|
Description
Andre Kovacs
2019-03-13 16:08:50 UTC
Jasem, Just to let you know, I managed to workaround this issue by overwriting the contents of the directory KStars.app/Contents/MacOS/astrometry with the ones from version 3.0.0 of KStars. Thanks, Andre Can you run this from the command line on the latest version from the DMG to see what fails? This won't give us enough information to diagnose the issue on your system. To find out what is wrong, you will need to run a test on your astrometry setup. 1. Plate solve something, it can be a simulator image, it doesn't matter what. 2. Then did you see in the log where it printed the commands that it tried? Try running those from a terminal window. From your Log these are the commands you should run. (You could copy and paste them to the terminal): export PATH=/Applications/kstars.app/Contents/MacOS/netpbm/bin:/Applications/kstars.app/Contents/MacOS/python/bin:/usr/local/bin:/usr/local/sbin:/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin:/opt/X11/bin export PYTHONPATH=/Applications/kstars.app/Contents/MacOS/python/bin/site-packages /Applications/kstars.app/Contents/MacOS/astrometry/bin/solve-field -O --no-plots --no-verify --resort --downsample 2 -3 351.405 -4 61.698 -5 30 --config /Applications/kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg -W /private/var/folders/l3/yznjwk917w9_f9qglkhkndpr0000gn/T/solution.wcs /private/var/folders/l3/yznjwk917w9_f9qglkhkndpr0000gn/T/fitstVCXWV.fits Just one note: the last part of the command where it says "/private/var/folders/l3/yznjwk917w9_f9qglkhkndpr0000gn/T/fitstVCXWV.fits". That is a temporary file, so that is what you would need to change. (also probably the wcs file right before that) Created attachment 118790 [details]
terminal window
The terminal window during the command
Created attachment 118791 [details]
command location
where to get the command for the terminal
Created attachment 118792 [details]
successful terminal run
what should happen in the terminal if everything works correctly.
Here's the verbose output of the failed execution (original version of astrometry.net solver): "2019-03-14T00:17:24 Solver failed. Try again. 2019-03-14T00:17:24 Error starting solver: Process crashed 2019-03-14T00:17:24 Starting solver... 2019-03-14T00:17:24 Index files index-4202 to index-4204 are missing. Astrometry.net would not be able to adequately solve plates until you install the missing index files. Download the index files from http://www.astrometry.net 2019-03-14T00:17:24 Image received. 2019-03-14T00:17:10 Capturing image... 2019-03-14T00:17:09 Changing filter to Green... 2019-03-14T00:15:56 World Coordinate System (WCS) is enabled. CCD rotation must be set either manually in the CCD driver or by solving an image before proceeding to capture any further images, otherwise the WCS information may be invalid. 2019-03-14T00:15:56 Detected Astrometry.net version 0.76 2019-03-14T00:15:56 Idle." Unfortunately, there's not much in it besides the strange error that the optional index files are missing. How do I enable the log level required to be able to see the commands that Ekos used to call the astrometry.net process? I couldn't get any information on that by enabling the Verbosity as Verbose, under KStars -> Preferences -> Advanced -> Logging Output. Here's the verbose output of the success execution (older version of astrometry.net solver): "2019-03-14T00:37:10 Solution coordinates: RA (01h 09m 14s) DEC (-55° 08' 05") Telescope Coordinates: RA (01h 09m 12s) DEC (-55° 08' 21") 2019-03-14T00:37:10 WCS information updated. Images captured from this point forward shall have valid WCS. 2019-03-14T00:37:10 Solver completed in 8 seconds. 2019-03-14T00:37:01 Starting solver... 2019-03-14T00:37:01 Index files index-4202 to index-4204 are missing. Astrometry.net would not be able to adequately solve plates until you install the missing index files. Download the index files from http://www.astrometry.net 2019-03-14T00:37:01 Image received. 2019-03-14T00:36:47 Capturing image... 2019-03-14T00:35:55 World Coordinate System (WCS) is enabled. CCD rotation must be set either manually in the CCD driver or by solving an image before proceeding to capture any further images, otherwise the WCS information may be invalid. 2019-03-14T00:35:54 Detected Astrometry.net version 0.73 2019-03-14T00:35:54 Idle." I managed to use your commands to successfully call the plate solver via the command line, and using the Options displayed on Ekos. Here're my commands: export PATH=/Applications/Astronomia/kstars.app/Contents/MacOS/netpbm/bin:/Applications/Astronomia/kstars.app/Contents/MacOS/python/bin:/usr/local/bin:/usr/local/sbin:/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin:/opt/X11/bin export PYTHONPATH=/Applications/Astronomia/kstars.app/Contents/MacOS/python/bin/site-packages /Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/solve-field -O --no-plots --no-verify --resort --downsample 2 -L 32.5717 -H 53.9383 -u aw -3 17.3038 -4 -55.1392 -5 30 --config /Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg -W ./solution.wcs ./test.fits Here's the console output of the successful execution via the command line (older version of astrometry.net solver): "/Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/solve-field -O --no-plots --no-verify --resort --downsample 2 -L 32.5717 -H 53.9383 -u aw -3 17.3038 -4 -55.1392 -5 30 --config /Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg -W ./solution.wcs ./test.fits Reading input file 1 of 1: "./test.fits"... Extracting sources... Downsampling by 2... simplexy: found 1474 sources. /Applications/Astronomia/kstars.app/Contents/MacOS/python/bin/site-packages/pyfits/__init__.py:22: PyFITSDeprecationWarning: PyFITS is deprecated, please use astropy.io.fits PyFITSDeprecationWarning) # noqa Reading sort column "FLUX" Sorting sort column mmapping input file Copying table header. Writing row 0 Done /Applications/Astronomia/kstars.app/Contents/MacOS/python/bin/site-packages/pyfits/__init__.py:22: PyFITSDeprecationWarning: PyFITS is deprecated, please use astropy.io.fits PyFITSDeprecationWarning) # noqa Solving... Reading file "./test.axy"... Only searching for solutions within 30 degrees of RA,Dec (17.3038,-55.1392) Field 1 did not solve (index index-4210.fits, field objects 1-10). Field 1 did not solve (index index-4209.fits, field objects 1-10). Field 1 did not solve (index index-4208.fits, field objects 1-10). Field 1 did not solve (index index-4207-11.fits, field objects 1-10). Field 1 did not solve (index index-4207-08.fits, field objects 1-10). Field 1 did not solve (index index-4207-04.fits, field objects 1-10). Field 1 did not solve (index index-4206-11.fits, field objects 1-10). log-odds ratio 295.194 (1.58973e+128), 53 match, 0 conflict, 44 distractors, 77 index. RA,Dec = (17.1018,-55.2418), pixel scale 0.590997 arcsec/pix. Hit/miss: Hit/miss: ++++++++-++-++++++++++-+++-++++++-++--+++++++-++-+-+--++---+-+--+++++---------+-----------------+(best)--- Field 1: solved with index index-4206-08.fits. Field 1 solved: writing to file ./test.solved to indicate this. Field: ./test.fits Field center: (RA,Dec) = (17.103504, -55.241262) deg. Field center: (RA H:M:S, Dec D:M:S) = (01:08:24.841, -55:14:28.542). Field size: 51.334 x 34.2598 arcminutes Field rotation angle: up is 89.8983 degrees E of N Field parity: neg Creating new FITS file "./test.new"... Here's the console output of the failed execution via the command line (original version of astrometry.net solver): "/Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/solve-field -O --no-plots --no-verify --resort --downsample 2 -L 32.5717 -H 53.9383 -u aw -3 17.3038 -4 -55.1392 -5 30 --config /Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg -W ./solution.wcs ./test.fits Reading input file 1 of 1: "./test.fits"... Illegal instruction: 4" Illegal Instruction 4?? Interesting. That is the error I have seen before when I try to run software that was built to run for a newer version of Mac OS X on an older OS X. But you said you are running High Sierra. . . Let me look into this one. Yes, I just updated from El Capitan to High Sierra to be able to run KStars 3.1.1. It definitely should not be able to run on 10.11 because QT now refuses to support that version of the OS. But I did set the deployment targets to 10.12 and I did test it out on a 10.12 system awhile back. I thought it was all working. I can try to run a few tests on it tomorrow. I haven't tried on 10.12 (Sierra), because I migrated directly to 10.13 (High Sierra). Maybe it's something specific to Astrometry.net 0.76 on 10.13 (High Sierra), and version 0.75, or 0.74, could work better. I noticed that KStars' internal version of Astrometry.net is a stripped down build of the nova.astrometry.net. From which repository did you get this version? If you want to just get it working now for yourself, you can run my build script and it should be guaranteed to run on your computer because then yours built it. I am using astrometry 0.76 in my script. You might want to give this a go: https://github.com/rlancaste/kstars-on-osx-craft Maybe it can be related, I use macports to install my libraries (mostly using universal i386+x86_64 versions), but I had to install libusb via home-brew to be able to startup the INDI webcam driver via INDI Control Panel (startup via the command console was working fine). If you're still following the steps from jamiesmith scripts, the Astrometry.net is copied into the kstars.app/Contents/MacOS/astrometry via the following steps (https://github.com/jamiesmith/kstars-on-osx/blob/master/README.md): mkdir -p ${KSTARS_DIR}/Applications/KDE/kstars.app/Contents/MacOS/astrometry cp -r $(brew --prefix astrometry-net)/bin ${KSTARS_DIR}/Applications/KDE/kstars.app/Contents/MacOS/astrometry/ cp -r $(brew --prefix astrometry-net)/lib ${KSTARS_DIR}/Applications/KDE/kstars.app/Contents/MacOS/astrometry/ cp $(brew --prefix astrometry-net)/etc/astrometry.cfg ${KSTARS_DIR}/Applications/KDE/kstars.app/Contents/MacOS/astrometry/bin/ So, it copies some of the folders from the Astrometry.net installed via Home-brew (which I don't have installed). Maybe something was left missing to copy for version 0.76, that could be referencing externally from the .app in your machine. To test this hypothesis, I forced the installation of the latest version of astrometry-net (0.76_2) via home-brew, without it's dependencies, and copied all its contents, and also the astrometry.cfg, into a new astrometry directory under kstars.app/Contents/MacOS/ Now the error changed to code 6: "/Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/solve-field -O --no-plots --no-verify --resort --downsample 2 -L 32.5717 -H 53.9383 -u aw -3 17.3038 -4 -55.1392 -5 30 --config /Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg -W ./solution.wcs ./test.fits dyld: Library not loaded: /usr/local/opt/cfitsio/lib/libcfitsio.7.dylib Referenced from: /Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/solve-field Reason: image not found Abort trap: 6" PS: The version of cfitsio installed in my machine, using macports, is 3.450_0+universal, apparently the same bundled with KStars 3.1.1 (libcfitsio.3.dylib) into kstars.app/Contents/MacOS/Frameworks. Also, I tried to install cfitsio using homebrew, without its dependencies, but it also installed version 3.450_1. so as I mentioned, https://github.com/rlancaste/kstars-on-osx-craft is now how it is built. Home-brew is used, but barely, just for python3 and a couple of these we couldn't build with craft yet. Most of it is built using craft including astrometry. And you should note that everything gets wrapped up inside the app bundle and the script runs over everything inside and bundles all of the dependencies up too. I forgot to check the actual name of the dylib inside /usr/local/Cellar/cfitsio/3.450_1/lib and it's really libcfitsio.7.3.45.dylib, with a link named libcfitsio.7.dylib.
I copied this link named libcfitsio.7.dylib into the kstars.app/Contents/MacOS/Frameworks, together with wcslib 6.2 (libwcs.6.2.dylib), and had to install also python3 and bumpy via homebrew.
Now, the none of the error codes 4 or 6 occurred, but I got a Python error with return code 1:
"/Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/solve-field -O --no-plots --no-verify --resort --downsample 2 -L 32.5717 -H 53.9383 -u aw -3 17.3038 -4 -55.1392 -5 30 --config /Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg -W ./solution.wcs ./test.fits
Reading input file 1 of 1: "./test.fits"...
Extracting sources...
Downsampling by 2...
simplexy: found 1474 sources.
Traceback (most recent call last):
File "/Applications/Astronomia/kstars.app/Contents/MacOS/python/bin/site-packages/numpy/core/__init__.py", line 16, in <module>
from . import multiarray
File "/Applications/Astronomia/kstars.app/Contents/MacOS/python/bin/site-packages/numpy/core/multiarray.py", line 12, in <module>
from . import overrides
File "/Applications/Astronomia/kstars.app/Contents/MacOS/python/bin/site-packages/numpy/core/overrides.py", line 6, in <module>
from numpy.core._multiarray_umath import (
ImportError: dlopen(/Applications/Astronomia/kstars.app/Contents/MacOS/python/bin/site-packages/numpy/core/_multiarray_umath.so, 2): Symbol not found: _PyBuffer_Type
Referenced from: /Applications/Astronomia/kstars.app/Contents/MacOS/python/bin/site-packages/numpy/core/_multiarray_umath.so
Expected in: flat namespace
in /Applications/Astronomia/kstars.app/Contents/MacOS/python/bin/site-packages/numpy/core/_multiarray_umath.so
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/removelines", line 14, in <module>
from astrometry.util.removelines import main
File "/usr/local/Cellar/astrometry-net/0.76_2/libexec/lib/python3.7/site-packages/astrometry/util/removelines.py", line 10, in <module>
import numpy as np
File "/Applications/Astronomia/kstars.app/Contents/MacOS/python/bin/site-packages/numpy/__init__.py", line 142, in <module>
from . import core
File "/Applications/Astronomia/kstars.app/Contents/MacOS/python/bin/site-packages/numpy/core/__init__.py", line 47, in <module>
raise ImportError(msg)
ImportError:
IMPORTANT: PLEASE READ THIS FOR ADVICE ON HOW TO SOLVE THIS ISSUE!
Importing the multiarray numpy extension module failed. Most
likely you are trying to import a failed build of numpy.
Here is how to proceed:
- If you're working with a numpy git repository, try `git clean -xdf`
(removes all files not under version control) and rebuild numpy.
- If you are simply trying to use the numpy version that you have installed:
your installation is broken - please reinstall numpy.
- If you have already reinstalled and that did not fix the problem, then:
1. Check that you are using the Python you expect (you're using /usr/local/Cellar/astrometry-net/0.76_2/libexec/bin/python3),
and that you have no directories in your PATH or PYTHONPATH that can
interfere with the Python and numpy versions you're trying to use.
2. If (1) looks fine, you can open a new issue at
https://github.com/numpy/numpy/issues. Please include details on:
- how you installed Python
- how you installed numpy
- your operating system
- whether or not you have multiple versions of Python installed
- if you built from source, your compiler versions and ideally a build log
Note: this error has many possible causes, so please don't comment on
an existing issue about this - open a new one instead.
Original error was: dlopen(/Applications/Astronomia/kstars.app/Contents/MacOS/python/bin/site-packages/numpy/core/_multiarray_umath.so, 2): Symbol not found: _PyBuffer_Type
Referenced from: /Applications/Astronomia/kstars.app/Contents/MacOS/python/bin/site-packages/numpy/core/_multiarray_umath.so
Expected in: flat namespace
in /Applications/Astronomia/kstars.app/Contents/MacOS/python/bin/site-packages/numpy/core/_multiarray_umath.so
augment-xylist.c:589:backtick Failed to run command: /Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/removelines /tmp/tmp.xyls.caV4ak /tmp/tmp.removelines.CLPZ8N
ioutils.c:567:run_command_get_outputs Command failed: return value 1"
Checking the Python version bundled with KStars 3.1.1 I noticed that it's 2.7 and not 3, as Astrometry.net requires.
So I created a new directory with Python 3, copied numpy and pyrites site-packages under it.
And the error changed to:
"Library not loaded: /usr/local/opt/openblas/lib/libopenblasp-r0.3.5.dylib"
I installed openblas 0.3.5 via homebrew, and copied it under kstars.app/Contents/MacOS/Frameworks (none version was present, like libwcs.6.2.dylib).
Aftrewards I had also to install gcc 8.3.0 via homebrew, and copied libgfortran.5.dylib and libquadmath.0.dylib under kstars.app/Contents/MacOS/Frameworks (none version was also present).
But then I stuck with this error:
"/Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/solve-field -O --no-plots --no-verify --resort --downsample 2 -L 32.5717 -H 53.9383 -u aw -3 17.3038 -4 -55.1392 -5 30 --config /Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg -W ./solution.wcs ./test.fits
Reading input file 1 of 1: "./test.fits"...
Extracting sources...
Downsampling by 2...
simplexy: found 1474 sources.
