Especially since as of now, the Scala notebook is really just a (fancier) port of an existing ImageJ tutorial. Keeping them together might (hopefully) help keep things in better sync.
I really like this idea. The only case I can imagine with notebooks not being appropriate for the tutorials repos is if the notebook is intended for something other than illustrating how to do stuff... like if they somehow start being used to do "real science" instead of training.
And in that case, we can create new repo(s) then, if it happens.
I'll try to be more responsive tomorrow so we can discuss further during work hours.
and a related thing, if i'm developing a script that uses annotations to grab inputs and testing it out by coding it in a separate project by symlinking the script into plugins/Scripts, then I can test code logic changes because the script is reread from file on each call, but the annotations seem to get read at menu generation time, so i end up restarting Fiji if I modify the input annotations to the script
loving this way of writing scripts for the biologists though, just making observations about the process
@kephale I made a short list in my workshop with @bnorthan in September: https://imagej.github.io/presentations/2015-09-04-imagej2-scripting/#/3/2
Please feel free to transcribe that onto the wiki! I suggest the script parameters page.
Oh sorry, I missed your reply, @rimadoma. Thanks.
It would be nice at some point to update this script: https://github.com/scijava/scripting-javascript/blob/scripting-javascript-0.4.3/src/main/resources/script_templates/Intro/Widgets.js
To include more things closer to the WidgetDemo.java I linked above.
In particular, multiple choice widgets (since those work now), and the message visibility thing.
done by lunchtime
Ok signing off for today. See you!
Next, we need to: 1. release imagej-launcher 4.0.4 (fixes Win10 32-bit, thanks to @stelfrich); 2. release TrackMate, Simple_Neurite_Tracer and IO; 3. update imagej/imagej and fiji/fiji accordingly (which I guess means another pom-imagej and pom-fiji, oops...); 4. test and upload all to Java-8.
This update will: 1. fix Win10 32-bit memory allocation as mentioned above; 2. fix the preprocessor cancelation bug; 3. eliminate the batik uber-JAR, closing fiji/fiji#83 and addressing this forum post; 4. fix the "1 errors during plugin discovery" message some people have noticed, and relatedly, fix execution of ImageJ1 plugins from the script editor and via
--run again.
I tried to run SystemInformation, but could not have a proper Service set to initialise from the main method. They are all null.
I have tried to run net.imagej.Main, etc up to no avail. I remember fighting this a while ago but could not retrieve the way to make it work.
This reminds me a bit of Quantum Theory: a snippet that creates the conditions in which you cannot determine why it is working.
@hinerm There are some remaining class clashes I am working out across Fiji relating to the XML classes. However, I think it is safe to upload the new scijava-common to the Java-8 update site any time. The imagej/imagej and fiji/fiji master branches will "catch up" to it very soon.
I will be in to the office ~1:30pm today, but feel free to upload it earlier if you wish.
Also, if you could cut the new release of imagej-launcher, that would be awesome. I didn't do it yet.
@kmader ok i'm stumped about the notebook displays (given that I'm puttering at this in Clojure), I see that the jupyter-scala package is providing the display function but it is behind so many layers of abstraction that i'm having trouble figuring out where the actual code is
although it looks like the scala world solved this, do you happen to know anything about how you trigger an image to be rendered, like does one actually need to hook into the Jupyter ZeroMQ instance and send a MIME encoded image via JSON?