I previously had this error a few weeks ago but had updated my intel stick to the latest anniversary patch. I at the same time installed the latest SGP beta. I thought something went wrong with the windows update so I reinstalled windows and all apps from scratch.
I had 1 session since then and SGP worked fine.
Tonight I got the same crash! There’s no trigger from what I see it just will randomly crash.
I will upload logs in the morning but its a hard crash. So much that it corrupts the sequence file when it happens. I remember this same corrupted problem when it happened the first time.
The error in the event log shows the same WindowsCodec.DLL crash. With a .NET runtime as well.
Buzz, I actually had windows updates off for almost a year but I had decided to update everything last month. This was when I first experienced this error.
It just happens theres a new creators patch out now but I have not upgraded to that so hope not to confuse which update I installed.
I also generally like to use the latest SGP betas
One common factor was trying the newest SGP beta 2.6.0.21 so it could be beta related. I originally thought it was something with windows because of the DLL involved.
I also saved my original log file when I last experienced this last month (prior to re-formatting and re-installing), both seemed to crash at similar instances during the “Performing Serialize” entry.
The randomness is also demonstrated by the error log 2017-03-22. The first one happened relatively quickly and then re-occurs the same night much later, in fact it had crashed while unattended in the second log.
Also included are Windows Event Viewers.
Finally if this helps, I’ve also included the resulting corrupted sequence file (m81.sgf). The messagebox says “Error loading sequence…End of Stream encountered before parsing was completed.” Hard to recover from this error out in the field.
Thank you for your time investigating this! Very much appreciated.
It is not immediately clear what is happening here, but I can confirm it has to do with auto saving the sequence. It’s not ideal, but you can probably prevent this by just turning that off. We will need to determine if this is an aspect of a new Windows update or something specific to your machine. Either is possible… please let us know if you see this on your machine.
FWIW, I have been testing with the Win 10 creators update for months and have never observed this behavior.
It certainly could be, but I cannot think of what. Sometimes changes in one area have unintended side effects. Also, something to consider is that this is a serious issue and we are not getting any widespread reports of this… The fact that the error is intermittent could mean there is an issue with the underlying drive (or network if it is a network drive).
Thanks Ken, I was thinking if perhaps it’s something with the hard disk (for example) then I can run a few stress tests on it to see if it reports anything.
The other Idea I had was to setup a cycle with the simulator exposures every 30 seconds for example and looping just to see if I can reproduce this error.
Do you think that would be a valid experiment with the simulator? or should loop it with EQASCOM and the real camera instead of the simulator.
Yes, Thank you Ken! I set the sequence to 25 seconds to get lots of hits.
One note I did not mention was I am also using windows 32-bit.
The other thing I noticed that, after all those tests, I did a windows update on my Windows 1607 version and it did an accumulative update that also updated the WindowsCodecs.DLL to a new version.
Finally, I may just update the whole thing to the new Creators update and re-test everything, hah.