Traceback (most recent call last):
File "/usr/local/opt/numpy/lib/python3.7/site-packages/numpy/core/__init__.py", line 40, in <module>
from . import multiarray
File "/usr/local/opt/numpy/lib/python3.7/site-packages/numpy/core/multiarray.py", line 12, in <module>
from . import overrides
File "/usr/local/opt/numpy/lib/python3.7/site-packages/numpy/core/overrides.py", line 6, in <module>
from numpy.core._multiarray_umath import (
ImportError: dlopen(/usr/local/opt/numpy/lib/python3.7/site-packages/numpy/core/_multiarray_umath.cpython-37m-darwin.so, 2): Symbol not found: ___addtf3
Referenced from: /usr/local/opt/gcc/lib/gcc/8/libquadmath.0.dylib
Expected in: /usr/lib/libSystem.B.dylib
in /usr/local/opt/gcc/lib/gcc/8/libquadmath.0.dylib
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/removelines", line 14, in <module>
from astrometry.util.removelines import main
File "/usr/local/Cellar/astrometry-net/0.76_2/libexec/lib/python3.7/site-packages/astrometry/util/removelines.py", line 10, in <module>
import numpy as np
File "/usr/local/opt/numpy/lib/python3.7/site-packages/numpy/__init__.py", line 142, in <module>
from . import core
File "/usr/local/opt/numpy/lib/python3.7/site-packages/numpy/core/__init__.py", line 71, in <module>
raise ImportError(msg)
ImportError:
IMPORTANT: PLEASE READ THIS FOR ADVICE ON HOW TO SOLVE THIS ISSUE!
Importing the multiarray numpy extension module failed. Most
likely you are trying to import a failed build of numpy.
Here is how to proceed:
- If you're working with a numpy git repository, try `git clean -xdf`
(removes all files not under version control) and rebuild numpy.
- If you are simply trying to use the numpy version that you have installed:
your installation is broken - please reinstall numpy.
- If you have already reinstalled and that did not fix the problem, then:
1. Check that you are using the Python you expect (you're using /usr/local/Cellar/astrometry-net/0.76_2/libexec/bin/python3),
and that you have no directories in your PATH or PYTHONPATH that can
interfere with the Python and numpy versions you're trying to use.
2. If (1) looks fine, you can open a new issue at
https://github.com/numpy/numpy/issues. Please include details on:
- how you installed Python
- how you installed numpy
- your operating system
- whether or not you have multiple versions of Python installed
- if you built from source, your compiler versions and ideally a build log
Note: this error has many possible causes, so please don't comment on
an existing issue about this - open a new one instead.
Original error was: dlopen(/usr/local/opt/numpy/lib/python3.7/site-packages/numpy/core/_multiarray_umath.cpython-37m-darwin.so, 2): Symbol not found: ___addtf3
Referenced from: /usr/local/opt/gcc/lib/gcc/8/libquadmath.0.dylib
Expected in: /usr/lib/libSystem.B.dylib
in /usr/local/opt/gcc/lib/gcc/8/libquadmath.0.dylib
augment-xylist.c:589:backtick Failed to run command: /Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/removelines /tmp/tmp.xyls.X34rQN /tmp/tmp.removelines.BMkQIi
ioutils.c:567:run_command_get_outputs Command failed: return value 1"
Resuming, it seems that many of the dependencies of Astrometry.net are externally referenced by KStars 3.1.1, since I had to include many of them into the app bundle.
so I have a fresh High Sierra virtual machine that I could test on that has never had home-brew, macports, kstars, or anything else installed on it. This evening I opened the DMG and installed kstars, then I downloaded GSC, I made sure the directory was right for the index files, then I downloaded the 4205 index files, I started an indi server, and I plate solved with no problem and no errors. The only issue I had was the astrometry folder in Application Support was not getting created and I had to do it manually, I will have to look into that. But none of the errors you reported with the illegal instructions etc. I suspect that some linux programs you had already installed in /usr/local or somewhere else on the system using homebrew or macports was in the system path and was being called on instead of the stuff in the kstars app bundle. And this was probably causing some linking problems because these programs are linked to other dylibs on your system. please note that issues with different versions of python, netpbm, numpy, pyfits, and astrometry installation issues were the reason I finally found that I had to just install a stripped down version of those programs all inside the app bundle so that kstars could just use that instead of trying to use the system files that might not be configured right. Doing this solved problems for so many people. We are using python 2.7 for astrometry which is still permitted by astrometry.net. There are two reasons for this. The first reason is to avoid conflicts with the users python which is usually python3. The second is that craft had a recipe already set up for building python2. it doesn't have one for building python3 because craft runs entirely in python3 for building software. using 2.7 causes less problems. All those errors you are observing is totally the reason I installed everything kstars should need into the app bundle pre configured so it uses that one instead. In trying to use python3 and stuff on your own system you are going down a rabbit hole I think. I really do suspect that part of your problem is that you had some things already located in /usr/local/ and maybe some of it wasn't compatible with the stuff in the kstars bundle. Possibly a solution to this would be for me to change the PATH variables kstars gives to astronomy to not include /usr/local Note that I said above that on a totally fresh installation of High Sierra using the current kstars dmg 3.1.1 that I posted on the website, there were no issues with plate solving using the internal astrometry, netpbm, python, pyfits, and numpy. When we first were working with kstars, we did need /usr/local to be in the path because that was where we were telling folks to install the astrometry stuff. Now since it is all in the bundle, we don't really need to do so. astrometry probably found python in your /usr/local on your system or something. I agree with you that something on my /usr/local or /opt/local is causing a conflict with the bundled versions, but you should have similar issues if you configure the development environment for indi and KStars locally. Also, last year I tried to build nova.astrometry.net from the repository, to avoid homebrew. Tomorrow I’ll check the list of libraries installed in my /usr/local and /opt/local via macports. Another difference I have in this machine is that I grouped all astronomy apps under a subdirectory of ~/Applications. As long as the subdirectory of applications does not have a space in the file path, it should be just fine. I made absolutely sure with everything I have done with kstars that it follows the idea of bundling. I did this because I tend to have multiple versions of kstars on my computer simultaneously and they need to work independently. Also because I too like to sort programs. I think I will run some tests maybe this weekend to see if I can remove the /usr/local directory from the path kstars passes to its programs. It shouldn't need that at all anymore. I was checking all Python references inside the kstars.app/Contents/MacOS/astrometry and I found that your bundle has a python/astrometry directory under kstars.app/Contents/MacOS/astrometry/lib that is not present in the /usr/local/Cellar/astrometry-net/0.76_2/lib that I copied into my bundle. It seems to me that this python/astrometry directory should contain the /usr/local/Cellar/astrometry-net/0.76_2/libexec/lib/python3.7/site-packages/astrometry/util/removelines.py that is causing the error in my execution using the version 0.76_2 from homebrew. So I copied also the contents of /usr/local/Cellar/astrometry-net/0.76_2/libexec/lib into kstars.app/Contents/MacOS/astrometry/lib. I also returned the kstars.app/Contents/MacOS/python to your bundled version. But the manual execution from the command line is still referencing the Python3 from /usr/local/Cellar/astrometry-net/0.76_2/libexec/bin/python3 and /usr/local/Cellar/astrometry-net/0.76_2/libexec/lib/python3.7/site-packages/astrometry/util/removelines.py (outside of the scope of the app), besides having a local copy of Python2.7 and the astrometry site-package into the app bundle. But I don't think the solution you suggested will be sufficient (remove the /usr/local from subprocess calls, to avoid any reference to any external homebrew library/package at /usr/local/Cellar), because, using the astrometry.net version from homebrew into the app bundle, I get this error if I try to plate solve after uninstalling home-brew's astrometry.net: "/Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/solve-field -O --no-plots --no-verify --resort --downsample 2 -L 32.5717 -H 53.9383 -u aw -3 17.3038 -4 -55.1392 -5 30 --config /Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg -W ./solution.wcs ./test.fits Reading input file 1 of 1: "./test.fits"... /bin/sh: /Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/image2pnm: /usr/local/Cellar/astrometry-net/0.76_2/libexec/bin/python3: bad interpreter: No such file or directory augment-xylist.c:589:backtick Failed to run command: /Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/image2pnm --infile ./test.fits --uncompressed-outfile /tmp/tmp.uncompressed.tCTq5i --outfile /tmp/tmp.ppm.QciqW8 --ppm --mydir /Applications/Astronomia/kstars.app/Contents/MacOS/astrometry/bin/solve-field ioutils.c:567:run_command_get_outputs Command failed: return value 126" Clearly the astrometry.net installed using homebrew has many static references to the contents of /usr/local/Cellar/astrometry-net/0.76_2/. Just to let you know, I tested KStars versions and the plate solver worked fine up to version 3.0.0. The issue started with version 3.1.0. The question is, what changed in plate solving from version 3.0.0 to 3.1.0? Actually the question is not what changed, the question is what didn't change. Basically EVERYTHING changed from 3.0.0 to 3.1.0. We were forced to abandon the old method we were using to build KStars entirely. The script we built over about 2 years no longer worked for us because qt stopped supporting OS 10.11 entirely. The old script would build using a combination of craft and homebrew. It was fairly complicated, sort of like a patchwork quilt, but it worked pretty well. With the old script, since we were using homebrew, we had to build kstars on the minimum system we wanted to support. Homebrew does not support building software for older systems. So if we wanted to support 10.11, 10.12, 10.13, and 10.14, we had to build on a 10.11 system. If we built on a 10.14 system, it will ONLY work on 10.14. I had been planning for months to get some craft recipes together so that we could build kstars using only craft and not having to install anything using macports or homebrew to /usr/local at all. Then we can build for older OSes I had actually written most of the recipes already over the last few months. But then our old script broke VERY suddenly, with no warning, with a QT update. So I very quickly wrote up a new build script based entirely on craft (well almost entirely). 3.1.0 was the first one we released with the new script. It actually worked almost completely. There were just a couple of mistakes in 3.1.0 that I corrected in 3.1.1. In fact 3.1.0 did NOT plate solve correctly because I misnamed the internal python executable. It needed to be called python2.7. I found that naming that python made it look in /usr/local for python instead of using the bundled one. But naming it python2.7 made it use the correct python. I just upgraded my other Mac (non-development machine) also from El Capitan to High Sierra, and KStars from 2.8.7 to 3.1.1 (and nothing else), and the solver also failed, but in this machine the verbosity of the console is working differently, as follows: "2019-03-16T16:07:41 Solver failed. Try again. 2019-03-16T16:07:41 Error starting solver: Process crashed 2019-03-16T16:07:41 Starting solver... 2019-03-16T16:07:41 Image received. 2019-03-16T16:07:29 Capturing image... 2019-03-16T16:07:29 Changing filter to LPR... 2019-03-16T16:07:18 World Coordinate System (WCS) is enabled. CCD rotation must be set either manually in the CCD driver or by solving an image before proceeding to capture any further images, otherwise the WCS information may be invalid. 2019-03-16T16:07:13 World Coordinate System (WCS) is enabled. CCD rotation must be set either manually in the CCD driver or by solving an image before proceeding to capture any further images, otherwise the WCS information may be invalid. 2019-03-16T16:07:11 Detected Astrometry.net version 0.76 2019-03-16T16:07:11 Idle." PS: Another issue I forgot to mention was that Mac OS is complaining about the developer identity of the app. I had to use the trick of opening the app by right-click+open in Finder to be able to use it. Here're the commands I used to test it via console: export PATH=/Applications/kstars.app/Contents/MacOS/netpbm/bin:/Applications/kstars.app/Contents/MacOS/python/bin:/usr/local/bin:/usr/local/sbin:/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin:/opt/X11/bin export PYTHONPATH=/Applications/kstars.app/Contents/MacOS/python/bin/site-packages /Applications/kstars.app/Contents/MacOS/astrometry/bin/solve-field -O --no-plots --no-verify --resort --downsample 2 -L 194.13 -H 321.847 -u aw -3 101.498 -4 -16.7448 -5 30 --config /Applications/kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg -W ~/Pictures/solution.wcs ~/Pictures/test.fits And here's the output of the execution: /Applications/kstars.app/Contents/MacOS/astrometry/bin/solve-field -O --no-plots --no-verify --resort --downsample 2 -L 194.13 -H 321.847 -u aw -3 101.498 -4 -16.7448 -5 30 --config /Applications/kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg -W ~/Pictures/solution.wcs ~/Pictures/test.fits Reading input file 1 of 1: "/Users/Andre/Pictures/test.fits"... Illegal instruction: 4 When you built your High Sierra machine you installed from scratch or upgraded from El Capitan? Actually I downloaded the High Sierra installer from apple and I installed it in my parallels virtual machine. It is totally fresh. "Another issue I forgot to mention was that Mac OS is complaining about the developer identity of the app. I had to use the trick of opening the app by right-click+open in Finder to be able to use it." This will always be true unless somebody wants to pay a yearly premium for a developer license and then code sign it. I do have a Sierra virtual machine. I could build kstars in that and it would definitely be built on an older computer than yours so then the illegal instruction error would most likely go away if the Mac OS X version is the issue. The goal is to be able to build kstars on any computer Sierra or newer and then have it work on any computer Sierra or newer. But maybe its not there yet fully? Since I'm a new High Sierra user, I don't know what surprises Apple left to us in this version. El Capitan was really traumatic to me, so I postponed the upgrade afterwards as long as I could. My guess is that we should try to get Astrometry.net to build and work on its own, because the KStars astrometry module works just fine with the older version of the Astrometry.net binaries. But never managed to get a local version of Astrometry.net to work. But that is what I did. The whole point is that I got python, netpbm, numpy, pyfits, and astrometry.net all wrapped up inside the app. None of them are built with homebrew, it is all built independent of that. And they do not require you to install any of that because when we first started with kstars on Mac, that was a real nightmare to get them all installed and working. It would be a good idea to figure out what about your system is different, but it indeed appears to work on most Sierra and High Sierra machines. I will try to run some tests this weekend to see if I can figure it out. The only two things I could figure out that could be different in my machines are macports (instead of homebrew), Python environment, and upgrade (instead of fresh install). Tomorrow I'll try to remove all macport installed libraries that were required to build the universal (i386+x86_64) version of the INDI Webcam driver with support for QT Kit. But, unfortunately, I cannot change any Python setup, for professional reasons. I uninstalled all ports, and also don't have any brew library installed, but the error (Illegal instruction: 4) continues. I never tried to build or install Astrometry.net in this machine. The only thing, besides KStars, containing Astrometry.net is CloudMakers' Astrometry app. Maybe I'm missing an update of some library or tool? Yes that is strange. Sorry I didn't get a chance to do my experiments yet. I got bogged down in another issue. Apparently craft didn't build dbus with the features we need in kstars. so I was working on a new recipe for dbus You had to install something in your (High) Sierra machine, other than KStars? Nope. The whole goal is for it to work out of the box. That is why astrometry, netpbm, python, pyfits, numpy, INDI, and xplanet are inside the app bundle. That being said, if you want to use simulators, you have to download GSC (an option in the startup wizard) and if you want to plate solve, you need index files. Regarding the configuration of the plate solver, it's set to use the internal kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg, and the index files are here: /Users/Andre/Library/Application Support/Astrometry/ (the directory where CloudMakers' Astrometry app downloads the index files). Also, when I installed version 3.1.0 I received a message asking if I'd like to change the Astrometry index file location from the dir /Users/rlancaster/... to my