Telescope.Sync() errors from TheSky causing FocusMax and ACP issues

[John Batts]

Hi Bob!

I've tried to fix this over the last week, but have been unable to execute FocusMax (v4.1.0.42) from ACP (v8.1). "Manually" executing FocusMax (that is, clicking the FM "Focus" button) results in a focus run, but neither an ACP run or the AutoFocus.vbs script does.

A log is attached ... the scope slews, takes & solves a pointing exposure, then the script terminates without error. I've tried reinstalling ACP (without uninstalling first ... don't worry!), checked both FM & ACP settings, but need to raise the white flag. I'd appreciate any suggestions.

Thank you!
John

[Dean Salman]

Hi John, this is a puzzle. WHen you start ACP is focusMax started and connected already. I am assuming nothing is running in admin mode. I am using build 40. Is this new when you installed build 42

[John Batts]

Hi Dean, thank you for responding ... a puzzle indeed! I forgot to mention that I thought it would be possible to debug this by setting ACP up in "simulator mode" (where ACP, MaxIm and FM use simulators), but focus runs are triggered as expected. So FocusMax runs normally when clicking "Focus," and when ACP is simulating a run, but not during live runs.

Unfortunately, nothing is being run in Admin mode. The latest FM release could be the culprit, but I'd think there'd be other similar error reports. Hoping our Fearless Leader will be able to sort this out.

[Dean Salman]

I assume you uninstalled and reinstall FM, that is safe to do

[John Batts]

I haven't tried uninstalling & reinstalling FM yet, because I seem to recall FM automatically does that when it upgrades (unlike ACP, which just overwrites the previous installation ... safer).

[Bob Denny]

It is a puzzle. So what's different between simulation and when things are actually connected? I understand that it runs fine when ACP and FMx are connected to simulators. The biggest puzzle is why you don't see anything when FMx fails to start.

In the 20161026_190553 log I see

01:05:54 (assuring that FocusMax is running now...)


It is at this point, at the beginning of every run, that ACP creates an instance of FocusMax and leaves it attached throughout the run. Are you saying that you never see FocusMax appear? And only when FocusMax and ACP are connected to real devices, not the simulators?

Also in that log I see an abrupt ending with no errors there

01:11:02 Imager sky position angle is 359.5 deg.
ACP console log closed 27-Oct-2016 01:11:03 UTC

and at this point you don't see any sort of error box, Bill Gates failure box, or anything appearing? Obviously no errors in the log. Since the "prime directive" is "do it right or raise an error" I'm stumped at the moment. Maybe you can fill in a puzzle piece from the above questions?

[John Batts]

I agree! In the uploaded log, I manually started FocusMax before beginning the run (that is, I can see it & I'm sure it's running), so maybe that's why there's no error message after "assuring that FM is running now ..."?

To clarify, (a) FocusMax is always running and connected, (b) a focus run occurs only when either the FM "Focus" button is clicked or when FM & ACP are connected to simulators (not real devices), and (c) there are no error messages (Win, ACP, FM or MaxIm).

Since there are no error messages, is it possible that FocusMax isn't called when connected to real devices? COM failure?

[Bob Denny]

I misunderstood you. I didn't realize that you had pre-started FocusMax and that it just wasn't focusing. I thought it wasn't starting. It is definitely not a COM failure. It will be a device problem or a weird bug in something that is triggered by as yet never seen condition.

It's bizarre enough that I'd like to see it with my own eyes. I'm going to call and try to set up a time.

[Bob Denny]

An hour and a half later, we have isolated it to what appears to be a malformed COM error response from TheSky when it is called to Sync(). This is an unusual problem because most people run TPOINT, and the driver is set to "inhibit sync to protect TPOINT model", and thus ACP never calls TheSky to Sync(), instead ACP jogs it to center the object. In this case, the COM error is not being trapped by ACP's Windows Script error handling... instead somehow it is causing the script to silently stop. That's why I suspect the error response is somehow strangely formed. Anyway we were able to trap it in a simple Javascript and it is "Object not set to a reference to an object".

John's going to continue to pursue this further as I have another online session and it's out of my hands for sure anyway.

[Colin Haig]

John - it would be good to know which Daily Build of TSX you are running, as a few of us use it as well.