SGP 2.5.0.20 crashes taking flats

Attached are 11 logs taking flats with both SBIG8300m and Canon 6D this morning with the new ver 20.
Ken it looks like your new code to handle the image display system is back to where we were prior to ver 17.

To summarize:
All the following assume taking biases, darks, or flats. I cannot reliably confirm that I ever had a crash related to image display during a normal evening’s image session, but it is possible.

  1. versions prior to 17 failed very quickly for me. I usually could not go more than a few images without a crash. However, after a while, the program would seem to get into a mode where I could finish 100 or more of the flats or darks I was taking successfully.
  2. versions 17 and 19 worked quite well for me taking biases, darks, or flats, and perfectly for normal imaging.
  3. version 20 again acts the same as the versions prior to 17.

This set of logs

is mostly for ver 20. The last couple document 2 crashes with ver 19 where I collected 200+ flat images. Version 20 would only go a few images before crashing.

OK. Thx for the feedback. I’m a little bummed, but not all that surprised. I’ll do some more thinking on this… I am interested if it made any difference for @spokeshave.

I replied in the other thread that I started. So far, no hangs for me on a sequence of 300 bias frames. It seems to be working well.

Tim

@jmacon

The new method for image preview in 2.5.0.20 is faster, recommended by Microsoft and seems more stable (in general). I know it doesn’t work for everyone (as is evidenced by your recent reports), but I think we will stay with this method.

2.5.0.21 uses some new code (not super thrilled about it) that is hopefully smart enough to figure out which method to use (meaning the 2.5.0.17 method or the new one introduced in 2.5.0.20). Hoping it will work OK for just about everyone. We will still move toward a unified display method.

I hate Windows sometimes…

Ken, you may want to forget the toggle to ver 17 for display.
Yesterday before I started the evening imaging with ver 20, I had time to rerun the earlier bias/flat runs with ver 20.

What a muddle. First run was the same 200 bias. It ran all the way through perfectly.
Next I ran lights, 40 per filter for 7 filters. Was running perfectly for the first 4+ filters and I needed to start imaging so I clicked the pause button. Crashed immediately. =====Only failure of the night =====

The really good news: no crashes where I had lots before. Don’t know what is happening, but I think we can put this issue to bed.

Version 20 ran perfectly for my entire nights imaging session.
Great new code for the display.

OK… Well that’s good. I was very unhappy about the toggle code anyhow. I’ll take a look at your logs… hopefully the crash on pause is completely unrelated to the preview display issue.

Looking at your logs, it is unfortunately related to the same image preview issue. I added some additional exception handling and I have one other safety mechanism to implement (wouldn’t make sense to anyone if I attempted to explain it).