directory /Users/Andre/Library/Application Support/Astrometry/, which I did. I tried update the kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg file with the one from the older version of KStars, but it didn't fix the issue, I just received a confirmation message to change the Astrometry index file location from the dir /Users/Andre/Library/Application Support/Astrometry to my directory /Users/Andre/Library/Application Support/Astrometry/ (same dir), which I did. Here's the output of python --version with the environment variables you said to set: export PATH=/Applications/kstars.app/Contents/MacOS/netpbm/bin:/Applications/kstars.app/Contents/MacOS/python/bin:/usr/local/bin:/usr/local/sbin:/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin:/opt/X11/bin export PYTHONPATH=/Applications/kstars.app/Contents/MacOS/python/bin/site-packages python --version Python 2.7.10 Yes, that all looks good to me As for the message for changing the astrometry.cfg file's path to your own index file directory. I was told by a few people after 3.1.0 that they didn't like that message and/or that it confused them, so in 3.1.1, I changed it to hopefully update the path automatically. I might still need to work on that part though, because when I ran the test that I described the other day on the fresh install, updating that path did work, but not seamlessly. So I will fix that along with anything we figure out about the issue with your system I was searching the net for the error message "Illegal instruction: 4" with astrometry.net and I found two threads of discussion. The first thread is from indilib.org Forum (I guess this one is the one you're talking about in your last message), and another one is this: https://github.com/muammar/mkchromecast/issues/4 Apparently, from the first thread KStars must be build per version of Mac OS X. From the second thread, the solution may be to build using VirtualBox, to force the build to use the older architecture. Yes, as I said when you first posted the error, I assume that the illegal instruction error means that it is trying to run code for a newer OS X version on an older OS X version. What I have done since 2017 is that I built on an old El Capitan machine and then that works on everything. But with the QT change, I had to abandon that and start using the new script. I am hoping that we can get everything to run on the older machines with the right compiler flags. So far until now it has seemed to work very well with craft. Yours is the first issue I have seen with this. The weird part is that I built on Mojave, but I tested on Sierra and High Sierra and it seems to work fine. I don't think there is really any issue with the internal astrometry, python, netpbm, or packages. I do suspect that one of them might not have been compiled for earlier versions of OS X based on your error. But the fact that it doesn't show the error on other Sierra or High Sierra machines though is weird. But thank you for sticking with it and not just getting frustrated. Maybe the issue is not OS related but, instead, hardware related. My machines are pretty old now, and I can’t upgrade them to Mojave due to the old specs. So I think that if you are compiling on Mojave, the binaries may be expecting something not available on my machines. That is an interesting idea. So it might not be the OS X version that is causing the problem. . .it might be something with the older machine's processor instructions or something. You might give the new build script a go. Did you say you tried that or not? Then it would be guaranteed to be compatible with your system. Do you use Parallels or VirtualBox VM as the build environment? I am today (5/29/19) getting the same plate solver "process crashed" message on my macBook Air running Mojave. Because the solver was working on my iMac but not my Air, I completely erased the HD on my Air, re-installed Mojave, re-installed kStars, including GSC and the Astrometry index files (copied from my iMac, where they work), and then ran it. It still crashes. I did this before installing anything funky. Happy to share log file... -stan in Brooklyn Created attachment 120379 [details]
Log file from solver crashing on my Air
Hi Stanley, What we really need is the message for why it crashed. the log file just says it crashed. To get that info, you need to run the solver in the terminal. When you try to plate solve an image, you will see some things print down below, just copy and pasted them into the terminal and see what happens. If everything is good it will solve, if not it will provide a lot more detail than astrometry.net gives back to kstars. Here is an example of the commands based on your log file: Command 1: export PATH=/Applications/kstars.app/Contents/MacOS/netpbm/bin:/Applications/kstars.app/Contents/MacOS/python/bin:/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin Command 2: export PYTHONPATH=/Applications/kstars.app/Contents/MacOS/python/bin/site-packages Command 3 (Changes with every solve): /Applications/kstars.app/Contents/MacOS/astrometry/bin/solve-field -O --no-plots --no-verify --resort --downsample 2 -L 19.3202 -H 26.6923 -u aw -3 57.033 -4 24.1727 -5 30 --config /Applications/kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg -W /private/var/folders/vd/vwnkltm92sx93mrg14plg82r0000gn/T/solution.wcs /private/var/folders/vd/vwnkltm92sx93mrg14plg82r0000gn/T/fitsVpiuva.fits The first two commands will be the same every time on your system. They are also printed in the log file and in the messages box at the bottom of the window. The third command you really have to copy from there because it wlll be based on the latest image. But literally, just try to solve an image, and when it fails, copy the commands to terminal and see what happens. It will print any errors and they should let us know what is wrong on your system. Note that each one of the 3 commands is one line, even though on this forum it might look like more. Created attachment 120435 [details] attachment-21781-0.html Hi Rob, First of all, big THANKS for spending time trying to help me. I’m attaching a text output of the exercise using Terminal. As with cases on the website, it’s an Illegal Instruction error. Don’t know if this output is helpful but I’m sure you’ll tell me. BTW I tried running both 3.0.0 and 3.2.2 with extensions off (shift key at startup) but that didn’t fix anything. Am next going to try v3.2.3 which I just downloaded. Many thanks again, stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On May 30, 2019, at 11:20 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #71 from Rob <rlancaste@gmail.com> --- > Hi Stanley, > > What we really need is the message for why it crashed. the log file just says > it crashed. To get that info, you need to run the solver in the terminal. > When you try to plate solve an image, you will see some things print down > below, just copy and pasted them into the terminal and see what happens. If > everything is good it will solve, if not it will provide a lot more detail than > astrometry.net gives back to kstars. > > Here is an example of the commands based on your log file: > > Command 1: > > export > PATH=/Applications/kstars.app/Contents/MacOS/netpbm/bin:/Applications/kstars.app/Contents/MacOS/python/bin:/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin > > Command 2: > > export > PYTHONPATH=/Applications/kstars.app/Contents/MacOS/python/bin/site-packages > > Command 3 (Changes with every solve): > > /Applications/kstars.app/Contents/MacOS/astrometry/bin/solve-field -O > --no-plots --no-verify --resort --downsample 2 -L 19.3202 -H 26.6923 -u aw -3 > 57.033 -4 24.1727 -5 30 --config > /Applications/kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg -W > /private/var/folders/vd/vwnkltm92sx93mrg14plg82r0000gn/T/solution.wcs > /private/var/folders/vd/vwnkltm92sx93mrg14plg82r0000gn/T/fitsVpiuva.fits > > > The first two commands will be the same every time on your system. They are > also printed in the log file and in the messages box at the bottom of the > window. The third command you really have to copy from there because it wlll > be based on the latest image. But literally, just try to solve an image, and > when it fails, copy the commands to terminal and see what happens. It will > print any errors and they should let us know what is wrong on your system. > > -- > You are receiving this mail because: > You are on the CC list for the bug. Created attachment 120436 [details]
attachment-21781-1.html
Created attachment 120437 [details]
Terminal Saved Output.txt
Hmm. Ok, the illegal instruction error is one I encountered when software was built on a newer version of OS X and then ran on an older version. I think that I carefully built everything with settings that would make it work on older versions of OS X, but I can go back and check all that. What is your OS X version information for reference? I have OS X 10.14.4 Created attachment 120442 [details] attachment-22443-0.html Hi Rob, I’m running a freshly-installed (a few days ago) v10.14.5 of Mojave—the latest commercial release. -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On May 31, 2019, at 10:19 AM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #76 from Rob <rlancaste@gmail.com> --- > Hmm. Ok, the illegal instruction error is one I encountered when software was > built on a newer version of OS X and then ran on an older version. I think > that I carefully built everything with settings that would make it work on > older versions of OS X, but I can go back and check all that. What is your OS > X version information for reference? > > I have OS X 10.14.4 > > -- > You are receiving this mail because: > You are on the CC list for the bug. Created attachment 120443 [details] attachment-22628-0.html Two more elements to the puzzle: -Just tried with v3.2.3 and the same process crash occurred. -I noticed in the solver options that it listed Tycho2 Catalog; but I had removed everything in my various attempts to fix the problem. I therefore re-installed Tycho2. Didn’t change anything. RE: OS X 10.14.5—maybe Apple changed something in the latest OS release which broke the solving? NO because solving works on my iMac with 10.14.5. Just not the Air. -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On May 31, 2019, at 10:19 AM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #76 from Rob <rlancaste@gmail.com> --- > Hmm. Ok, the illegal instruction error is one I encountered when software was > built on a newer version of OS X and then ran on an older version. I think > that I carefully built everything with settings that would make it work on > older versions of OS X, but I can go back and check all that. What is your OS > X version information for reference? > > I have OS X 10.14.4 > > -- > You are receiving this mail because: > You are on the CC list for the bug. Created attachment 120444 [details] attachment-23276-0.html Just tried running v3.2.3 in Terminal. Got a different error message—please see attached. -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On May 30, 2019, at 11:20 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #71 from Rob <rlancaste@gmail.com> --- > Hi Stanley, > > What we really need is the message for why it crashed. the log file just says > it crashed. To get that info, you need to run the solver in the terminal. > When you try to plate solve an image, you will see some things print down > below, just copy and pasted them into the terminal and see what happens. If > everything is good it will solve, if not it will provide a lot more detail than > astrometry.net gives back to kstars. > > Here is an example of the commands based on your log file: > > Command 1: > > export > PATH=/Applications/kstars.app/Contents/MacOS/netpbm/bin:/Applications/kstars.app/Contents/MacOS/python/bin:/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin > > Command 2: > > export > PYTHONPATH=/Applications/kstars.app/Contents/MacOS/python/bin/site-packages > > Command 3 (Changes with every solve): > > /Applications/kstars.app/Contents/MacOS/astrometry/bin/solve-field -O > --no-plots --no-verify --resort --downsample 2 -L 19.3202 -H 26.6923 -u aw -3 > 57.033 -4 24.1727 -5 30 --config > /Applications/kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg -W > /private/var/folders/vd/vwnkltm92sx93mrg14plg82r0000gn/T/solution.wcs > /private/var/folders/vd/vwnkltm92sx93mrg14plg82r0000gn/T/fitsVpiuva.fits > > > The first two commands will be the same every time on your system. They are > also printed in the log file and in the messages box at the bottom of the > window. The third command you really have to copy from there because it wlll > be based on the latest image. But literally, just try to solve an image, and > when it fails, copy the commands to terminal and see what happens. It will > print any errors and they should let us know what is wrong on your system. > > -- > You are receiving this mail because: > You are on the CC list for the bug. Created attachment 120445 [details]
attachment-23276-1.html
Created attachment 120446 [details]
Terminal Saved Output 2.txt
That's because a forward slash is missing. There is no file called: Applications/kstars.app/Contents/MacOS/astrometry/bin/solve-field it is called: /Applications/kstars.app/Contents/MacOS/astrometry/bin/solve-field I have a suspicion that this may be a python issue. I will see what I can do about it. Created attachment 120449 [details] attachment-26555-0.html Ah yes, my mistake. Sorry! Here it is again with the correct commands; Illegal Instruction 4 once again. Running v3.2.3. OK, I’m very grateful, Rob. -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On May 31, 2019, at 12:12 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #82 from Rob <rlancaste@gmail.com> --- > That's because a forward slash is missing. There is no file called: > > Applications/kstars.app/Contents/MacOS/astrometry/bin/solve-field > > it is called: > > /Applications/kstars.app/Contents/MacOS/astrometry/bin/solve-field > > -- > You are receiving this mail because: > You are on the CC list for the bug. Created attachment 120450 [details]
attachment-26555-1.html
Created attachment 120451 [details]
Terminal Saved Output 3.txt
So I worked this weekend on trying to get python better wrapped up inside the app bundle. Can you check to see if this helps? https://drive.google.com/open?id=19hX15yz1Ryr-xkJxsWKYWeb_y7vcDl6T This one is better than the last one I just posted. The last one had python that was compatible with 10.14 Mojave computers. This should be compatible with everything down to 10.12 I think. drive.google.com/open?id=1wmkuli2tjKK9GF3oZwnJznkIYWpx7OY_ Thanks, Rob Created attachment 120523 [details] attachment-6462-0.html Hi Rob, Thank you so much for working on this during the weekend! I’m sorry to say it still didn’t work—I used the last build you sent me. As previously, after it didn’t work, I ran it again in Terminal and am sending you the logs. FYI I am running Mojave 10.14.5 on my Air. I am heading up to the mountains tonight and will try it again with real equipment attached rather than the simulators. Will let you know tomorrow how that went. Many thanks again, -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Jun 3, 2019, at 12:44 AM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #88 from Rob <rlancaste@gmail.com> --- > This one is better than the last one I just posted. The last one had python > that was compatible with 10.14 Mojave computers. This should be compatible with > everything down to 10.12 I think. > > drive.google.com/open?id=1wmkuli2tjKK9GF3oZwnJznkIYWpx7OY_ > > Thanks, > > Rob > > -- > You are receiving this mail because: > You are on the CC list for the bug. Created attachment 120524 [details]
attachment-6462-1.html
Created attachment 120525 [details]
attachment-6462-2.html
Created attachment 120526 [details]
log_10-59-02.txt
Created attachment 120527 [details]
Terminal Saved Output
For now if you want to get it working, since this is proving to be more difficult than I first thought. The following commands will install homebrew, python3, Astropy, and numpy to your system. The final command will delete my broken internal python. Just type the commands into a terminal. Plate solving in Kstars should work as expected. I would still like to resolve this. I don't want users to have to do all this if possible. /usr/bin/ruby -e "$(curl -fsSL raw.githubusercontent.com/Homebrew/install/master/install )" brew install python3 ln -s /usr/local/bin/python3 /usr/local/bin/python ln -s /usr/local/bin/pip3 /usr/local/bin/pip pip install astropy rm -r /Applications/kstars.app/Contents/MacOS/python/* Thanks, Rob Thanks, Rob,
That’s still wonderfully kind of you.
In the meantime I will see if I can use Cloudmakers’ software…
Let me know if there’s anything I can do to help you. I love Ekos!
-stan
> On Jun 4, 2019, at 12:46 AM, Rob <bugzilla_noreply@kde.org> wrote:
>
> https://bugs.kde.org/show_bug.cgi?id=405437
>
> --- Comment #94 from Rob <rlancaste@gmail.com> ---
> For now if you want to get it working, since this is proving to be more
> difficult than I first thought.
> The following commands will install homebrew, python3, Astropy, and numpy to
> your system.
> The final command will delete my broken internal python.
> Just type the commands into a terminal.
> Plate solving in Kstars should work as expected.
> I would still like to resolve this. I don't want users to have to do all this
> if possible.
>
> /usr/bin/ruby -e "$(curl -fsSL
> raw.githubusercontent.com/Homebrew/install/master/install )"
> brew install python3
> ln -s /usr/local/bin/python3 /usr/local/bin/python
> ln -s /usr/local/bin/pip3 /usr/local/bin/pip
> pip install astropy
> rm -r /Applications/kstars.app/Contents/MacOS/python/*
>
> Thanks,
>
> Rob
>
> --
> You are receiving this mail because:
> You are on the CC list for the bug.
