GitLab

I was doing more testing, and the range on the sliders needs to be expanded just slightly so that it includes all the data.

In preparation for the OS X release of KStars 2.9.0, I hunted around for any bugs I could find and I made several small changes.

1. Set the size of the fitsviewer AFTER showing it.  This fixes a bug that showed up in QT 5.10
2. Widened the fitsviewer just a little to show all buttons.
3. Added a cutoff enabling button in the 3d Graph so that the cutoff is not enabled by default and this button toggles it on and off.
4. Made the fitsviewer not immediately load the 3d Graph view before the user selects something to view.
5. Fixed a minor error in the 3D graph zoom algorithm.
6. Set the Star Detection Algorithm for the FITS_Normal and FITS_Align views so that the user can get a good HFR of the stars in the frame.

In case of conflict in i18n, keep the version of the branch "ours"
To resolve a particular conflict, "git checkout --ours path/to/file.desktop"

In case of conflict in i18n, keep the version of the branch "ours"
To resolve a particular conflict, "git checkout --ours path/to/file.desktop"

This patch is for KStars.  It will set the path in the environment variable for the DSI to function on Mac.  Second, I thought the environment variables were getting a little messy, so I wrote a function that would clean them up a bit and make it easier to add more.  And finally, I found a better way to get the QImage for the Star Profile plot using a copy of a portion of the main QImage that matches the tracking box instead of using the tracking box pixmap function.  This way the overlays will not be included and the image zoom will not affect the profile.  The x and z axes are now actually pixels.

This patch and the accompanying files should create a 3D star profile plot that will be available in the guide and focus modules.  Note that currently, it uses the generatePixmap method of the fitsview to get a pixmap in the tracking box to plot in the 3d graph, so as of now the scales on the graph are all wrong and the zoom level on the image will change the result.  But it does work.

In case of conflict in i18n, keep the version of the branch "ours"
To resolve a particular conflict, "git checkout --ours path/to/file.desktop"

1. I added more functions to the nonlinear spin box to make it easier for KStars to add new recommended value items and to get the list of recommended values.
2. I made sure that in those new functions, it would not mess up the overall range of the spin box.  It will expand the range if the new recommended value is outside the current range but will not make the range smaller.
3. I made sure that the recommended values are always sorted correctly.
4. I changed the exposure spin box in the capture module to use the new functions.  This also ensures that 0 does not get added to the list if the min and the max do not get set from the current CCD.
5. I made the guide module request the valid exposure values and set the spin box to those recommended values on equipment connection, and save the message so if the user tries to set it to an invalid value, it doesn’t have to ask PHD2 what the valid values are every time.

As I said, I would try to take care of the Dither issues with PHD2.  This patch should prevent the observed problems.

1. It adds a Timer that will keep track of the amount of time since the dither request was initiated.  Then if PHD2 doesn’t respond within the dither timeout that it dithered, then either the capture sequence will continue or it will be aborted depending on whether the user selected ditherFailAbortsAutoGuide.
2. It eliminates some of the dither messages I added when dithering wasn’t working so well.
3. It adds a “Dither” label to the drift graph so that you know when dithering occurred.
4. It prevents GuideSteps during the dither and settle time from being added to the drift Graph.
5. It changes conditions of the dither failure after settle occurs so that if the user didn’t select that ditherFailAbortsAutoGuide it won’t abort the sequence

CCMAIL:kstars-devel@kde.org

1. Adds the NonLinearDoubleSpinBox class and makes both Capture Module and Guide Module use them with appropriate values (These can be changed).
2. Handles displaying an error message when the PHD2 pixel scale is incorrect because the user didn’t enter equipment info in PHD2.
3. Displays Connecting and disconnecting Messages for PHD2 and PHD2 equipment.
4. Improves some of the Guiding/Settling/Dithering code so that it gives more feedback and might fail to dither less often.
5. Adds the environment variable for the apogee camera firmware.

