Now I'm confused. I looked at the 5.5.1 release notes
https://github.com/Unidata/netcdf-java/releases/tag/v5.5.1 . It
specifically says you still support Java 8 (and will try to until
netcdf-java v6) and that you tested with Temurin 8 from the (Eclipse
Adoptium project), which is what I test with.
And I'll add that my tests found some additional problems similar to the
one I reported previously:
Caused by: java.lang.NoSuchMethodError:
java.nio.ShortBuffer.position(I)Ljava/nio/ShortBuffer;
at ucar.nc2.iosp.IospHelper.readData(IospHelper.java:335)
at ucar.nc2.iosp.IospHelper.readDataFill(IospHelper.java:288)
at ucar.nc2.internal.iosp.hdf4.H4iosp.readData(H4iosp.java:168)
...
Caused by: java.lang.NoSuchMethodError:
java.nio.FloatBuffer.position(I)Ljava/nio/FloatBuffer;
at ucar.nc2.iosp.IospHelper.readData(IospHelper.java:367)
at ucar.nc2.iosp.IospHelper.readDataFill(IospHelper.java:288)
at ucar.nc2.internal.iosp.hdf5.H5iospNew.readData(H5iospNew.java:226)
...
Thanks for looking into this.
On Wed, Dec 22, 2021 at 8:58 AM Bob Simons - NOAA Federal <
bob.simons@xxxxxxxx> wrote:
> (Sorry for so many emails lately.)
>
> I installed netcdfAll 5.5.1. It compiles. But when I ran my regression
> tests last night, my saveAsGeotiff method failed with
>
> java.lang.NoSuchMethodError:
> java.nio.ByteBuffer.flip()Ljava/nio/ByteBuffer;
> at ucar.nc2.geotiff.GeoTiff.writeHeader(GeoTiff.java:250)
> at ucar.nc2.geotiff.GeoTiff.writeMetadata(GeoTiff.java:227)
> at ucar.nc2.geotiff.GeotiffWriter.writeMetadata(GeotiffWriter.java:261)
> at ucar.nc2.geotiff.GeotiffWriter.writeGrid(GeotiffWriter.java:155)
> at ucar.nc2.geotiff.GeotiffWriter.writeGrid(GeotiffWriter.java:101)
>
> When I look in the Java 8 documentation (which I'm still using as it has
> Long Term Support), I don't see a flip() method. With 5.5.1, are you
> stopping support for Java 8? What version of Java is now recommended? Or
> do you think the error message has some other cause?
>
> (I know I'm old, but I liked it when the compiler could tell me if my code
> was basically valid and all the needed components were present. With the
> current trends (in other code, too) I need extensive regression tests to
> check that everything still works when I update libraries. And even then, I
> am much less confident.)
>
>
>