Created attachment 120613 [details]
screenshot of python installer
Ok, I worked for many hours during the last couple of days on a better solution. I think I have it. I have set up an installer button and put it into the Astrometry configuration options for now. (I will see what others think about it, they might end up being moved somewhere else. But for right now, I would like to see if this new configuration and installer works. Please test this to make sure plate solving works correctly. Before you can use it, you MUST run the installer. There is no more embedded python. Please see my screenshot for details. drive.google.com/open?id=1K3anbNAcsLC3ddy4UVYpMbcjYC60v8xl Please test plate solving with your system, to be even more helpful, you might want to try removing astropy, pyfits, numpy, python, and/or homebrew to see how the installers work. Let me know how it goes. Thanks, Rob Created attachment 120621 [details] attachment-23871-0.html Hi Rob, Well, I just tried it out, clicked on the Setup button and followed instructions. I believe the Command Line Tools were already installed. Your script installed HomeBrew, etc. Unless I did something wrong…I did get an error message on the CLT installation saying “Can’t install the software because it is not currently available from the Software Update server” but my understanding is that that message means the command line tools were already installed... The bad news is that it still didn’t work, illegal instruction errors #4. I’m attaching both the terminal output of my installing it, as well as the terminal output from running kStars in Terminal. I’m really sorry! -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Jun 5, 2019, at 9:43 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #96 from Rob <rlancaste@gmail.com> --- > Created attachment 120613 [details] > --> https://bugs.kde.org/attachment.cgi?id=120613&action=edit > screenshot of python installer > > -- > You are receiving this mail because: > You are on the CC list for the bug. Created attachment 120622 [details]
attachment-23871-1.html
Created attachment 120623 [details]
attachment-23871-2.html
Created attachment 120624 [details]
Terminal Saved Output 6-6-19
Created attachment 120625 [details]
Terminal Saved Output 2
I think your Xcode command line tools are not installed correctly, so the homebrew installation failed to complete Try this in a terminal: xcode-select --install And if that fails, maybe this: /usr/bin/sudo /usr/bin/xcode-select --switch /Library/Developer/CommandLineTools Do you have Xcode on your machine? If so have you started it and accepted the user agreement? Created attachment 120639 [details] attachment-31301-0.html Just tried the first command. 1. A dialog box popped up asking me to install the Command Line Developer Tools, with an option to install Xcode. I just now chose to install Xcode, which took me to the App Store, where I was able to access it. It’s in the process of downloading now. I see that choosing this option also includes the CMDT as well as xCode. I had NOT chosen this option previously, so perhaps when I ran the new build, my computer was not correctly set up for it. I’ll install xCode, accept the User Agreement and then try again, and let you know what happens. Thank you, Rob, stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com --- Comment #103 from Rob <rlancaste@gmail.com <mailto:rlancaste@gmail.com>> --- I think your Xcode command line tools are not installed correctly, so the homebrew installation failed to complete Try this in a terminal: xcode-select --install And if that fails, maybe this: /usr/bin/sudo /usr/bin/xcode-select --switch /Library/Developer/CommandLineTools Do you have Xcode on your machine? If so have you started it and accepted the user agreement? Created attachment 120642 [details] Terminal Saved Output 3 Hi Rob, I downloaded and installed Xcode. Then I re-booted, then re-downloaded your special version of kStars from today. Tried it out within Terminal and it didn’t work. I’m attaching the Terminal record of my installation, as well as the three command lines from the failure. It said Homebrew was not installed and put up a terminal window to install it. I said yes but the installation failed. I did already install Xcode… So then I tried the set of instructions you had sent me to install home-brew, in your email from 1:03am. Didn’t work, am attaching the Terminal text (Terminal Saved Output last). I’m mystified...?? -stan > On Jun 6, 2019, at 4:38 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #103 from Rob <rlancaste@gmail.com> --- > I think your Xcode command line tools are not installed correctly, so the > homebrew installation failed to complete > > Try this in a terminal: xcode-select --install > > And if that fails, maybe this: > > /usr/bin/sudo /usr/bin/xcode-select --switch > /Library/Developer/CommandLineTools > > Do you have Xcode on your machine? If so have you started it and accepted the > user agreement? > > -- > You are receiving this mail because: > You are on the CC list for the bug. Created attachment 120643 [details]
Terminal Saved Output last
Look at the error message that printed here: Basidium:~ sf$ /usr/bin/ruby -e "$(curl -fsSL > raw.githubusercontent.com/Homebrew/install/master/install )" curl: no URL specified! In copying the command to terminal, you added a "return." I think this is actually this forum's fault for that because it put it on two lines. You can copy it from their official website: https://brew.sh Created attachment 120659 [details] attachment-16410-0.html Hi Rob, OK, I installed HomeBrew using the link in your email below. No problem. I then launched kStars again. I clicked in the python “Setup” button in the Astrometry configuration pane. It said Homebrew was installed already (good!) and Python3 installed (good?). Got an “All installations are complete and ready to use.” (Great!) I then tried clicking on Capture & Slew again. Got the same error message: “Error starting solver: Process crashed” after reading input file 1 of 1: “/private/var/folders/vd/vwnkltm92sx93mrg14plg82r0000gn/T/fitslixbl.r.fits” I did check, online plate solving works, just not offline. Rebooted the computer and tried again, just in case. Same result (sigh)… Any thoughts? Thanks again, stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Jun 6, 2019, at 7:46 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #107 from Rob <rlancaste@gmail.com> --- > Look at the error message that printed here: > > Basidium:~ sf$ /usr/bin/ruby -e "$(curl -fsSL >> raw.githubusercontent.com/Homebrew/install/master/install )" > curl: no URL specified! > > > In copying the command to terminal, you added a "return." I think this is > actually this forum's fault for that because it put it on two lines. > > You can copy it from their official website: > > https://brew.sh > > -- > You are receiving this mail because: > You are on the CC list for the bug. Created attachment 120661 [details] attachment-16518-0.html Just checked again, when I click in the Astrometry “Setup” box again, it says, “Homebrew, Python and Astropy are already installed.” -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Jun 7, 2019, at 10:24 AM, Stanley Fertig <stanleyf@nyc.rr.com> wrote: > > Hi Rob, > > OK, I installed HomeBrew using the link in your email below. No problem. > > I then launched kStars again. I clicked in the python “Setup” button in the Astrometry configuration pane. It said Homebrew was installed already (good!) and Python3 installed (good?). Got an “All installations are complete and ready to use.” (Great!) > > I then tried clicking on Capture & Slew again. Got the same error message: “Error starting solver: Process crashed” after reading input file 1 of 1: “/private/var/folders/vd/vwnkltm92sx93mrg14plg82r0000gn/T/fitslixbl.r.fits” > > I did check, online plate solving works, just not offline. > > Rebooted the computer and tried again, just in case. Same result (sigh)… > > Any thoughts? > > Thanks again, > stan > > > -- > Stanley Fertig > 35 Sidney Place > Brooklyn, New York 11201 > +1 917-361-0632 > stanleyf@nyc.rr.com <mailto:stanleyf@nyc.rr.com> > > > > > > > > > > >> On Jun 6, 2019, at 7:46 PM, Rob <bugzilla_noreply@kde.org <mailto:bugzilla_noreply@kde.org>> wrote: >> >> https://bugs.kde.org/show_bug.cgi?id=405437 <https://bugs.kde.org/show_bug.cgi?id=405437> >> >> --- Comment #107 from Rob <rlancaste@gmail.com> --- >> Look at the error message that printed here: >> >> Basidium:~ sf$ /usr/bin/ruby -e "$(curl -fsSL >>> raw.githubusercontent.com/Homebrew/install/master/install )" >> curl: no URL specified! >> >> >> In copying the command to terminal, you added a "return." I think this is >> actually this forum's fault for that because it put it on two lines. >> >> You can copy it from their official website: >> >> https://brew.sh >> >> -- >> You are receiving this mail because: >> You are on the CC list for the bug. > So just to clarify, the machine causing the astrometry error is running 10.14.5? Also this is a MacBook Air? And when you type: which python3 into a terminal, what prints? Created attachment 120663 [details] attachment-20222-0.html Hi Rob, Yes, it’s an Air running 10.14.5. When I typed “which python3” into Terminal, it gave me the following: /usr/local/bin/python3 That’s all… -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Jun 7, 2019, at 1:40 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #110 from Rob <rlancaste@gmail.com> --- > So just to clarify, the machine causing the astrometry error is running > 10.14.5? Also this is a MacBook Air? > > And when you type: > > which python3 > > into a terminal, what prints? > > -- > You are receiving this mail because: > You are on the CC list for the bug. Created attachment 120664 [details]
Test file for astrometry
Ok then lets see if there is something different about your processor or if it is an issue with solve-field or netpbm. This will give us information about the processor, so we can see if anything is weird there: sysctl -n machdep.cpu.brand_string This should give you a list of what solve-field can do (I assume this won't give a illegal instruction error, if it does, that says a lot): /Applications/kstars.app/Contents/MacOS/astrometry/bin/solve-field -h Now lets run a test of solve field that should not use netpbm: This should download a test file that I made for astrometry to use: curl -o ~/Downloads/test.xyls https://bugsfiles.kde.org/attachment.cgi?id=120664 This will put python3 in your path for the solver: export PATH=/usr/local/opt/python/libexec/bin:$PATH And this should try to solve it in solve field: /Applications/kstars.app/Contents/MacOS/astrometry/bin/solve-field -O --no-verify --config /Applications/kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg ~/Downloads/test.xyls --x-column X_IMAGE --y-column Y_IMAGE --width 640 --height 512 Created attachment 120667 [details] attachment-23175-0.html Hi Rob, I got as far as the download of the test file; then I’m not sure I did the right thing in Terminal for the test file download I marked below with question marks. I’m attaching a printout of the Terminal session. Can you please help me with this? then I’ll do the rest… Thanks, Stan > On Jun 7, 2019, at 2:44 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #113 from Rob <rlancaste@gmail.com> --- > Ok then lets see if there is something different about your processor or if it > is an issue with solve-field or netpbm. > > This will give us information about the processor, so we can see if anything is > weird there: > > sysctl -n machdep.cpu.brand_string > > This should give you a list of what solve-field can do (I assume this won't > give a illegal instruction error, if it does, that says a lot): > > /Applications/kstars.app/Contents/MacOS/astrometry/bin/solve-field -h > > Now lets run a test of solve field that should not use netpbm: > > This should download a test file that I made for astrometry to use: > curl -o ~/Downloads/test.xyls. ??????????????????????????? > https://bugsfiles.kde.org/attachment.cgi?id=120664 <https://bugsfiles.kde.org/attachment.cgi?id=120664>. ?????????????????? > This will put python3 in your path for the solver: > export PATH=/usr/local/opt/python/libexec/bin:$PATH > And this should try to solve it in solve field: > /Applications/kstars.app/Contents/MacOS/astrometry/bin/solve-field -O > --no-verify --config > /Applications/kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg > ~/Downloads/test.xyls --x-column X_IMAGE --y-column Y_IMAGE --width 640 > --height 512 > > -- > You are receiving this mail because: > You are on the CC list for the bug. Created attachment 120668 [details]
attachment-23175-1.html
Created attachment 120669 [details]
Terminal Saved Output 6:7:19
curl -o ~/Downloads/test.xyls "https://bugsfiles.kde.org/attachment.cgi?id=120664" Do you see ??? here or a URL? Oh I think I see, the url got copied to the next line in your terminal. This is just one line that you need to copy and paste in the terminal: curl -o ~/Downloads/test.xyls "https://bugsfiles.kde.org/attachment.cgi?id=120664" this is also just one line of code, no return in the middle of it or anything /Applications/kstars.app/Contents/MacOS/astrometry/bin/solve-field -O --no-verify --config /Applications/kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg ~/Downloads/test.xyls --x-column X_IMAGE --y-column Y_IMAGE --width 640 --height 512 Created attachment 120671 [details]
attachment-23392-0.txt
OK, it took me a bit to figure out how to deal with the line breaks, all good now.
I’ve run it all, the Terminal session text file is attached.
Good luck and thanks!
-stan
Created attachment 120672 [details]
Terminal Saved Output 6:7:2019 2
I looked at your terminal output, it was very close, but somehow a space disappeared between -O and --no-verify. note the message: solve-field: unrecognized option `---no-verify' Here is the command again. /Applications/kstars.app/Contents/MacOS/astrometry/bin/solve-field -O --no-verify --config /Applications/kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg ~/Downloads/test.xyls --x-column X_IMAGE --y-column Y_IMAGE --width 640 --height 512 Created attachment 120673 [details]
attachment-23813-0.txt
Ok Rob,
I tried it with and without a space. Didn’t appear to like either. I’m attaching the Terminal file for both, first without and then with an extra space.
!!
Thanks,
Stan
Created attachment 120674 [details]
Terminal Output 6:7:19 3.rtf
Still a problem with spaces. Here I will attach a text file with the commands in it. Created attachment 120677 [details]
terminal commands for astrometry test
Created attachment 120678 [details] attachment-24560-0.html Hi Rob, I’m not sure what to do with this. I clicked on the link and it was a blank form… -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Jun 7, 2019, at 5:57 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #126 from Rob <rlancaste@gmail.com> --- > Created attachment 120677 [details] > --> https://bugs.kde.org/attachment.cgi?id=120677&action=edit > terminal commands for astrometry test > > -- > You are receiving this mail because: > You are on the CC list for the bug. Created attachment 120679 [details] attachment-24587-0.html I see a url, that’s all. Clicked on it and it’s a blank form. -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Jun 7, 2019, at 5:00 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #117 from Rob <rlancaste@gmail.com> --- > curl -o ~/Downloads/test.xyls > "https://bugsfiles.kde.org/attachment.cgi?id=120664" > > Do you see ??? here or a URL? > > -- > You are receiving this mail because: > You are on the CC list for the bug. Created attachment 120680 [details] attachment-24613-0.html Oh, wait, I didn’t run the line above, let me try again. -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Jun 7, 2019, at 5:00 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #117 from Rob <rlancaste@gmail.com> --- > curl -o ~/Downloads/test.xyls > "https://bugsfiles.kde.org/attachment.cgi?id=120664" > > Do you see ??? here or a URL? > > -- > You are receiving this mail because: > You are on the CC list for the bug. When you click this link, https://bugs.kde.org/attachment.cgi?id=120677 you see a blank form? You should see a text file with the 3 commands in it, labelled with command 1, command 2, command 3 They are the same commands I was telling you to run, just carefully put into a text file so that there are no more issues with copying and pasting from a forum. Created attachment 120683 [details] attachment-26856-0.html Hi Rob, Yes, I do see the commands. Wasn’t the case before. Will try and let you know how it goes. -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Jun 7, 2019, at 6:27 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #130 from Rob <rlancaste@gmail.com> --- > When you click this link, > > https://bugs.kde.org/attachment.cgi?id=120677 > > you see a blank form? You should see a text file with the 3 commands in it, > labelled with command 1, command 2, command 3 > > -- > You are receiving this mail because: > You are on the CC list for the bug. That sounds good. My thought is, if it runs without an illegal argument exception, then the issue is probably with netpbm. If it fails, the issue is probably related to python. If it is python, then I believe that the issue is with some instructions in the python binaries that your processor can't handle. We can probably fix that by running: brew reinstall --build-from-source python3 which would be sure to make sure that your processor can handle everything since it would be the one building it. Created attachment 120699 [details]
attachment-8937-0.txt
Hi Rob,
OK, I re-installed Homebrew as per your code, ran everything including the text file with the 3 commands, etc.
Still getting an illegal instruction error! If you look at the bottom of the file, appears to be when it attempts to read the FITS file downloaded...
Long Terminal file attached.
Thanks,
stan
Created attachment 120700 [details]
Terminal Saved Output 6:8:19
Ok, so it's not python, and probably not netbpm since my last command shouldn't use netpbm. Let's try this. brew install astrometry-net And then the same commands as I sent you in the text file before to see if it works, with one difference, we will use the homebrew solve-field: command 1: (the same) curl -o ~/Downloads/test.xyls "https://bugsfiles.kde.org/attachment.cgi?id=120664" command 2: (the same) export PATH=/usr/local/opt/python/libexec/bin:$PATH command 3: (a slight change, just the path to solve-field) /usr/local/bin/solve-field -O --no-verify --config /Applications/kstars.app/Contents/MacOS/astrometry/bin/astrometry.cfg ~/Downloads/test.xyls --x-column X_IMAGE --y-column Y_IMAGE --width 640 --height 512 You might copy and paste the commands from the text file again, except the last one needs that slight change. Created attachment 120703 [details]
attachment-12009-0.txt
Hi Rob,
Aha! That seemed to work! Please see the attached Terminal printout.
So do you therefore know what the problem has been?
Many thanks again,
stan
Created attachment 120704 [details]
Terminal Saved Output 6:8:19 2
Yes, that did solve. Actually I figured that it would. Basically through the process of elimination, we determined that your system has a problem with the astrometry.net that I built on my system for kstars. The test you just ran was basically running astrometry.net entirely in homebrew. The only thing we used from the app bundle was the configuration file. An easy solution for you in the short term would be to just go into the astrometry options in kstars and set it to use the homebrew solver for now, since you have installed it. As to why the internal one doesn't work, I will have to look into that since I think it should. So you can go ahead and start using it with the homebrew astrometry solver on this system and do some astronomy. But if I come up with a possible solution for the internal solver, I might ask you to see if you can test it again for me. Thanks, Rob Created attachment 120705 [details] attachment-12956-0.html Rob, you are truly an awesome person. I can’t believe the number of hours and amount of patience you put into helping me. I do have to believe that if I had this problem, then others probably do too. Where on Earth are you? (Just curious). So, practically…how do I do this from inside kStars? I just tried opening the last version you sent me and tried it without changing any settings and it failed. What setting do I need to do in it? Thanks again, stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Jun 8, 2019, at 3:12 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #139 from Rob <rlancaste@gmail.com> --- > Yes, that did solve. Actually I figured that it would. Basically through the > process of elimination, we determined that your system has a problem with the > astrometry.net that I built on my system for kstars. The test you just ran was > basically running astrometry.net entirely in homebrew. The only thing we used > from the app bundle was the configuration file. > > An easy solution for you in the short term would be to just go into the > astrometry options in kstars and set it to use the homebrew solver for now, > since you have installed it. As to why the internal one doesn't work, I will > have to look into that since I think it should. So you can go ahead and start > using it with the homebrew astrometry solver on this system and do some > astronomy. But if I come up with a possible solution for the internal solver, > I might ask you to see if you can test it again for me. > > Thanks, > > Rob > > -- > You are receiving this mail because: > You are on the CC list for the bug. Well thanks, considering the number of hours that I put into getting KStars to work on Macs, and then the build scripts, etc. If something like this is preventing it from working on some computers, I think that needs to be fixed. So yes I did put in some time trying to fix it. I am in Wilmington Delaware, USA. In order to make KStars use the external, homebrew based astrometry.net, you would need to go into the Astrometry settings in the align panel and uncheck the box that makes it use the internal solver and make sure it has the right path to the solver. I will send a screenshot. Created attachment 120734 [details]
Astrometry setting
Created attachment 120736 [details]
attachment-14467-0.txt
Thanks indeed, Rob. On behalf of the other Mac kstars users, I’m very grateful.
Just tried it…almost there. But I apparently had a WCS error at the end of the process—please see the attached log file from kStars.
I’m hoping this isn’t major...
(Sigh!)
