Java 1.6.0 pop up

[tischi]

Hi @ctrueden,

I have two questions:

1. Is it possible to suppress the Console popping up with the message: 3D [dev] 1.6.0-scijava-2-pre11-daily-experimental daily. It is a bit disturbing for some applications...
2. Could you point me to the place in the code where the Java 1.6 is launched? I am just very curious...are you launching a second JVM or how does that work?

[ctrueden]

It's not Java 6... it's Java 3D 1.6.0. Specifically, the SciJava fork thereof.

And I agree the message is annoying and confusing. There are a lot of System.err.println statements in that codebase. Highly likely that one of them is the culprit.

[tischi]

Hi @ctrueden

I did some research and I found the lines in question:

• https://github.com/scijava/java3d-core/blob/master/src/org/scijava/java3d/VirtualUniverse.java#L236

If you look there, the lines are only a System.out.println, so should not cause the pop-up, right?

What is interesting is that for me (using parent-pom version 25.0.0) those lines read:

System.err.println(str);
System.err.println();

This is: Version 4.0.2 of the 3DViewer and version 1.6.0-scijava-2 of the j3dcore library.

So, maybe it is just a matter of updating the 3DViewer and the j3dcore library in the parent pom?

[tischi]

Hi @ctrueden,
Would you maybe find a minute to look at above issue?
Thank you for your time!

[ctrueden]

It looks like with scijava/java3d-core#1, @haesleinhuepf fixed the System.err issue in Java 3D.

As for 3D Viewer, there are only four places that System.err is used:

$ git grep System.err
src/main/java/ij3d/ImageWindow3D.java:     System.err.println("Java3D error: "
src/main/java/ij3d/ImageWindow3D.java:     System.err.println("Java3D < 1.5 detected");
src/main/java/ij3d/UniverseSettings.java:  System.err.println(e.getMessage());
src/main/java/ij3d/UniverseSettings.java:  System.err.println(e.getMessage());

I do not think any of these are common. And if they do occur, I think having the console pop up is the right thing to happen anyway.

So I guess we just need to cut a new release of j3dcore, and update Fiji to include it.

[tischi]

So I guess we just need to cut a new release of j3dcore, and update Fiji to include it.

Yes, I came to a similar conclusion:

So, maybe it is just a matter of updating the 3DViewer and the j3dcore library in the parent pom?

Is that something that could happen within the next few weeks? I guess you are the one who would have to do it? Or does, e.g. @haesleinhuepf, also have the power to do so? :-)

[ctrueden]

I have not released the SciJava Java 3D artifacts in ~3 years, and they are structured differently than most SciJava ecosystem projects (e.g. they do not extend pom-scijava and do not use Travis CI). And 3 years is beyond the edge of my human memory as well as my shell history log. 🙁 So it's taking me awhile to figure out how to even do it. But I'm trying.

[tischi]

Thank you so much! If it is too much of a headache, maybe just leave it. The pop-up is a bit annoying but not a show stopper. On the other hand, maybe it is good to properly integrate the Java 3D artifacts anyway?! I cannot judge this, as I am not into these dev-ops aspects of Fiji. Anyway: thanks a lot already for looking into it!

[ctrueden]

I tried 6 or so times today to release it, and got a little farther each time. My last attempt still failed, getting as far as the GPG key signing timing out. How can something like signing a GPG key fail with a timeout, you ask? I dunno, why are ~/.gnupg/S.gpg-agent* sockets? Why does gpg fail with Inappropriate ioctl for device unless you export GPG_TTY=$(tty) first?

I guess I should just switch over to the known-to-work pom-scijava-based approach and let Travis do it. Will try to work on it again soon.

[tischi]

OMG, that sounds like real "fun".....

[imagesc-bot]

This issue has been mentioned on Image.sc Forum. There might be relevant details there:

https://forum.image.sc/t/change-font-size-for-the-axes-labels-in-3d-viewer/68689/6