PHD2 Changes:
1.  Clean up PHD2 header file to match the skeleton of PHD2 methods like I did before for the cpp file
2.  Removes the former connect/disconnect code and fits them into the request/response/event format of the rest of the PHD2 communications
3.  The change above corrects some longstanding bugs with connecting, so I could remove the band-aid I put into the last patch.
4.  Prevents the calibration success event from accidentally starting guiding when first connecting
5.  Moves the request list removal stuff to the bottom of the cpp file so it is all together.
6.  Removes requests that had errors now and handles some of them.
7.  Simplifies the handling of the JSON objects so they ALL fit into either events, responses, or errors.
8.  Adds a LOSTLOCK abortTimer so that it waits 5 seconds after the star lock is lost before aborting guiding, in case it is reacquired
9.  Adds a list of GuideParamChanges but doesn’t implement them.
10.  Pauses between requests from PHD2 so that it doesn’t get too many requests at once.
11.  Adds the DE Guide Mode change communications between Ekos and PHD2.
12. Adds the Set Exposure Method so Ekos can tell PHD2 how long of an exposure to set.
13.  Disables the RA mode buttons and SWAP button for PHD2.

The first is that it would state “Calibration Completed” when it first connects even when PHD2 had not even started calibrating yet.  Basically when first told to guide, all guiders would  be told to calibrate and PHD2’s method would just emit that the calibration was successful then move on to guiding when the calibration would actually occur.  To solve this, I just prevent it from trying to calibrate PHD2 in the first place and just go straight to guiding, where PHD2 will automatically calibrate it.  Now it doesn’t print that the calibration is completed until it really is completed.

The second is an issue with the situation where PHD2 is already guiding when KStars connects to it.  Multiple state commands are issued, that it is guiding, that it is connected, etc.  The issue was that the Guide button was being disabled because it was guiding, then enabled because the equipment was connected, and then it tried to disable it again because it was guiding but it couldn’t because the first time it did it the PI was set to animated.  So I disabled the lines of code that prevented the setBusy command from executing if the PI indicator is running, but only for PHD2.  Now the Guide button works correctly.

I also deleted an unnecessary line of code from the patch I made yesterday that added the pulse graphs since the line above it already takes care of it.

+ Fix issue with non-guide dithering getting triggered while guiding.
+ Robert fixes for PHD2.

This patch should add the pulse graphs for RA and DEC with settings to turn them off.  It includes a slider to adjust the height of the corrections axis independently.  It also corrects the display of the RA error on the graph, which was upside down. It increases the amount of information that is saved when the guide data is exported.  And finally, it adds pulse data to the tooltip when any is available.

This has been corrected for the former sign difficulties I believe.

This patch adds tooltips for the guide buttons I added in the last patch.  It also remembers 3 of the settings in the KStars config file.  And finally comments some of the code.

1.  A Time Slider for the guide plots, which also functions as a scrolling view for the drift graph.
2.  A spin box for the desired accuracy for the drift plot.
3.  A save button for the guide data to export the data as a csv file.
4.  An autoscale function for the graphs.
5.  The ability to turn off RA or DEC plots individually in the drift graph.
6.  A clear button for the drift data.

Add defualt focuser option and make default devices selection more reliable and consistent across modules

Remove lock policy from filter manager and let focus module handle it as it reduces complexity and fixes some corner issues.
BUG:388129
FIXED-IN:2.9.0

Make sure to set the ST4 and Guide drivers on explicit set since combobox index does not necessary trigger signal to setST4

Summary:
Mostly builds it again, it seems that some things were not up to date.
Also there's some workarounds that don't seem to be required, at least not
on our docker sdk.

Test Plan: Been building it in https://hub.docker.com/r/sgclark/kde-ci-android/

Reviewers: mutlaqja, ckertesz

Reviewed By: ckertesz

Subscribers: ckertesz, #kde_edu

Tags: #kde_edu

Differential Revision: https://phabricator.kde.org/D8859

This should make sure that the host and port get set correctly for PHD2 and linguider when switching profiles.