so I converted it to PDF:

pom.pdf (15.2 KB)

I hope that helps. I did a bit of investigation and I think that Aspose calls into a C/C++ program using JavaCpp and it is that linkage that does not support os/400-ppc (Power PC). Perhaps I am wrong - I certainly hope so. I think that Linux x86 64 bit is probably the closest.



Thanks for sharing these details.

We need to further investigate the reasons behind this issue and for the purpose, we have logged an investigation ticket as OCRJAVA-82 in our issue management system. We will further look into its details and keep you informed with the status of its rectification. Please be patient and give us some time.

We apologize for the inconvenience.

Much appreciated. If I can help in any way with testing etc. then just let me know.

Thanks very much, Mark.


Sure, we will let you know in case we need some additional information during ticket investigation.

Hello, just wondering what the latest is (because I will definitely be asked today).

Thanks, Mark.


Regretfully, the earlier logged ticket is not yet resolved. It is under the phase of investigation and as soon as it is done, we will be able to share some updates with you. We highly appreciate your patience and comprehension in this matter. Please give us some time.

We are sorry for the inconvenience.

Hi Asad,

Please can you give me an update on this issue. Our customer is planning
to go live at the end of October and we really do need to have an OCR
solution in place before then.

Can you give me any technical updates and some indication of progress
towards a solution.

Thanks for your efforts and time.

Regard, Mark


We have investigated the earlier logged ticket. Please note that our API has onnxruntime library in dependencies, and this library supports only Windows, Linux, Mac. Unfortunately, it doesn’t support IBM. You can read it on their site