-stan
Created attachment 120737 [details]
kstars log 6:9:19.rtf
So you just unchecked the box to use the solver in homebrew instead of the solver in the app bundle. Maybe uncheck the box right next to the one you just did to use the homebrew wcsinfo instead of the wcsinfo in the app bundle? wcsinfo is a part of astrometry.net and should be installed already just like solve-field. Created attachment 120738 [details] attachment-15965-0.html Ha! That did it. Had replaced the link for the “solver” checkbox but hadn’t touched the wcsinfo box. I just simply unchecked the wcsinfo box as you suggested and it WORKED. Woo-hoo! I’m back in business doing astrophotography thanks to you! Thank you so much, Rob! -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Jun 9, 2019, at 2:51 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #145 from Rob <rlancaste@gmail.com> --- > So you just unchecked the box to use the solver in homebrew instead of the > solver in the app bundle. Maybe uncheck the box right next to the one you just > did to use the homebrew wcsinfo instead of the wcsinfo in the app bundle? > wcsinfo is a part of astrometry.net and should be installed already just like > solve-field. > > -- > You are receiving this mail because: > You are on the CC list for the bug. Created attachment 120739 [details] attachment-16081-0.html BTW if I can ever be of help to you testing or helping to debug newer versions, of course I’m available. -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Jun 9, 2019, at 2:51 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #145 from Rob <rlancaste@gmail.com> --- > So you just unchecked the box to use the solver in homebrew instead of the > solver in the app bundle. Maybe uncheck the box right next to the one you just > did to use the homebrew wcsinfo instead of the wcsinfo in the app bundle? > wcsinfo is a part of astrometry.net and should be installed already just like > solve-field. > > -- > You are receiving this mail because: > You are on the CC list for the bug. Rob, After struggling with similar issues when using CCDciel, I found their solution to the problem of Astrometry.net on Mac OS X easier to handle: https://www.ap-i.net/ccdciel/en/documentation/astrometry.net Their way of solving this issue was to use a script to invoke the CloudMaker's Astrometry.app plate solver, that has no compatibility issues between Mac OS X versions, and is fully customizable. I think it's a nice solution, since we already have to use CloudMaker's Astrometry.app to manage the local Astrometry.net catalogs. Regards, Andre Thanks for the idea. I think we almost have it all sorted out. In the past I have tried the cloud makers app for plate solving and I have found it to be nice for downloading and managing index files, but not as nice for the actual solving. The really good news is that the stuff you had to do from the command line is no longer necessary because I found a way to integrate the error output so that it all prints in the Ekos window now. No more needing to copy and paste the commands to see what is wrong. When you said. That you need to use the cloud makers app to manage your index files, why is that necessary? I spent a lot of time about 2 years ago making an index file manager in Ekos. Is that not working for you? PS I am on vacation in Alaska and won’t be back for two weeks. I have spotty internet service. So I might not get messages right away. Created attachment 122049 [details]
solve-field.sh
Hi Rob, I tried KStars 3.3.4, and even installed the required Python libraries using Mac Ports, but I couldn't make it work. Even the workaround (replacement of the astrometry directory with the one from the older version of KStars) that I used with previous versions. Alternatively, I started to work on the plate solving script (solve-field.sh, based on the one from CCDciel), to use the internal astrometry engine from Astrometry.app. I managed to make it produce the required WCS file in the T temp directory, but I'm still receiving the "Solver failed. Try again." message from Ekos. I also tried to replace the wcsinfo with the one that comes with Astrometry.app (/Applications/Astrometry.app/Contents/MacOS/wcsinfo). Here's the output from the log file (verbose enabled): [2019-08-10T14:09:54.463 -03 INFO ][ org.kde.kstars.ekos.align] - "Capturing image..." [2019-08-10T14:10:01.727 -03 INFO ][ org.kde.kstars.fits] - Loading FITS file "/private/var/folders/t2/pdxc8twj5q7c1mfhc61y5ts40000gw/T/fitsGVRPDr.fits" [2019-08-10T14:10:02.060 -03 INFO ][ org.kde.kstars.ekos.align] - "Image received." [2019-08-10T14:10:02.101 -03 INFO ][ org.kde.kstars.ekos.align] - "Starting solver..." [2019-08-10T14:10:06.976 -03 INFO ][ org.kde.kstars.ekos.align] - "Solver failed. Try again." Could you please help me check what I might be missing here? Thanks, Andre Hmm, I had the same problem but i found it went away completely with v3.3.3 and v3.3.4. Did a fresh installation of Python 3 and HomeBrew, then used the Setup button in the Astrometry panel to check all was well and finish setup. Since then, no problems at all... Am running latest version of Mojave on a macBook Air. -stan in Brooklyn (In reply to Andre Kovacs from comment #0) > SUMMARY > > After I upgraded KStars from version 2.9.8 to 3.1.0, and 3.1.1 afterwards, > the internal astrometry server crashes every time I try to plate solve using > Ekos Align feature. > Also, when I installed version 3.1.0 I received a message asking if I'd like > to change the Astrometry configuration file from the dir > /Users/rlancaster/... to my directory, which I did. > I also verified the configurations of the astrometry server under the Align > options button, and it is exactly as I used before, except for the Index > File Location field that didn't exist in the previous version, but the > Astrometry.cfg file is exactly the same. > > STEPS TO REPRODUCE > 1. Open Ekos and go to Align tab; > 2. Click capture & Solve; > 3. Output console displays a message that the local solver crashed (it works > fine if I choose the Online solver option, instead of the Offline option). > > OBSERVED RESULT > > The output console displays the following messages: > 2019-03-13T12:52:28 Solver failed. Try again. > 2019-03-13T12:52:28 Error starting solver: Process crashed > 2019-03-13T12:52:28 Starting solver... > 2019-03-13T12:52:28 Image received. > 2019-03-13T12:52:15 Capturing image... > > EXPECTED RESULT > > Here's the output for the online solver: > 2019-03-13T13:07:42 Solution coordinates: RA (01h 09m 16s) DEC (-55° 07' > 49") Telescope Coordinates: RA (01h 09m 14s) DEC (-55° 08' 05") > 2019-03-13T13:07:42 Solver completed in 65 seconds. > 2019-03-13T13:06:39 Upload complete. Waiting for astrometry.net solver to > complete... > 2019-03-13T13:06:37 Uploading file... > 2019-03-13T13:06:37 Image received. > 2019-03-13T13:06:24 Capturing image... > > SOFTWARE/OS VERSIONS > Windows: > MacOS: High Sierra (10.13.6) > Linux/KDE Plasma: > (available in About System) > KDE Plasma Version: > KDE Frameworks Version: > Qt Version: > > ADDITIONAL INFORMATION I had the same problem, as Rob knows.
After I upgraded to v3.3.3 and v3.3.4, the problem went away completely. Did a fresh install of Python 3 and of HomeBrew, and then checked the Setup button in the astrometry panel. All works fine now.
Have tried it both on an iMac and on my macBook Air. No more problems.
-stan in Brooklyn
> On Aug 10, 2019, at 2:54 PM, Andre Kovacs <bugzilla_noreply@kde.org> wrote:
>
> https://bugs.kde.org/show_bug.cgi?id=405437
>
> --- Comment #151 from Andre Kovacs <andre.kovacs@gmail.com> ---
> Hi Rob,
>
> I tried KStars 3.3.4, and even installed the required Python libraries using
> Mac Ports, but I couldn't make it work. Even the workaround (replacement of the
> astrometry directory with the one from the older version of KStars) that I used
> with previous versions.
>
> Alternatively, I started to work on the plate solving script (solve-field.sh,
> based on the one from CCDciel), to use the internal astrometry engine from
> Astrometry.app.
> I managed to make it produce the required WCS file in the T temp directory, but
> I'm still receiving the "Solver failed. Try again." message from Ekos. I also
> tried to replace the wcsinfo with the one that comes with Astrometry.app
> (/Applications/Astrometry.app/Contents/MacOS/wcsinfo).
> Here's the output from the log file (verbose enabled):
> [2019-08-10T14:09:54.463 -03 INFO ][ org.kde.kstars.ekos.align] -
> "Capturing image..."
> [2019-08-10T14:10:01.727 -03 INFO ][ org.kde.kstars.fits] - Loading
> FITS file
> "/private/var/folders/t2/pdxc8twj5q7c1mfhc61y5ts40000gw/T/fitsGVRPDr.fits"
> [2019-08-10T14:10:02.060 -03 INFO ][ org.kde.kstars.ekos.align] - "Image
> received."
> [2019-08-10T14:10:02.101 -03 INFO ][ org.kde.kstars.ekos.align] - "Starting
> solver..."
> [2019-08-10T14:10:06.976 -03 INFO ][ org.kde.kstars.ekos.align] - "Solver
> failed. Try again."
>
> Could you please help me check what I might be missing here?
>
> Thanks,
> Andre
>
> --
> You are receiving this mail because:
> You are on the CC list for the bug.
Hi Stan, Thank you for the update, but I use Mac Ports instead of Homebrew (for development reasons), so the Homebrew libraries conflicts with my existing Mac Port libraries. Thanks, Andre Created attachment 122053 [details]
solve-field.sh
Updated script to match the generated file names by Ekos
Hi Rob, I updated the script to generate the files according to the names used by Ekos' internal solve-field. But I have a couple of questions: 1- The WCS file must be named as "solution.wcs" or could be the FITS file name with extension .wcs? 2- The file with .new extension is the updated FITS file with the WCS information (output of new-wcs), right? Finally, do you expect a return value from the solve-field, or only the generated files in the T temp directory? Thanks, Andre Created attachment 122054 [details]
solve-field.sh
Updated script to generate the WCS file as "solution.wcs"
Hi Rob, I tested a version of the script that generates the WCS file as "solution.wcs", and now the error changed to "WCS header missing or corrupted. Solver failed.", using the internal wcsinfo. If I configure Ekos to use /Applications/Astrometry.app/Contents/MacOS/wcsinfo, Ekos logs "Solver completed in 6 seconds.", but the solving process in Ekos keeps running. Any idea of what might be happening? Do I need another script for wcsinfo? Thanks, Andre Created attachment 122055 [details]
solution.wcs
WCS file generated by the solve-field script
Hi Rob, I was checking the internal wcsinfo against the solution.wcs generated by the script (attached file), and I found that it's aborting with a message "Illegal instruction: 4". Here's the output of the internal wcsinfo (Revision 0.76): /Applications/Astronomia/KStars_3.3.4.app/Contents/MacOS/astrometry/bin/wcsinfo /private/var/folders/t2/pdxc8twj5q7c1mfhc61y5ts40000gw/T/solution.wcs crpix0 1761.64859009 crpix1 2545.80651855 crval0 78.4784437837 crval1 -8.0552038294 ra_tangent 78.4784437837 dec_tangent -8.0552038294 pixx_tangent 1761.64859009 pixy_tangent 2545.80651855 imagew 5208 imageh 3476 cd11 0.00017980392322 cd12 9.8135450351e-08 cd21 -9.8135450351e-08 cd22 0.00017980392322 det 3.23294604359e-08 parity 1 pixscale 0.647294220003 orientation 0.031271546 ra_center 78.631476441 dec_center -8.20041425873 orientation_center 0.0098274657 ra_center_h 5 ra_center_m 14 ra_center_s 31.5543458478 dec_center_sign -1 dec_center_d 8 dec_center_m 12 dec_center_s 1.49133141871 ra_center_hms 05:14:31.554 dec_center_dms -08:12:01.491 ra_center_merc 0.21842077 dec_center_merc 0.4771429 fieldarea 0.58526 fieldw 56.18 fieldh 37.5 fieldunits arcminutes Illegal instruction: 4 Here's the output of the Astrometry.app wcsinfo: /Applications/Astrometry.app/Contents/MacOS/wcsinfo /private/var/folders/t2/pdxc8twj5q7c1mfhc61y5ts40000gw/T/solution.wcs crpix0 1761.64859009 crpix1 2545.80651855 crval0 78.4784437837 crval1 -8.0552038294 ra_tangent 78.4784437837 dec_tangent -8.0552038294 pixx_tangent 1761.64859009 pixy_tangent 2545.80651855 imagew 5208 imageh 3476 cd11 0.00017980392322 cd12 9.8135450351e-08 cd21 -9.8135450351e-08 cd22 0.00017980392322 det 3.23294604359e-08 parity 1 pixscale 0.647294220003 orientation 0.031271546 ra_center 78.631476441 dec_center -8.20041425873 orientation_center 0.0098274657 ra_center_h 5 ra_center_m 14 ra_center_s 31.5543458478 dec_center_sign -1 dec_center_d 8 dec_center_m 12 dec_center_s 1.49133141871 ra_center_hms 05:14:31.554 dec_center_dms -08:12:01.491 ra_center_merc 0.21842077 dec_center_merc 0.4771429 fieldarea 0.58526 fieldw 56.18 fieldh 37.5 fieldunits arcminutes decmin -8.51291 decmax -7.88757 ramin 78.158 ramax 79.1048 ra_min_merc 0.219736 ra_max_merc 0.217106 dec_min_merc 0.478021 dec_max_merc 0.476266 merc_diff 0.00262998 merczoom 9 Thanks, Andre Hi Rob, I just tested my solution with the solve-field.sh and the /Applications/Astronomia/KStars_3.3.4.app/Contents/MacOS/astrometry/bin/wcsinfo with KStars 3.3.3 and it worked perfectly. And, strangely, it also worked just fine after closing KStars 3.3.3 and opening KStars 3.3.4. Looks like it requires a fresh start after the configurations are changed. Thanks, Andre So it shouldn't matter if you are using macports or home-brew to install Python3. The only thing we are using homebrew for is python actually. As long as you have installed python3 somewhere on your computer and it is working, and you put a path into the kstars options that can help it find your installation if you are not using homebrew, it should work ok. In fact if macports is a very common way to install python, and it always installs in the same spot, maybe a combo box that lets you select that python would help? I am not sure what was causing the illegal argument exception before, but hopefully that is now solved based on Stanley's report?? I had told you to install astrometry.net in homebrew as a temporary solution, but hopefully that is not needed anymore. Perhaps another good addition might be to have a selection to allow the user to use the cloud maker's app to plate solve would be a good option to add as well? Your script could be used as a template to add this option to kstars. Hi Rob, The beauty of the scripting approach is that there's no need to change anything in Ekos. Just providing the script into the Contents directory of the app, and a documentation of the configuration process, is sufficient. Thanks, Andre Created attachment 124244 [details] attachment-1428-0.html Hi Rob, Hope this finds you well. I just downloaded kStars v3.3.8 and tried it out (using the simulators). When plate-solving using Astrometry.net <http://astrometry.net/> online, it works fine. But when I try doing it offline, it immediately fails. I’ve copied the session comments below (although not through Terminal as last time). I haven’t moved any of the Astrometry indexes, although I have installed Catalina on my Air. BTW this problem does not occur on my iMac, same as last time. Do you have any advice? Many thanks again, Stanley Fertig 2019-12-01T13:37:42 Solver failed. Try again. 2019-12-01T13:37:42 Error starting solver: Process crashed 2019-12-01T13:37:42 Reading input file 1 of 1: "/private/var/folders/vd/vwnkltm92sx93mrg14plg82r0000gn/T/fitsdfuyiD.fits"... 2019-12-01T13:37:41 /Applications/KStars.app/Contents/MacOS/astrometry/bin/solve-field -O --no-plots --no-verify --resort --downsample 2 -3 292.875 -4 28.0033 -5 30 --config /Applications/KStars.app/Contents/MacOS/astrometry/bin/astrometry.cfg -W /private/var/folders/vd/vwnkltm92sx93mrg14plg82r0000gn/T/solution.wcs /private/var/folders/vd/vwnkltm92sx93mrg14plg82r0000gn/T/fitsdfuyiD.fits 2019-12-01T13:37:41 Starting solver... 2019-12-01T13:37:41 export PATH=/Applications/KStars.app/Contents/MacOS/netpbm/bin:/usr/local/opt/python/libexec/bin:/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin 2019-12-01T13:37:41 Solver iteration #1 2019-12-01T13:37:41 Image received. 2019-12-01T13:37:38 Capturing image... -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com And on this device, you have python3 all set up in homebrew? Created attachment 124246 [details] attachment-2807-0.html Hi Rob, In answer to the python3 question below, when I click on the “Setup” button under Astrometry.net <http://astrometry.net/> panel in Ekos, I get the following message: “Home-brew, python and satrapy are already installed” Thanks, stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Dec 1, 2019, at 2:25 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #166 from Rob <rlancaste@gmail.com> --- > And on this device, you have python3 all set up in homebrew? > > -- > You are receiving this mail because: > You are on the CC list for the bug. Created attachment 124247 [details] attachment-3036-0.html Just checked. Python v. 3.7.3 is running on my Air (where astrometry does not work) whereas v. 2.7 is running on my iMac (where it does work). -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Dec 1, 2019, at 2:25 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #166 from Rob <rlancaste@gmail.com> --- > And on this device, you have python3 all set up in homebrew? > > -- > You are receiving this mail because: > You are on the CC list for the bug. Er...that's "satrapy" not "satrapy"--so much for dyslexia! (In reply to Stanley Fertig from comment #167) > Created attachment 124246 [details] > attachment-2807-0.html > > Hi Rob, > > In answer to the python3 question below, when I click on the “Setup” button > under Astrometry.net <http://astrometry.net/> panel in Ekos, I get the > following message: “Home-brew, python and satrapy are already installed” > > Thanks, > stan > > > -- > Stanley Fertig > 35 Sidney Place > Brooklyn, New York 11201 > +1 917-361-0632 > stanleyf@nyc.rr.com > > > > > > > > > > > > > On Dec 1, 2019, at 2:25 PM, Rob <bugzilla_noreply@kde.org> wrote: > > > > https://bugs.kde.org/show_bug.cgi?id=405437 > > > > --- Comment #166 from Rob <rlancaste@gmail.com> --- > > And on this device, you have python3 all set up in homebrew? > > > > -- > > You are receiving this mail because: > > You are on the CC list for the bug. Actually its "astropy" Maybe spell check changed it. But Did you try running the commands Ekos put in the log into the terminal to see what happens? Don't copy and paste it from here, copy and paste it from KStars into the terminal export PATH=/Applications/KStars.app/Contents/MacOS/netpbm/bin:/usr/local/opt/python/libexec/bin:/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin /Applications/KStars.app/Contents/MacOS/astrometry/bin/solve-field -O --no-plots --no-verify --resort --downsample 2 -3 292.875 -4 28.0033 -5 30 --config /Applications/KStars.app/Contents/MacOS/astrometry/bin/astrometry.cfg -W /private/var/folders/vd/vwnkltm92sx93mrg14plg82r0000gn/T/solution.wcs /private/var/folders/vd/vwnkltm92sx93mrg14plg82r0000gn/T/fitsdfuyiD.fits OK, I've been trying to get plate solving to work again on my macBook Air, using kStars v3.3.9. Becoming desperate. 1. I did another Erase and Install Mac OS Catalina v10.15.3 2. I installed HomeBrew and its Command Line tools, accepted everything. 3. Did not install ANYTHING else on my Air before the following. EXCEPT for the Astrometry index files, which I had backed up. 4. Did a fresh installation of kStars. Did the Startup Wizard, etc. 5. In the Solver Panel clicked on the Edit Solver Options icon. 6. Clicked on the Setup button for Python. 7. Detectd Homebew. When trying to install Python, got an error: "python install error" 8. Quit kStars, rebooted Mac. 9. Tried again. Now I get the following message: "Homebrew installed. Python3 installed. Astropy will install when you click OK." 10. When I click OK, I immediately get the following message: "Astropy install failure." 11. Trying "Capture and Solve" anyway just fails immediately, not a surprise if it needs Astropy. 12. Tried installing Astropy outside of kStars using Anaconda. Didn't change anything, still fails. I'm really at a loss. It all works fine on my iMac, just not on my Air. Although it previously worked on my Air under version 3.3.1 of kStars. Can you PLEASE help? I really don't want to go back to using Maxim... Thank you! stan in Brooklyn Yes I apologize, I have been very busy with my real job. Now that my midterm exams are over and I have entered my grades, I should be able to take a look again at this in the next few days. Sorry about the delay Big thanks, Rob!!!
