Astrometric Clamshell misbehaving and hanging ACP

[Peter Prendergast]

I formerly had this under the forum for hardware problems NOT related to ACP but I'm satisfied now that this is indeed an ACP problem.

Planewave has gone through their stuff and is satisfied that the park state is being properly reported.

I've used TSX to park and TSX correctly detects the parked state.

When ACP sends a park command the scope parks correctly, but, ACP just sits there waiting indefinitely thereafter even though the scope
was long ago parked.

This failure is ONE HUNDRED PERCENT reproducible which I am sure will make it tons easier to solve.

Their is NO LOG as the program just locks up.

In fact, ACP when told to abort the park command seems to change state, partially, the clock is dead, but, if you close ACP by clicking on the upper right hand
corner ACP will then come back and ask if you want to park the scope. At that point I say no and the program closes out.

Upon relaunch all is well again.

What do you think? Would you like to team viewer in???

Peter

[Bob Denny]

Yes I will log on and set up a test. I will call to set up a time.

[Bob Denny]

OK, we're on it. Conferencing with us all now.

[Peter Prendergast]

Problem SOLVED
So their actually were three issues.
Issue that SOLVED the problem was that the Astrometric "dome" controller reports falsely that the "dome" can be parked. Well, this is an Astrohaven CLAMSHELL not a "dome"
ACP queiries the driver and is falsely told the "dome" can be parked. When the weather safety script/shutdown script is run the "dome" and scope are told to park. Maestro
keeps telling ACP the "dome" is not parked. So, ACP waits and waits and waits. But, and this is issue number two, ACP doesn't tell you that its waiting for the dome it looks like
ACP is waiting for the scope which is why we ran down a rabbit hole.

Once inside preferences, Dome, I unchecked the box that says to park the "dome" all worked fine. We could see all this via PIPE, a great tool for sure.

The other issue had to do with the Sitech driver. The Sitech driver was reading a "atparkmodel", this model was wrong, this model was using the very very first
model I made after the mount/scope was plopped on the pier and the initial model was off by 3 degrees. So, trying to park the scope up ended the math.
Once the correct model was installed, with an error of 30 arc seconds, all went well with the Sitech driver. Up until then, the Sitech driver never found park.
Once the correct model was installed in this arcane location, park happens immediately.

The other issue uncovered is to say the least, RARE as hens teeth and just an accidental find, if you have ACP park the scope, but OUTSIDE ACP you
unpark the scope, ACP chokes because no one told ACP a change of state occurred. This likely effects no one, but was an interesting find.

So, Astrometric will fix that bug, in the mean time, I've unchecked the park dome metric and we're all good. Phew

Let me thank Bob and Kevin Iverson for working together to sort this one out!!!!!!!!!!
As usual superb support!!!!!!!

Peter

[Bob Denny]

Wow!!! Excellent!!! I will file a ticket on the "hens' teeth" one.

The fact that the Astrometric controller accepted the Park() command without protesting (with an error) is the worst bug. That violates the "prime directive" of "do it right or report an error". It didn't do it at all and it didn't report the error.

ACP-1717 - Unparking outside of ACP can cause ACP to hang

[Russell Sandlin]

Peter,

Thanks for figuring this out. I have the same setup and I'm seeing this issue.
I have unchecked the park dome in preferences. I will test it tonight.


Thanks again,


Russell Sandlin

[Peter Prendergast]

Peter,

Thanks for figuring this out. I have the same setup and I'm seeing this issue.
I have unchecked the park dome in preferences. I will test it tonight.


Thanks again,


Russell Sandlin

the Astrometric Firmware was fixed quickly so that should not be a problem
Peter

[Bob Denny]

Thanks Peter!

[Bob Denny]

Wow, this is a combination of the old internal ACP logic to "freeze" the app to accommodate the old (ancient) unbuffered CCD cameras... minimizing CPU to prevent horizontal banding (sordid story that will seemingly never end), and internal park tracking for the old Telescope V1 drivers that had no way to determine whether they were parked or not. Fixed for the next release (Gemini 1717).