Hi Asher,
please use public mails (i.e. reply-to-all) to discuss Open Source. I will
work via private mails for you if you insist, but you will have to pay me
what my work is worth in that case.
On Wed, 25 Jun 2014, Asher Wright wrote:
> Thanks for your quick reply.
You are welcome.
> Unfortunately, this is still not working for me. I type "Fiji -macro
> <Path>" and it works well (with output "07:26:35.657
> [SciJava-4ef18d37-Thread-0] DEBUG loci.common.NIOByteBufferProvider - Using
> mapped byte buffer? false")).
>
> But then when I do this: "Fiji --headless -macro <Path> -batch" I get a
> null pointer exception.
Please note that the headless mode is a hack intended to get as much as
possible working in truly headless environments such as a cluster.
However, the design of ImageJ 1.x does not allow a perfect headless mode:
most of the code assumes that a graphical desktop is available.
Having said that, it still might be possible to support your use case. But
it is too hard to guess exactly what your macro is doing that causes this
exception...
Therefore, if you want your use case to be supported, please make a
minimal example macro that reproduces the problem. In other words, make a
copy of your macro and then reduce it as much as possible so long as the
exception still occurs in headless mode (if it contains a loop, try to
remove the loop and instead only execute the first iteration, if it tries
to load a 500MB file, test whether it also fails using the Clown or the
Blobs sample image, etc).
Once you have such a minimal macro, it should be relatively easy for me to
squeeze this project in and avoid the NullPointerException.
> Do you have any more ideas?
Lots. Tons. Too many, in fact. :-)
Ciao,
Johannes
--
ImageJ mailing list:
http://imagej.nih.gov/ij/list.html
Locked
