I have now had the opportunity to try out the Kstars 3.5.0 beta version for my last 3 imaging nights.

I can only say: Chapeau!

Not a major problem, Kstars ran smoothly through the night, perfect refocusing, realignment and guiding after the flip. I got great frames from those 3 nights!

The only minor problem (which predates this version) is that if you were guiding before starting the scheduler, say, to capture a perfectly framed image to use as a template for repeated realignments, and you forgot to stop the guider before starting the scheduler, then the scheduler will hang with the message "Guiding already in progress" or something to that effect.

I wonder if it is possible to just stop guiding whenever the scheduler is started to obviate that problem.

Other than that, I have found noting else so far.

GREAT WORK!

I wish I could do that...

Cheers,

Jo

Glad to hear that! I'll defer to either Wolfgang or Eric regarding the scheduler behavior since they're the experts on this subject matter.

Thanks for the kind words. We should be adding capture and scheduler tests in the coming months, which hopefully should allow us to detect those issues. There are two issues that we need to fix very soon: scheduler insisting on guiding when not required and capture doing nothing when needing to guide after a flip. There is a merge request on invent.kde.org for the second point that needs testing.

-Eric

Compiling it right now. Will let you know what I find.

Jo

With my latest installation from the invent.kde.org repository as of last night I had no problem with capturing or guiding after the flip. Everything performed automatically as commanded by the scheduler and executed perfectly.

I read about the "capture doing nothing when needing to guide after a flip" problem here on the forum, but I have not encountered it.

Also, I want to point out the vast improvements to the guider Hy implemented with the GPG algorithm. That has improved guiding accuracy with my low end iOptron SmartEQPro+ mount enormously!

Jo

Hi Jo,

I would like to also add how well the 3.5.0 Beta is working. I am using 3.5.0 Beta Build: 2020-08-07. The current Astro-Physics Experimental driver in Git is under a major rewrite so I have compiled an earlier version of the driver from INDI 1.8.4 or 1.8.5 which does have a problem computing the Park 3 position (this was working fine in a version from last fall), however I have worked around that by using the "Custom" park position.

I am running it on a Raspberry Pi4, latest Raspberry PI OS and compiled versions of Kstars/INDI as described above. I modified a version of the script from Robert Lancaster for my system. In my use and testing the automated meridian flip, parking (using the Custom parking position), the wonderful new SEP Multi Star and GPG RA Guider enabled, and the the plate solving with Astrometry.net using Sextractor (this is very fast now) works a treat.

I did an automated run with the scheduler a couple of nights back capturing M27 and it all worked great. This included an automated meridian flip, auto focus, guiding, plate solving, parking when finished, etc. The Fits viewer is much more stable now as well and I was able to comfortably use the Fits viewer with no problems throughout the 3 hour session.

To qualify I have used my system very little in recent months mostly due to weather and lately the driver issue which I have worked around as described above. There have been a lot of changes since I last actively used KStars/Ekos and I am really impressed with where it is now. I have always liked the software and find it an even greater pleasure to use in the latest incarnation. I will be backing up my SD card and use this as my working version going forward.

Kudos to the development team! Thank you all!

btw, Markus patch for AP Experimental was merged in a couple of days ago. Is that working for you? It's suppose to fix the issues reported before.

Hi Jasem...
3.5.0 using SEP Multistar and GPG was fantastic last night - RMS very low - I have been testing 3.5.0 due to the ONStep issue and I consider that solved as it is never a problem now. However last night Kstars just disappeared - unfortunately I had no logging turned on except to console (and I was asleep ), But I do know it crashed after the last frame of one filter (HA) and moving to the next (OIII) - the directory for the capture run on that filter had been created, focus was still on the position for HA filter so that leaves me with filter change as the last actions as autofocus had not kicked in - when I inspected the rig this morning, INDI service was still running so I connected and found the FW (ASI EFW) was reporting some kind of error (communication) but i did not mange to get details.....sorry about the lack of logs (now turned on for next run)

I don't think that is a systematic problem. I was also running 3.5.0 (built on 8/7/2020) last night with the ASI-EFW and multiple filter changes throughout the night with no issues, either last night or the week prior. In my hands, the system is performing flawlessly. Maybe a malfunction of your EFW. Is a filter scraping and has the EFW lost synchronization?

My current problem - totally unrelated to Kstars(!) - is that the WiFi connection for unexplicable reasons occasionally crashes, sometimes in the middle of file transfer. The rig continues to run fine, though, images continue to be collected locally, all Kstars functions execute and I find the telescope parked correctly in the morning.

I wonder whether the WiFi on the Pi4 may be giving out over time? I may open a separate topic on that to ask around.

Jo

If you get consistent crashes, please run KStars vis ekos-debugger. This way you can send the logs back with full useful information.

Will do...I plan on running a full but small test tonight so i can see end to end (3 filters of 2 images each + scheduler + park/unpark) - if it is re-creatable that should do it...thx

Hi Jasem,

I will fetch and compile from the bleeding edge GIT and test with Marcus' latest driver and let you know how that goes. I have a couple SD cards I use just for this type of testing. I hope to have a break in the rain and clouds in a few days.