These are chat archives for fiji/fiji
+ curl -O https://update.imagej.net/bootstrap.js && jrunscript bootstrap.js update-force-pristinto work with openjdk-8?
update-force-pristines in a row.
mvnwith the scijava-maven-plugin. But it doesn’t do all the needed things yet.
docker run --rm -v /Users/at22/PIPELINE/stitching/img:/img fiji:latest fiji-linux64 --headless -macro /img/macro.ijm /img/test#/img/out
Script1.groovy: 1: expecting '!', found '@' @ line 1, column 2. #@File(label="Select your directory with your exported images", style="directory") theDir ^
Step 7/11 : RUN bash ImageJ.sh --headless --update add-update-site PTBIOP http://biop.epfl.ch/Fiji-Update ---> Running in 7791d6ae90a4 Error: Could not find or load main class net.imagej.updater.CommandLine The command '/bin/sh -c bash ImageJ.sh --headless --update add-update-site PTBIOP http://biop.epfl.ch/Fiji-Update' returned a non-zero code: 1
ImageJ.shis a complex shell script attempting to emulate a bunch of them...
expecting '!', found '@'the problem is that you are invoking an ImageJ1 macro with the ImageJ1-style
-macroflag, which does not understand ImageJ2 script parameters.
--->We will try to work on 1 wells in parallel and 41 extra threads to process your data Processing Well [1, 2]
ImageJ.shis: it works on POSIX-friendly systems that don't have a native ImageJ launcher executable. :-)
ImageJ.shscript ultimately invokes
java, so if there is some weirdness with the ImageJ native executable linking to
libjvm, and it also fails to "fall back to system Java" (i.e., call
javaw) for some reason, then the shell script is a potential way forward.
@frauzufall @HedgehogCode I ran across this in my notes from last December:
Plan for TensorFlow: ship native libs both CPU and GPU in
lib/<platform>directly; load them with
System.loadLibraryso that TensorFlow Java library is already happy. Name
libtensorflow_gpufor the GPU version; load that by default first; if it fails, load CPU. UNLESS! a system property is set to "cpu" then use that preferentially.
Did that work get completed? Or, does it make sense to file an issue in imagej/imagej-tensorflow so we don't forget the details?
fiji-win32.exe done [ERROR] Error updating java.lang.RuntimeException: Incorrect file size for /home/fiji/Fiji.app/fiji-win32.exe: 964 (expected 382976) at net.imagej.updater.Installer.verify(Installer.java:293) at net.imagej.updater.Installer$VerifyFiles.itemDone(Installer.java:270)
Downloading... (fiji-win32.exe) [ERROR] Error updating java.net.SocketException: Connection reset at java.net.SocketInputStream.read(SocketInputStream.java:210) at java.net.SocketInputStream.read(SocketInputStream.java:141) at java.io.BufferedInputStream.read1(BufferedInputStream.java:284)
tensorflow/NativeLibrary.javato enable loading the GPU if possible and the CPU as fallback but never created a PR there... :( It's on my list though..