Trouble with BatchPreProcessing script

[EorEquis]

Hello all,

Of late, I've had some intermittent and variable trouble with BPP.  It almost always occurs after integrating flat frames, and the error message is usually different, but almost always includes some reference to [this._____] is undefined.

The latest :



Can anyone point me in the right direction to start troubleshooting?

[EorEquis]

I should also mention that USUALLY (not always, however) simply re-running the script with identical files and settings works fine.

[EorEquis]

Update from this morning.  Another error, again right after "End integration of flat frames" appears in the Processing window.  This morning's error is :

this.setValue is not a function

Once again, simply clicking "Run" again worked fine the next trip through.

[EorEquis]

Another error this morning.  it now seems to be happening every time I run BPP.



Still looking for input.

[bitli]

I have seen the

Code: [Select]

this.setValue is not a functionat times. But I was not able to reproduce it in any systematic way. I did clean the cache and restart PI, this seems to have solved the problem. 
-- bitli

[EorEquis]

Hrm.  Have to give that a try, I suppose.  How does one "clean the cache"?

[bitli]

See attached screen dump.
-- bitli

[Juan Conejero]

This is a very strange problem that should never happen under normal conditions. It could be a bug in the BPP script, but I cannot reproduce it on any platform with any of our test data sets.

This looks like some kind of corruption of the JavaScript engine. Have you tried uninstalling/reinstalling PixInsight?

[bitli]

On my system it occurred only when I did first quite a few other operations.  Restarting PI in general also helped (I do not know why clearing  the cache helped, anyhow it did).  But I can work days without that problem.
One of the case which seems to cause problems is when creating previews on the Blink (and doing a few operations like Blinking other stuff, moving the preview to another window, and so on...).  I had 'memory access violation' a few times after this kind of operations, but I have unfortunately no way to reproduce it systematically. After a memory violation it is usually much safer to restart PI.
I'll tell you if I find a repetitive way to crash PI  (but in general this version is pretty stable).
-- bitli