I have considered an API interface, but then a API interface program has then to be started by autostart and stay resident in memory. The ASTAP program is also compiled for Mac, Linux AMD64, Raspberry PI3 and 4. It has to work in every environment. There are some hurdles to take for an API interface.
What’s wrong with a command-line interface? The program is given a search limit so it should stop when the search is finished. Secondly, you should be able stop it anytime by a killing the application.
In principle ANSVR is an API interface for the command-line driven Astrometry.net. Astrometry.net is a real collection of command-line driven programs written in C and Python 2.
Nevertheless, I have considered and seperate API interface program to emulate Astrometry.net. Maybe that will be realised in the future
I will keep an eye on this forum to see if there are any startup problems.
Yes, I have just switched over to ASTAP also after running a few tests. Wow, it’s fast! The true test will be under the stars, but I have 100% confidence in very fast solves now. Thanks Han!
It is not designed for inter-process communication. It can to be fragile and tends leaves the calling application in a “blind” state where the only feedback we can give after calling is solved or not solved… no specific issues or reasons the solve failed. I saw that you are cross platform which is why I asked specifically about a REST API as opposed to a .NET assembly or COM interface.
Maybe in the interim, to help bridge this gap, ASTAP could ALWAYS write a text file after it is invoked. In cases of a valid solve, it will remain the same format and in cases where it fails, give as specific a reason as possible (e.g. Catalog not installed, unsupported image type, etc.)
Yes writing some info is not a problem. Besides the .apm (PlateSolve2 style) with a solved or not solved report, ASTAP writes a .txt file with in the first line the used command line and in the second the solver results if i remember well. This is for debugging, but I can extend that with some other info like you suggested.
Using the native ASTAP commandline it produces a .wcs file like Astrometry.net and an .ini file with the same keyword based info. The contains is identical and maybe the .ini is easier to read. A missing star database can be reported in these files.
A missing star database is already reported by a message window.
If the search spiral has reached a distance more then 2 degrees from the start position, then a popup notifier will show the actual search distance and solver settings. There is also a tray icon with hint info, but tray icons are default off in the latest Win10 versions. so most users will not see the tray icon.
Sure, that seems great to me. I assume you mean that you would publish a list of custom error levels and we could integrate that into our own messaging within SGPro?
Wow! That’s a fast plate solver. PS2 has been very reliable for me with 100% successful solves in 2-3 seconds. This solved in less than a second and no waiting for the window to close so SGP can move on. I’ll let it run for a few nights to see how it goes but initial tests are great!
Apologies if this is starting to sound boring but, Wow! And Wow! Just completed a comparison and found PS2 takes around 14 seconds and ASTAP only 3 seconds. Tested on 14 different raw images of varying quality (Aeroplane streaks, out of focus, in focus, etc.) and ASTAP never missed a beat.
However…I did have an issue getting SGPro to see the executable (version 3.1.0.359). What I found was when using SGPro to locate the astap.exe file, SGPro would omit the ending “\”.
So I needed to manually add this in by editing the “sgpro_settings.json” file in notepad, adding a “\\” to the end of the file so it looks like: ,"_settingAstapDirectory":“C:\\Program Files\\astap\\”}
Now I don’t know if this is specific to something I’ve done, but I just followed the simple install instructions.
Since there are some errors regarding Mercury is not the right answe of the forum of HNSKY, I just want to add one bug of ASTAP here. If you opened the ASTAP on the secondary screen on Windows, then disconnect the 2nd screen, the ASTAP won’t show again unless you recoonect it and move to your main screen.
Just some more info – I have been using ASTAP in a test application and have done about 100 plate solves. None have failed and the average plate solve time is < 3 seconds. The RA and DEC values in the FITS header are reasonably accurate, which speeds things up a lot.