I am grateful.
-stan
> On Feb 5, 2020, at 11:24 PM, Rob <bugzilla_noreply@kde.org> wrote:
>
> https://bugs.kde.org/show_bug.cgi?id=405437
>
> --- Comment #172 from Rob <rlancaste@gmail.com> ---
> Yes I apologize, I have been very busy with my real job. Now that my midterm
> exams are over and I have entered my grades, I should be able to take a look
> again at this in the next few days. Sorry about the delay
>
> --
> You are receiving this mail because:
> You are on the CC list for the bug.
I did some experiments last night with the issue you reported in a virtual machine with python3 installed in different ways to see why python3 sometimes is not working properly with astrometry.net to plate solve. Another user, who also installed non-homebrew python3 has reported a similar issue in this forum: https://indilib.org/forum/general/6362-kstars-mac-dmg-3-4-0-beta-testing-needed.html?start=72#48945 , and we were trying to diagnose it there too. In both cases, we did a number of experiments but couldn't find the problem. I decided to try a different approach. It was much easier for me to diagnose the issue in a virtual machine. I will post what I found last night in both places. The problem: The issue happens when the user has installed non-homebrew python. Python3 is installed properly, the site packages include astropy and numpy, it can find the site packages, the correct python3 is first in the PATH variable, KStars is installed correctly, and NOTE this is not the "illegal argument exception" caused by running software compiled for a newer computer on an older computer. But when the plate solve is run, it gives an error message which means it can't find astropy. When run from the command line, it gives the same error! My finding: It seems to me that the issue is that the calls to python astrometry.net are making use the name python not python3 when the calls are made. OS X has a python2.7 installed by default that is used by the system and should not be changed. Python3 is installed in various ways and could be in different locations. All the different python versions that I tested last night put a simlink in /usr/local/bin for python3 to redirect calls for python3 to wherever python3 is installed. There is not a simlink to python put in that location because the assumption is that if you call python, you want python2.7 to run the code and if you call python3, you want your self installed python to run it. I ran into this issue before with the homebrew python, and it gave me big headaches before, but later I found that there was a folder homebrew made called /usr/local/opt/python/libexec/bin where there actually was a simlink placed from python to the homebrew python3 and if I put that in the path, it automatically fixed the problem! I had assumed that the other versions of python would have a similar folder and you could just put them in the PATH variable and it would work. I was surprised last night when I didn't find a python simlink in /Library/Frameworks/Python.framework/Versions/3.7/bin (or similar folder for other installation), nor did I find anything in /Library/Frameworks/Python.framework/Versions/3.7/opt. So thus even with that folder in the PATH, it still called the system python whenever python was called for. Now that I think I know what the problem is, I should be able to come up with a way to fix it for non homebrew pythons. Just to verify, can you check for symlinks to python (not python3) in your /usr/local/bin folder, your /usr/local/opt/python/libexec/bin folder, and in whichever python3 bin and opt folder you would currently like to use (for example: /Library/Frameworks/Python.framework/Versions/3.7/bin and /Library/Frameworks/Python.framework/Versions/3.7/opt). If none of those folders contain a file or simlink called python, I think we have found the python3 issue. Created attachment 125793 [details] attachment-15595-0.html Hi Rob, Thanks so much for this. I’m attaching 4 screen shots of folders (labeled with numbers). The first is from /usr/local/bin The second is from /usr/local/opt/python/libexec/bin The 3rd is from /Library/Frameworks/Python.framework/Versions/3.7/bin The 4th is from /Library/Frameworks/Python.framework/Versions/3.7/opt I note I do see symlinks to python, python3, python3.7 and python3.8 in the first, a simlink to python in the second, no links but a python folder in the third (with a screenshot of the contents of that subfolder labeled 3.5), and when I choose “Go To Folder…” for the fourth link, I just get a sysbeep and I see the same folder as for the third link. Note: I have a Python3.8 folder in my Applications folder which contains the items is screenshot #5. Note 2: if I substitute “3.8” for “3.7” in the 3rd & 4th links, I get what appears to be the same folder as when I typed “3.7." Don’t know if any of that is helpful, please let me know what else you want me to test. Many thanks again for your help and time, Stan > On Feb 9, 2020, at 12:04 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #174 from Rob <rlancaste@gmail.com> --- > I did some experiments last night with the issue you reported in a virtual > machine with python3 installed in different ways to see why python3 sometimes > is not working properly with astrometry.net to plate solve. Another user, who > also installed non-homebrew python3 has reported a similar issue in this forum: > > https://indilib.org/forum/general/6362-kstars-mac-dmg-3-4-0-beta-testing-needed.html?start=72#48945 > > , and we were trying to diagnose it there too. In both cases, we did a number > of experiments but couldn't find the problem. I decided to try a different > approach. It was much easier for me to diagnose the issue in a virtual > machine. I will post what I found last night in both places. > > > > The problem: > The issue happens when the user has installed non-homebrew python. Python3 is > installed properly, the site packages include astropy and numpy, it can find > the site packages, the correct python3 is first in the PATH variable, KStars is > installed correctly, and NOTE this is not the "illegal argument exception" > caused by running software compiled for a newer computer on an older computer. > But when the plate solve is run, it gives an error message which means it can't > find astropy. When run from the command line, it gives the same error! > > My finding: > It seems to me that the issue is that the calls to python astrometry.net are > making use the name python not python3 when the calls are made. OS X has a > python2.7 installed by default that is used by the system and should not be > changed. Python3 is installed in various ways and could be in different > locations. All the different python versions that I tested last night put a > simlink in /usr/local/bin for python3 to redirect calls for python3 to wherever > python3 is installed. There is not a simlink to python put in that location > because the assumption is that if you call python, you want python2.7 to run > the code and if you call python3, you want your self installed python to run > it. I ran into this issue before with the homebrew python, and it gave me big > headaches before, but later I found that there was a folder homebrew made > called /usr/local/opt/python/libexec/bin where there actually was a simlink > placed from python to the homebrew python3 and if I put that in the path, it > automatically fixed the problem! I had assumed that the other versions of > python would have a similar folder and you could just put them in the PATH > variable and it would work. I was surprised last night when I didn't find a > python simlink in /Library/Frameworks/Python.framework/Versions/3.7/bin (or > similar folder for other installation), nor did I find anything in > /Library/Frameworks/Python.framework/Versions/3.7/opt. So thus even with that > folder in the PATH, it still called the system python whenever python was > called for. > > Now that I think I know what the problem is, I should be able to come up with a > way to fix it for non homebrew pythons. Just to verify, can you check for > symlinks to python (not python3) in your /usr/local/bin folder, your > /usr/local/opt/python/libexec/bin folder, and in whichever python3 bin and opt > folder you would currently like to use (for example: > /Library/Frameworks/Python.framework/Versions/3.7/bin and > /Library/Frameworks/Python.framework/Versions/3.7/opt). If none of those > folders contain a file or simlink called python, I think we have found the > python3 issue. > > -- > You are receiving this mail because: > You are on the CC list for the bug. Created attachment 125794 [details]
1Screen Shot 2020-02-09 at 1.52.29 PM.png
Created attachment 125795 [details]
2Screen Shot 2020-02-09 at 1.50.01 PM.png
Created attachment 125796 [details]
3Screen Shot 2020-02-09 at 1.55.01 PM.png
Created attachment 125797 [details]
3.5Screen Shot 2020-02-09 at 1.55.09 PM.png
Created attachment 125798 [details]
4Screen Shot 2020-02-09 at 2.01.57 PM.png
Created attachment 125799 [details]
5Screen Shot 2020-02-09 at 1.58.45 PM.png
I just had an idea today, that it might just be best to add an option to use sextractor with astrometry.net, for folks who can't get python to behave. If it works really well, then maybe we can just move away from python, since it seems there are so many different configurations out there. The good news is that I experimented with this before and still have my notes/code from that. So I think I will try it. Ok, after working really hard on this for many hours in my spare time in last 2 days, I think I have a solution for you: Try this: https://www.indilib.org/forum/general/6362-kstars-mac-dmg-3-4-0-beta-testing-needed.html?start=96#49352 See if this solves the problem please. Created attachment 125853 [details] attachment-21425-0.html OK. Just tried the new beta using the CCD, telescope and focuser simulators and the Pleiades as a target. Using the offline solver after configuring as you said, I immediately got a solver failure: “Starting Solver” then “Error starting solver: Process crashed” then “Sover failed. try again.” Tried that a few times, got the same error. Then I tried the online astrometry.net <http://astrometry.net/> solver (from the popup menu) and contrary to previous versions, it just hangs after uploading image file, does not solve. After a few minutes got a message: Upload failed.” Checked my internet connection and all is fine. ?? Thanks again, Rob, I’m sure we’ll get there... stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Feb 11, 2020, at 7:42 AM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #183 from Rob <rlancaste@gmail.com> --- > Ok, after working really hard on this for many hours in my spare time in last 2 > days, I think I have a solution for you: > > Try this: > https://www.indilib.org/forum/general/6362-kstars-mac-dmg-3-4-0-beta-testing-needed.html?start=96#49352 > > See if this solves the problem please. > > -- > You are receiving this mail because: > You are on the CC list for the bug. I assume that this is the "Illegal Instruction 4" error correct? I will try running this on an older Mac and try to determine which program is compiled for an newer version of OS X and thus won't run on older machines. Created attachment 125862 [details] attachment-26421-0.html I honestly don’t know if it’s the illegal instruction…can you please tell me again how to use Terminal to get more debug info? Thanks, stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Feb 11, 2020, at 1:34 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #185 from Rob <rlancaste@gmail.com> --- > I assume that this is the "Illegal Instruction 4" error correct? > > I will try running this on an older Mac and try to determine which program is > compiled for an newer version of OS X and thus won't run on older machines. > > -- > You are receiving this mail because: > You are on the CC list for the bug. I’m using a macbook Air under Catalina. Not so old...
-stan
Sent from my iPhone
> On Feb 11, 2020, at 1:34 PM, Rob <bugzilla_noreply@kde.org> wrote:
>
> https://bugs.kde.org/show_bug.cgi?id=405437
>
> --- Comment #185 from Rob <rlancaste@gmail.com> ---
> I assume that this is the "Illegal Instruction 4" error correct?
>
> I will try running this on an older Mac and try to determine which program is
> compiled for an newer version of OS X and thus won't run on older machines.
>
> --
> You are receiving this mail because:
> You are on the CC list for the bug.
That is the weird part then, if it is running Catalina, I don't know why it wouldn't be able to run the code. I think this shows it best: https://www.indilib.org/forum/general/6362-kstars-mac-dmg-3-4-0-beta-testing-needed.html?start=12#48752 Just note that the actual solve command will look a little different if you are using the sextractor xy list to solve. But it is still just copy and paste to the terminal. Created attachment 125870 [details] attachment-29217-0.html OK, I’ve configured it as in your screen shot. But copy WHAT to the Terminal? -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Feb 11, 2020, at 2:27 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #188 from Rob <rlancaste@gmail.com> --- > That is the weird part then, if it is running Catalina, I don't know why it > wouldn't be able to run the code. > > I think this shows it best: > https://www.indilib.org/forum/general/6362-kstars-mac-dmg-3-4-0-beta-testing-needed.html?start=12#48752 > > Just note that the actual solve command will look a little different if you are > using the sextractor xy list to solve. But it is still just copy and paste to > the terminal. > > -- > You are receiving this mail because: > You are on the CC list for the bug. In case you didn't see, there was a part after the first screen shot that says: Then plate solve an image in the simulator. You should see lots of things print out in the console. Copy and paste the line that says "export PATH= . . . " Hit enter. Also copy and paste the line that says "/Applications/KStars.app . . ." Then hit enter. Be sure you don't copy anything that is not commands, like dates, time stamps, etc. In this screen shot, I show you the exact commands that I copied and pasted. And how it looks in terminal when I do so. Just be careful you don't get extra stuff like "Starting solver" in there. Then there is a second screenshot showing an example. Created attachment 125887 [details] attachment-14537-0.html Hi Rob, I did plate solve, it failed and I copied what I think are the lines in the console at the bottom of the Align window. Not the results I was expecting…am I doing something wrong in the copy/paste? Last login: Tue Feb 11 23:16:00 on ttys000 sf@Stanleys-Air ~ % export PATH=/Applications/KStars 3.4.0 Beta.app/Contents/MacOS/netpbm/bin:/usr/local/opt/python/libexec/bin:/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin export: not an identifier: 3.4.0 sf@Stanleys-Air ~ % /Applications/KStars 3.4.0 Beta.app/Contents/MacOS/astrometry/bin/sex -CATALOG_NAME /private/var/folders/c6/k5dsm_n545d46b72wgyxqxc80000gn/T/SextractorList.xyls -CATALOG_TYPE FITS_1.0 /private/var/folders/c6/k5dsm_n545d46b72wgyxqxc80000gn/T/fitskpsDuK.fits zsh: no such file or directory: /Applications/KStars sf@Stanleys-Air ~ % I do have the "use Sextractor not python" box checked as well as the internal Sextractor box. But I see python and netpbm in the printout above... Thanks again, Stan PS—Because I had other kStars versions in my Applications folder with the name “kStars," I renamed the beta you sent me as it appears above. > On Feb 11, 2020, at 5:12 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #190 from Rob <rlancaste@gmail.com> --- > In case you didn't see, there was a part after the first screen shot that says: > > Then plate solve an image in the simulator. You should see lots of things print > out in the console. Copy and paste the line that says "export PATH= . . . " Hit > enter. Also copy and paste the line that says "/Applications/KStars.app . . ." > Then hit enter. Be sure you don't copy anything that is not commands, like > dates, time stamps, etc. In this screen shot, I show you the exact commands > that I copied and pasted. And how it looks in terminal when I do so. Just be > careful you don't get extra stuff like "Starting solver" in there. > > Then there is a second screenshot showing an example. > > -- > You are receiving this mail because: > You are on the CC list for the bug. There is a space in your PATH!! Make sure there are no spaces in your path to kstars: KStars 3.4.0 Beta.app Use underscores if you like, but never spaces, command line programs like astrometry choke on them. Created attachment 125888 [details] attachment-15742-0.html Aha! OK, I changed the name of the Application to remove the spaces. Now when I copy those 2 lines after plate solving I get the following: Last login: Tue Feb 11 23:21:10 on ttys000 sf@Stanleys-Air ~ % export PATH=/Applications/KStars3.4.0Beta.app/Contents/MacOS/netpbm/bin:/usr/local/opt/python/libexec/bin:/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin sf@Stanleys-Air ~ % /Applications/KStars3.4.0Beta.app/Contents/MacOS/astrometry/bin/sex -CATALOG_NAME /private/var/folders/c6/k5dsm_n545d46b72wgyxqxc80000gn/T/SextractorList.xyls -CATALOG_TYPE FITS_1.0 /private/var/folders/c6/k5dsm_n545d46b72wgyxqxc80000gn/T/fitsNhWxjI.fits > WARNING: default.sex not found, using internal defaults ----- SExtractor 2.19.5 started on 2020-02-11 at 23:44:26 with 1 thread > Setting catalog parameters > *ERROR*: can't read default.param sf@Stanleys-Air ~ % Hope that’s helpful… -stan > On Feb 11, 2020, at 11:31 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #192 from Rob <rlancaste@gmail.com> --- > There is a space in your PATH!! Make sure there are no spaces in your path to > kstars: > > KStars 3.4.0 Beta.app > > Use underscores if you like, but never spaces, command line programs like > astrometry choke on them. > > -- > You are receiving this mail because: > You are on the CC list for the bug. Sorry, there is one difference for using sextractor. There is an additional call for using sextractor. You should not need to copy that line to the terminal because I think it worked originally. You should copy the one to solve field like the one in my screenshot. Sorry if that wasn’t clear. Created attachment 125902 [details] attachment-6005-0.html Thanks Rob, OK. Don’t know if I got the right line, but I did get an illegal instruction when I did the following: Last login: Wed Feb 12 10:08:51 on ttys000 sf@Stanleys-Air ~ % export PATH=/Applications/KStars3.4.0Beta.app/Contents/MacOS/netpbm/bin:/usr/local/opt/python/libexec/bin:/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin sf@Stanleys-Air ~ % /Applications/KStars3.4.0Beta.app/Contents/MacOS/astrometry/bin/solve-field -O --no-plots --no-verify --resort --downsample 2 --width 320 --height 256 --x-column X_IMAGE --y-column Y_IMAGE --sort-column MAG_AUTO --sort-ascending --no-remove-lines --uniformize 0 -3 10.9511 -4 41.3797 -5 30 --config /Applications/KStars3.4.0Beta.app/Contents/MacOS/astrometry/bin/astrometry.cfg -W /private/var/folders/c6/k5dsm_n545d46b72wgyxqxc80000gn/T/solution.wcs /private/var/folders/c6/k5dsm_n545d46b72wgyxqxc80000gn/T/SextractorList.xyls Reading input file 1 of 1: "/private/var/folders/c6/k5dsm_n545d46b72wgyxqxc80000gn/T/SextractorList.xyls"... zsh: illegal hardware instruction /Applications/KStars3.4.0Beta.app/Contents/MacOS/astrometry/bin/solve-field - sf@Stanleys-Air ~ % I’m also copying here the entire output from kStars: 2020-02-12T10:17:09 Solver failed. Try again. 2020-02-12T10:17:09 Error starting solver: Process crashed 2020-02-12T10:17:09 Reading input file 1 of 1: "/private/var/folders/c6/k5dsm_n545d46b72wgyxqxc80000gn/T/SextractorList.xyls"... 2020-02-12T10:17:08 /Applications/KStars3.4.0Beta.app/Contents/MacOS/astrometry/bin/solve-field -O --no-plots --no-verify --resort --downsample 2 --width 320 --height 256 --x-column X_IMAGE --y-column Y_IMAGE --sort-column MAG_AUTO --sort-ascending --no-remove-lines --uniformize 0 -3 10.9511 -4 41.3797 -5 30 --config /Applications/KStars3.4.0Beta.app/Contents/MacOS/astrometry/bin/astrometry.cfg -W /private/var/folders/c6/k5dsm_n545d46b72wgyxqxc80000gn/T/solution.wcs /private/var/folders/c6/k5dsm_n545d46b72wgyxqxc80000gn/T/SextractorList.xyls 2020-02-12T10:17:08 Starting solver... 2020-02-12T10:17:08 > WARNING: default.sex not found, using internal defaults [1M> [1A----- SExtractor 2.19.5 started on 2020-02-12 at 10:17:09 with 1 thread [1M> Setting catalog parameters [1A[1M> Reading detection filter [1A[1M> Initializing catalog [1A[1M> Looking for fitsCdnvQe.fits [1A----- Measuring from: fitsCdnvQe.fits "M_31" / no ext. header / 320x256 / 16 bits (integers) Detection+Measurement image: [1M> Setting up background maps [1A[1M> Filtering background map(s) [1A[1M> Computing background d-map [1A[1M> Computing background-noise d-map [1A(M+D) Background: 231.941 RMS: 11.4251 / Threshold: 17.1377 [1M> Scanning image [1A[1M> Line: 25 Objects: 1 detected / 0 sextracted [1A[1M> Line: 50 Objects: 2 detected / 0 sextracted [1A[1M> Line: 75 Objects: 6 detected / 0 sextracted [1A[1M> Line: 100 Objects: 8 detected / 0 sextracted [1A[1M> Line: 125 Objects: 11 detected / 0 sextracted [1A[1M> Line: 150 Objects: 13 detected / 0 sextracted [1A[1M> Line: 175 Objects: 16 detected / 0 sextracted [1A[1M> Line: 200 Objects: 17 detected / 0 sextracted [1A[1M> Line: 225 Objects: 22 detected / 0 sextracted [1A[1M> Line: 250 Objects: 25 detected / 0 sextracted [1A[1M> Line: 256 Objects: 27 detected / 0 sextracted [1A Objects: detected 27 / sextracted 27 [1M> Closing files [1A[1M> [1A> All done (in 0.0 s: 13398.3 lines/s , 1413.1 detections/s) 2020-02-12T10:17:08 /Applications/KStars3.4.0Beta.app/Contents/MacOS/astrometry/bin/sex -CATALOG_NAME /private/var/folders/c6/k5dsm_n545d46b72wgyxqxc80000gn/T/SextractorList.xyls -CATALOG_TYPE FITS_1.0 /private/var/folders/c6/k5dsm_n545d46b72wgyxqxc80000gn/T/fitsCdnvQe.fits 2020-02-12T10:17:08 Starting sextractor... 2020-02-12T10:17:08 export PATH=/Applications/KStars3.4.0Beta.app/Contents/MacOS/netpbm/bin:/usr/local/opt/python/libexec/bin:/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin 2020-02-12T10:17:08 Solver iteration #1 2020-02-12T10:17:08 Image received. 2020-02-12T10:17:07 Capturing image... 2020-02-12T10:09:26 World Coordinate System (WCS) is enabled. CCD rotation must be set either manually in the CCD driver or by solving an image before proceeding to capture any further images, otherwise the WCS information may be invalid. 2020-02-12T10:09:26 Detected Astrometry.net version 0.76 2020-02-12T10:09:26 Idle. Hope that helps. Many thanks again, Stan > On Feb 12, 2020, at 1:47 AM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #194 from Rob <rlancaste@gmail.com> --- > Sorry, there is one difference for using sextractor. There is an additional > call for using sextractor. You should not need to copy that line to the > terminal because I think it worked originally. You should copy the one to > solve field like the one in my screenshot. Sorry if that wasn’t clear. > > -- > You are receiving this mail because: > You are on the CC list for the bug. Perfect, yes, that is exactly what you needed to do. Looks like it is the same error. What I don't understand is why can't a MacBook Air run code that a MacBook Pro can run??? What processor does your MacBook Air have? Created attachment 125923 [details] attachment-26286-0.html Hi Rob, I’ve never heard of anything like that chez Apple. In any case, my Air’s processor is a 1.8GHz Dual-Core Intel Core I5, 4 GB of RAM and Intel HD Graphics 4000 1536MB, from “About this Mac.” What’s also strange is that this all worked at some point in the past. Maybe I should try to downgrade my OS to El Capitan or one of the Sierras… -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Feb 12, 2020, at 2:46 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #196 from Rob <rlancaste@gmail.com> --- > Perfect, yes, that is exactly what you needed to do. Looks like it is the same > error. What I don't understand is why can't a MacBook Air run code that a > MacBook Pro can run??? What processor does your MacBook Air have? > > -- > You are receiving this mail because: > You are on the CC list for the bug. Nothing sounds weird about that. Mine is an Intel i9, but that should not matter if yours is an intel i5 I don't think. Created attachment 125924 [details] attachment-26779-0.html And it runs fine on my iMac, which is a 2.7 GHz Quad-Core Intel Core i5… -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Feb 12, 2020, at 9:11 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #198 from Rob <rlancaste@gmail.com> --- > Nothing sounds weird about that. Mine is an Intel i9, but that should not > matter if yours is an intel i5 I don't think. > > -- > You are receiving this mail because: > You are on the CC list for the bug. So your statement that it worked in the past . . .actually it worked for you when I built it on an ancient 10 year old computer running El Capitan. So then your computer was a newer one running a newer operating system. Typically that works well. The challenge happens when you try to run newer stuff on older machines. I expect the error you report if you are running an older operating system than I built it on, but you are running this on a new OS. I can no longer build software on that older computer for this since craft uses the newest version of QT and QT decided that they won't support any operating system older than 10.12 Sierra. That old computer cannot run Sierra. And this is the weirdest part I think, the iMac is also an i5! (In reply to Stanley Fertig from comment #199) > Created attachment 125924 [details] > attachment-26779-0.html > > And it runs fine on my iMac, which is a 2.7 GHz Quad-Core Intel Core i5… > > -stan > -- > Stanley Fertig > 35 Sidney Place > Brooklyn, New York 11201 > +1 917-361-0632 > stanleyf@nyc.rr.com > > > > > > > > > > > > > On Feb 12, 2020, at 9:11 PM, Rob <bugzilla_noreply@kde.org> wrote: > > > > https://bugs.kde.org/show_bug.cgi?id=405437 > > > > --- Comment #198 from Rob <rlancaste@gmail.com> --- > > Nothing sounds weird about that. Mine is an Intel i9, but that should not > > matter if yours is an intel i5 I don't think. > > > > -- > > You are receiving this mail because: > > You are on the CC list for the bug. Created attachment 125925 [details] attachment-27157-0.html Rob I mean it worked one year ago on my Air! -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Feb 12, 2020, at 9:19 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #201 from Rob <rlancaste@gmail.com> --- > And this is the weirdest part I think, the iMac is also an i5! > > (In reply to Stanley Fertig from comment #199) >> Created attachment 125924 [details] >> attachment-26779-0.html >> >> And it runs fine on my iMac, which is a 2.7 GHz Quad-Core Intel Core i5… >> >> -stan >> -- >> Stanley Fertig >> 35 Sidney Place >> Brooklyn, New York 11201 >> +1 917-361-0632 >> stanleyf@nyc.rr.com >> >> >> >> >> >> >> >> >> >> >> >>> On Feb 12, 2020, at 9:11 PM, Rob <bugzilla_noreply@kde.org> wrote: >>> >>> https://bugs.kde.org/show_bug.cgi?id=405437 >>> >>> --- Comment #198 from Rob <rlancaste@gmail.com> --- >>> Nothing sounds weird about that. Mine is an Intel i9, but that should not >>> matter if yours is an intel i5 I don't think. >>> >>> -- >>> You are receiving this mail because: >>> You are on the CC list for the bug. > > -- > You are receiving this mail because: > You are on the CC list for the bug. Right, it worked a year ago on your Air with software that I built on a really old computer, not the software I build now on a newer computer with the latest operating system. It was in KStars 3.1.0 that I started building on the new computer. So maybe I should revert to kStars 2.9.8? Cheaper than getting a new computer. I’d miss out on all the improvements you’ve made since but I could still do photography...
-stan
Sent from my iPhone
> On Feb 12, 2020, at 9:28 PM, Rob <bugzilla_noreply@kde.org> wrote:
>
> https://bugs.kde.org/show_bug.cgi?id=405437
>
> --- Comment #204 from Rob <rlancaste@gmail.com> ---
> It was in KStars 3.1.0 that I started building on the new computer.
>
> --
> You are receiving this mail because:
> You are on the CC list for the bug.
As an alternative, have you tried doing this: https://indilib.org/forum/ekos/5259-test-on-kstars-3-2-3-on-mac-missing-astrometry-net-binaries.html?start=24#40216 Created attachment 125926 [details] attachment-27941-0.html Did you mean in Terminal? I tried and got the following: > Last login: Wed Feb 12 22:00:05 on console > sf@Stanleys-Air ~ % brew install astrometry.net > Updating Homebrew... > ==> Auto-updated Homebrew! > Updated 1 tap (homebrew/core). > ==> New Formulae > cacli gh libdeflate wasm3 > cctz kcptun pueue > ==> Updated Formulae > abcmidi gdb openjdk@11 > ace gimme-aws-creds openjdk@12 > activemq git-delta paket > akka glooctl pdfcpu > allure glslviewer petsc > ammonite-repl golo petsc-complex > angular-cli gradle phpunit > ansible grafana picard-tools > ant gsasl pipx > antlr hapi-fhir-cli procs > antlr@2 helmfile profanity > apache-archiva hlint proteinortho > apache-arrow homebank pulumi > apache-arrow-glib http-parser pumba > apache-ctakes hugo pyinvoke > apache-forrest i386-elf-gdb pympress > apache-spark imagemagick python-markdown > apollo imagemagick@6 qxmpp > armadillo ioping rav1e > artifactory jenkins repo > asymptote jhipster rex > aws-cdk jsonnet riff > aws-sdk-cpp just run > ballerina kops scons > bash kubecfg screen > bazel lazygit serverless > bazelisk legit sfk > beast libcdr sile > benthos libgee simgrid > bit libgpg-error skopeo > broot libmtp snakemake > buildifier libpsl source-to-image > caf linkerd sphinx-doc > cfn-lint liquibase spotbugs > checkbashisms lldpd ssh-vault > citus manticoresearch starship > clamav maven stress-ng > closure-compiler maven@3.2 subnetcalc > cmake maven@3.3 subversion > cobalt maven@3.5 swiftlint > conan micro sysdig > contentful-cli minikube telegraf > dartsim minio terraformer > dateutils mk-configure terragrunt > dive movgrab texlab > dvc nano tile38 > dynare nats-streaming-server tippecanoe > elixir navi traefik > embree ncdu ttyd > erlang nesc v8 > exploitdb netlify-cli vala > eye-d3 ngt verilator > faiss nng wireguard-tools > fcl nnn wtf > findomain node wtfutil > fio node@10 xcodegen > fits node@12 xtensor > flow now-cli yadm > fluxctl nss yarn > flyway ocp yq > folly octave zeek > fuseki opa > gdal openjdk > ==> Deleted Formulae > python@2 > > Error: No available formula with the name "astrometry.net" > ==> Searching for a previously deleted formula (in the last month)... > Warning: homebrew/core is shallow clone. To get complete history run: > git -C "$(brew --repo homebrew/core)" fetch --unshallow > > Error: No previously deleted formula found. > ==> Searching for similarly named formulae... > This similarly named formula was found: > astrometry-net > To install it, run: > brew install astrometry-net > ==> Searching taps... > ==> Searching taps on GitHub... > Error: No formulae found in taps. > sf@Stanleys-Air ~ % > On Feb 12, 2020, at 9:42 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #206 from Rob <rlancaste@gmail.com> --- > As an alternative, have you tried doing this: > > https://indilib.org/forum/ekos/5259-test-on-kstars-3-2-3-on-mac-missing-astrometry-net-binaries.html?start=24#40216 > > -- > You are receiving this mail because: > You are on the CC list for the bug. Created attachment 125927 [details] attachment-28009-0.html Sorry, my mistake, it is brew install astrometry-net > On Feb 12, 2020, at 10:05 PM, Stanley Fertig <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #207 from Stanley Fertig <stanleyf@nyc.rr.com> --- > Did you mean in Terminal? I tried and got the following: > >> Last login: Wed Feb 12 22:00:05 on console >> sf@Stanleys-Air ~ % brew install astrometry.net >> Updating Homebrew... >> ==> Auto-updated Homebrew! >> Updated 1 tap (homebrew/core). >> ==> New Formulae >> cacli gh libdeflate wasm3 >> cctz kcptun pueue >> ==> Updated Formulae >> abcmidi gdb openjdk@11 >> ace gimme-aws-creds openjdk@12 >> activemq git-delta paket >> akka glooctl pdfcpu >> allure glslviewer petsc >> ammonite-repl golo petsc-complex >> angular-cli gradle phpunit >> ansible grafana picard-tools >> ant gsasl pipx >> antlr hapi-fhir-cli procs >> antlr@2 helmfile profanity >> apache-archiva hlint proteinortho >> apache-arrow homebank pulumi >> apache-arrow-glib http-parser pumba >> apache-ctakes hugo pyinvoke >> apache-forrest i386-elf-gdb pympress >> apache-spark imagemagick python-markdown >> apollo imagemagick@6 qxmpp >> armadillo ioping rav1e >> artifactory jenkins repo >> asymptote jhipster rex >> aws-cdk jsonnet riff >> aws-sdk-cpp just run >> ballerina kops scons >> bash kubecfg screen >> bazel lazygit serverless >> bazelisk legit sfk >> beast libcdr sile >> benthos libgee simgrid >> bit libgpg-error skopeo >> broot libmtp snakemake >> buildifier libpsl source-to-image >> caf linkerd sphinx-doc >> cfn-lint liquibase spotbugs >> checkbashisms lldpd ssh-vault >> citus manticoresearch starship >> clamav maven stress-ng >> closure-compiler maven@3.2 subnetcalc >> cmake maven@3.3 subversion >> cobalt maven@3.5 swiftlint >> conan micro sysdig >> contentful-cli minikube telegraf >> dartsim minio terraformer >> dateutils mk-configure terragrunt >> dive movgrab texlab >> dvc nano tile38 >> dynare nats-streaming-server tippecanoe >> elixir navi traefik >> embree ncdu ttyd >> erlang nesc v8 >> exploitdb netlify-cli vala >> eye-d3 ngt verilator >> faiss nng wireguard-tools >> fcl nnn wtf >> findomain node wtfutil >> fio node@10 xcodegen >> fits node@12 xtensor >> flow now-cli yadm >> fluxctl nss yarn >> flyway ocp yq >> folly octave zeek >> fuseki opa >> gdal openjdk >> ==> Deleted Formulae >> python@2 >> >> Error: No available formula with the name "astrometry.net" >> ==> Searching for a previously deleted formula (in the last month)... >> Warning: homebrew/core is shallow clone. To get complete history run: >> git -C "$(brew --repo homebrew/core)" fetch --unshallow >> >> Error: No previously deleted formula found. >> ==> Searching for similarly named formulae... >> This similarly named formula was found: >> astrometry-net >> To install it, run: >> brew install astrometry-net >> ==> Searching taps... >> ==> Searching taps on GitHub... >> Error: No formulae found in taps. >> sf@Stanleys-Air ~ % > >> On Feb 12, 2020, at 9:42 PM, Rob <bugzilla_noreply@kde.org> wrote: >> >> https://bugs.kde.org/show_bug.cgi?id=405437 >> >> --- Comment #206 from Rob <rlancaste@gmail.com> --- >> As an alternative, have you tried doing this: >> >> https://indilib.org/forum/ekos/5259-test-on-kstars-3-2-3-on-mac-missing-astrometry-net-binaries.html?start=24#40216 >> >> -- >> You are receiving this mail because: >> You are on the CC list for the bug. > > -- > You are receiving this mail because: > You are the assignee for the bug. Wow! That worked. Plate solving worked after doing that installation and configuring as in your screenshot!
Big enormous thank you, Rob. But what does it all mean?
Thanks again, you are such a dedicated guy!
-stan
> On Feb 12, 2020, at 10:07 PM, Rob <bugzilla_noreply@kde.org> wrote:
>
> https://bugs.kde.org/show_bug.cgi?id=405437
>
> --- Comment #208 from Rob <rlancaste@gmail.com> ---
> Sorry, my mistake, it is brew install astrometry-net
>
>> On Feb 12, 2020, at 10:05 PM, Stanley Fertig <bugzilla_noreply@kde.org> wrote:
>>
>> https://bugs.kde.org/show_bug.cgi?id=405437
>>
>> --- Comment #207 from Stanley Fertig <stanleyf@nyc.rr.com> ---
>> Did you mean in Terminal? I tried and got the following:
>>
>>> Last login: Wed Feb 12 22:00:05 on console
>>> sf@Stanleys-Air ~ % brew install astrometry.net
>>> Updating Homebrew...
>>> ==> Auto-updated Homebrew!
>>> Updated 1 tap (homebrew/core).
>>> ==> New Formulae
>>> cacli gh libdeflate wasm3
>>> cctz kcptun pueue
>>> ==> Updated Formulae
>>> abcmidi gdb openjdk@11
>>> ace gimme-aws-creds openjdk@12
>>> activemq git-delta paket
>>> akka glooctl pdfcpu
>>> allure glslviewer petsc
>>> ammonite-repl golo petsc-complex
>>> angular-cli gradle phpunit
>>> ansible grafana picard-tools
>>> ant gsasl pipx
>>> antlr hapi-fhir-cli procs
>>> antlr@2 helmfile profanity
>>> apache-archiva hlint proteinortho
>>> apache-arrow homebank pulumi
>>> apache-arrow-glib http-parser pumba
>>> apache-ctakes hugo pyinvoke
>>> apache-forrest i386-elf-gdb pympress
>>> apache-spark imagemagick python-markdown
>>> apollo imagemagick@6 qxmpp
>>> armadillo ioping rav1e
>>> artifactory jenkins repo
>>> asymptote jhipster rex
>>> aws-cdk jsonnet riff
>>> aws-sdk-cpp just run
>>> ballerina kops scons
>>> bash kubecfg screen
>>> bazel lazygit serverless
>>> bazelisk legit sfk
>>> beast libcdr sile
>>> benthos libgee simgrid
>>> bit libgpg-error skopeo
>>> broot libmtp snakemake
>>> buildifier libpsl source-to-image
>>> caf linkerd sphinx-doc
>>> cfn-lint liquibase spotbugs
>>> checkbashisms lldpd ssh-vault
>>> citus manticoresearch starship
>>> clamav maven stress-ng
>>> closure-compiler maven@3.2 subnetcalc
>>> cmake maven@3.3 subversion
>>> cobalt maven@3.5 swiftlint
>>> conan micro sysdig
>>> contentful-cli minikube telegraf
>>> dartsim minio terraformer
>>> dateutils mk-configure terragrunt
>>> dive movgrab texlab
>>> dvc nano tile38
>>> dynare nats-streaming-server tippecanoe
>>> elixir navi traefik
>>> embree ncdu ttyd
>>> erlang nesc v8
>>> exploitdb netlify-cli vala
>>> eye-d3 ngt verilator
>>> faiss nng wireguard-tools
>>> fcl nnn wtf
>>> findomain node wtfutil
>>> fio node@10 xcodegen
>>> fits node@12 xtensor
>>> flow now-cli yadm
>>> fluxctl nss yarn
>>> flyway ocp yq
>>> folly octave zeek
>>> fuseki opa
>>> gdal openjdk
>>> ==> Deleted Formulae
>>> python@2
>>>
>>> Error: No available formula with the name "astrometry.net"
>>> ==> Searching for a previously deleted formula (in the last month)...
>>> Warning: homebrew/core is shallow clone. To get complete history run:
>>> git -C "$(brew --repo homebrew/core)" fetch --unshallow
>>>
>>> Error: No previously deleted formula found.
>>> ==> Searching for similarly named formulae...
>>> This similarly named formula was found:
>>> astrometry-net
>>> To install it, run:
>>> brew install astrometry-net
>>> ==> Searching taps...
>>> ==> Searching taps on GitHub...
>>> Error: No formulae found in taps.
>>> sf@Stanleys-Air ~ %
>>
>>> On Feb 12, 2020, at 9:42 PM, Rob <bugzilla_noreply@kde.org> wrote:
>>>
>>> https://bugs.kde.org/show_bug.cgi?id=405437
>>>
>>> --- Comment #206 from Rob <rlancaste@gmail.com> ---
>>> As an alternative, have you tried doing this:
>>>
>>> https://indilib.org/forum/ekos/5259-test-on-kstars-3-2-3-on-mac-missing-astrometry-net-binaries.html?start=24#40216
>>>
>>> --
>>> You are receiving this mail because:
>>> You are on the CC list for the bug.
>>
>> --
>> You are receiving this mail because:
>> You are the assignee for the bug.
>
> --
> You are receiving this mail because:
> You are on the CC list for the bug.
Sorry, I should have told you to do that sooner. But I do want to get to the bottom of this silly strange issue with the code not running on a MacBook Air, I don't get that. Maybe come compiler flag needs to be set to not use a certain instruction set or something, but I don't understand it if it works on your other i5. . . Neither do I but if there’s any test you want me to do which will help you, consider me your loyal assistant.
I did this on the beta you sent me. Am going to try it now on the official v3.3.1. Will let you know how it goes…
-stan
> On Feb 12, 2020, at 10:31 PM, Rob <bugzilla_noreply@kde.org> wrote:
>
> https://bugs.kde.org/show_bug.cgi?id=405437
>
> --- Comment #210 from Rob <rlancaste@gmail.com> ---
> Sorry, I should have told you to do that sooner. But I do want to get to the
> bottom of this silly strange issue with the code not running on a MacBook Air,
> I don't get that. Maybe come compiler flag needs to be set to not use a
> certain instruction set or something, but I don't understand it if it works on
> your other i5. . .
>
> --
> You are receiving this mail because:
> You are on the CC list for the bug.
Just tried it on v3.3.1. Worked like a charm and solved in about 2 seconds.
Thank you again, Rob!
-stan
> On Feb 12, 2020, at 10:31 PM, Rob <bugzilla_noreply@kde.org> wrote:
>
> https://bugs.kde.org/show_bug.cgi?id=405437
>
> --- Comment #210 from Rob <rlancaste@gmail.com> ---
> Sorry, I should have told you to do that sooner. But I do want to get to the
> bottom of this silly strange issue with the code not running on a MacBook Air,
> I don't get that. Maybe come compiler flag needs to be set to not use a
> certain instruction set or something, but I don't understand it if it works on
> your other i5. . .
>
> --
> You are receiving this mail because:
> You are on the CC list for the bug.
Created attachment 125928 [details] attachment-28559-0.html And I just tried in v3.3.9. Worked like a charm. -stan -- Stanley Fertig 35 Sidney Place Brooklyn, New York 11201 +1 917-361-0632 stanleyf@nyc.rr.com > On Feb 12, 2020, at 10:31 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #210 from Rob <rlancaste@gmail.com> --- > Sorry, I should have told you to do that sooner. But I do want to get to the > bottom of this silly strange issue with the code not running on a MacBook Air, > I don't get that. Maybe come compiler flag needs to be set to not use a > certain instruction set or something, but I don't understand it if it works on > your other i5. . . > > -- > You are receiving this mail because: > You are on the CC list for the bug. Hi Rob, I hope you are surviving the pandemic. My dreaded astrometry plate-solving problem has returned—Astrometry crashes immediately. FYI I’ve moved and changed email address, please note that I’m now at stratocaster@icloud.com. I’ve established a new account with kde bug tracking with that address. I’ve re-installed everything to no avail. My setup is the following: -2014 MacBook Air -Mohave OS (10.14.6)—had to downgrade because SBIG has problems in Catalina. -I have Python 3.8.5 installed (using the Python installer, not through kStars) -kStars v3.4.3 -I followed your last Astrometry install procedure in Terminal from our last go-round (brew install astronomy-net) -So after tracking an object in kStars (simulators) I get the following message despite having already installed home-brew & Python 3 and the above: "Astrometry.net uses python3 and the astropy package for plate solving images offline. These were not detected on your system. Please go into the Align Options and either click the setup button to install them or uncheck the default button and enter the path to python3 on your system and manually install astropy.” -OK. So when I click on the “Setup” button in the Astrometry.net settings in kStars, I get 3 successive messages: 1. This installer will install the following requirements for astrometry.net if they are not installed: Homebrew -an OS X Unix Program Package Manager Python3 -A Powerful Scripting Language Astropy -Python Modules for Astronomy Do you wish to continue?" 2. And after I click on “Yes” I get this second message: "Homebrew installed Python3 installed Astropy will install when you click Ok (Note: this might take a few minutes, please be patient.)” 3. And after clicking on “OK” I get this: “Astropy Install Failure” And it will not solve. If I repeat the above exercise, I get the exact same results. I’m really sorry to bother you again. Also downloaded and tried ASTAP but that wouldn’t solve either. Any advice? Many thanks, Stan — Stanley Fertig 20 Henry Street, PH3S Brooklyn, NY 11201 +1 917 361-0632 stratocaster@icloud.com On Feb 12, 2020, at 10:31 PM, Rob <bugzilla_noreply@kde.org> wrote: https://bugs.kde.org/show_bug.cgi?id=405437 --- Comment #210 from Rob <rlancaste@gmail.com> --- Sorry, I should have told you to do that sooner. But I do want to get to the bottom of this silly strange issue with the code not running on a MacBook Air, I don't get that. Maybe come compiler flag needs to be set to not use a certain instruction set or something, but I don't understand it if it works on your other i5. . . -- You are receiving this mail because: You are on the CC list for the bug. Hi Stanley, I was going through all the old bug reports. I think this bug should no longer be present now because I have now updated KStars to include stellarsolver that I built during the pandemic in 2020. It should solve all these issues because it doesn't have all the dependencies that astrometry.net has. So it should work out of the box with no issues. Created attachment 145918 [details] attachment-6803-0.html Hi Rob, I’m just testing it out now during the day with kStars 3.5.7. I’m getting an error: “[ERROR] ReadoutLine: CC_READ_SUBTRACT_Line./cc_READOUT_LINE -> (RX /Timeout) Is that an error of my camera or its USB cord or of kStars? Thanks, stan — Stanley Fertig 20 Henry Street, PH3S Brooklyn, NY 11201 +1 917 361-0632 stratocaster@icloud.com > On Jan 22, 2022, at 11:19 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > Rob <rlancaste@gmail.com> changed: > > What |Removed |Added > ---------------------------------------------------------------------------- > Resolution|--- |FIXED > Status|REPORTED |RESOLVED > > --- Comment #215 from Rob <rlancaste@gmail.com> --- > Hi Stanley, I was going through all the old bug reports. I think this bug > should no longer be present now because I have now updated KStars to include > stellarsolver that I built during the pandemic in 2020. It should solve all > these issues because it doesn't have all the dependencies that astrometry.net > has. So it should work out of the box with no issues. > > -- > You are receiving this mail because: > You are on the CC list for the bug. Created attachment 145922 [details] attachment-9765-0.html Hmm, just tried it out with a “camera only” profile (no telescope, no focuser, etc.) and the camera appears to be working fine. Could that error message have been generated by my focuser (Moonlite)? -stan — Stanley Fertig 20 Henry Street, PH3S Brooklyn, NY 11201 +1 917 361-0632 stratocaster@icloud.com > On Jan 25, 2022, at 10:26 AM, Stanley Fertig <stratocaster@icloud.com> wrote: > > Hi Rob, > > I’m just testing it out now during the day with kStars 3.5.7. I’m getting an error: “[ERROR] ReadoutLine: CC_READ_SUBTRACT_Line./cc_READOUT_LINE -> (RX /Timeout) > > Is that an error of my camera or its USB cord or of kStars? > > Thanks, > stan > > > > > — > Stanley Fertig > 20 Henry Street, PH3S > Brooklyn, NY 11201 > +1 917 361-0632 > stratocaster@icloud.com <mailto:stratocaster@icloud.com> > > > > >> On Jan 22, 2022, at 11:19 PM, Rob <bugzilla_noreply@kde.org <mailto:bugzilla_noreply@kde.org>> wrote: >> >> https://bugs.kde.org/show_bug.cgi?id=405437 <https://bugs.kde.org/show_bug.cgi?id=405437> >> >> Rob <rlancaste@gmail.com> changed: >> >> What |Removed |Added >> ---------------------------------------------------------------------------- >> Resolution|--- |FIXED >> Status|REPORTED |RESOLVED >> >> --- Comment #215 from Rob <rlancaste@gmail.com> --- >> Hi Stanley, I was going through all the old bug reports. I think this bug >> should no longer be present now because I have now updated KStars to include >> stellarsolver that I built during the pandemic in 2020. It should solve all >> these issues because it doesn't have all the dependencies that astrometry.net >> has. So it should work out of the box with no issues. >> >> -- >> You are receiving this mail because: >> You are on the CC list for the bug. > So I am not sure what would cause this to print out. I did a google search for "CC_READ_SUBTRACT_LINE" and found that it was probably associated with the SBIG driver. You might want to post in the INDI forum and see if anybody knows. OK Rob, thanks. You are the greatest!
-stan
> On Jan 25, 2022, at 2:48 PM, Rob <bugzilla_noreply@kde.org> wrote:
>
> https://bugs.kde.org/show_bug.cgi?id=405437
>
> --- Comment #218 from Rob <rlancaste@gmail.com> ---
> So I am not sure what would cause this to print out. I did a google search for
> "CC_READ_SUBTRACT_LINE" and found that it was probably associated with the SBIG
> driver. You might want to post in the INDI forum and see if anybody knows.
>
> --
> You are receiving this mail because:
> You are on the CC list for the bug.
Hello Rob, Long time no speak! I hope this finds you well. I’m the guy who went back and forth with you about kStars v3.? not platesolving correctly using my old MacBook Air. Problem which you solved. I’m not sure how to request a new feature in kStars so I’m writing to you. Please feel free to forward as appropriate if it's not you. I recently bought an AG Optical iDK scope which is wonderful. (Agoptical.com). They are similar to Planewaves. They come with a temperature control system featuring sensors on the main and secondary mirrors as well as ambient, and cooling fans as well as dew heaters on the mirrors. Unfortunately for me, the software package to control all this, (“TCS”) is Windows-only and the developer sees no point in open-sourcing this or doing any additional work on it, thus no porting to Mac. (Discussion with him her: https://www.cloudynights.com/topic/868614-ag-optical-thermal-control-system-app-no-longer-supported/) I’m not wild about having to run a second computer (Windows) just for the thermal control of the scope. I see that kStars already supports Planewave’s Delta-T software, so maybe it wouldn’t be so much of a stretch to support AGO TCS as well. So that’s my feature request: add AGO TCS as an “AUX” option in Ekos. Do you think this is possible or likely? Many thanks again, Stan > On Feb 12, 2020, at 10:31 PM, Rob <bugzilla_noreply@kde.org> wrote: > > https://bugs.kde.org/show_bug.cgi?id=405437 > > --- Comment #210 from Rob <rlancaste@gmail.com> --- > Sorry, I should have told you to do that sooner. But I do want to get to the > bottom of this silly strange issue with the code not running on a MacBook Air, > I don't get that. Maybe come compiler flag needs to be set to not use a > certain instruction set or something, but I don't understand it if it works on > your other i5. . . > > -- > You are receiving this mail because: > You are on the CC list for the bug. |