Mac Calibration issue with 3.5.2
Robert Lancaster
rlancaste at gmail.com
Mon Mar 1 22:20:52 GMT 2021
Jasem,
I expect this would require compiling on the user’s part? He has said he didn’t want to compile.
Thanks,
Rob
> On Mar 1, 2021, at 5:04 PM, Jasem Mutlaq <mutlaqja at ikarustech.com> wrote:
>
> Hello Hy,
>
> A git-bisect would be ideal, but that requires the user to do that themselves. I think that's not hard to do on Linux, but I'm not sure on MacOS.
>
> --
> Best Regards,
> Jasem Mutlaq
>
>
>
> On Tue, Mar 2, 2021 at 12:54 AM Hy Murveit <murveit at gmail.com <mailto:murveit at gmail.com>> wrote:
> Rob,
>
> I believe Jean-Claude is willing to test what we send him, but not compile.
> There are only a handful of commits that touch guiding between 3.5.1 -> 3.5.2 and I don't really see anything affecting mounts.
> I have no clue how much work it is to make a DMG, so let me know if this is too much to ask.
> Could you make several DMGs, so we could narrow it down to a commit?
>
> Could you make DMGs for JCJ after the following commits:
>
> feb 3: c5c04f2e9cd77a095550f364aa4bda3f18927416
> feb 8: 8962f1cc618977b342b3565eb7c04606b7018805
> feb 9: c97f2adb4f7060812d8bd591593dd93c07e8e11b
> feb 27: d80cef232934af61b150610ccaaac966245f9f24
> feb 27: e6c2c9c4ccfced6960a872843d0e3aebf021af46
>
> Could you also make a DMG from the start of 3.5.2 (which should be about the same as 3.5.1)
> jan 10: c5e8b6234636e144b6272a72deae50865e018de7
> as a sanity check, to make sure it isn't something in the way you make DMGs.
>
> If someone has a better idea of how to find this, let me know.
>
> Thanks
> Hy
>
>
>
> More details about commits affecting guide.
>
> I just ran
> git log --name-only
> and copied the commits that have guide files
>
>
> commit e6c2c9c4ccfced6960a872843d0e3aebf021af46
> Author: Hy Murveit <hy-2 at murveit.com <mailto:hy-2 at murveit.com>>
> Date: Sat Feb 27 16:19:22 2021 -0800
>
> Save the estimated GPG period.
>
> kstars/ekos/guide/internalguide/gpg.cpp
>
>
> commit d80cef232934af61b150610ccaaac966245f9f24
> Author: Eric Dejouhanet <eric.dejouhanet at gmail.com <mailto:eric.dejouhanet at gmail.com>>
> Date: Sat Feb 27 14:59:48 2021 +0000
>
> Fix Guider capture/reacquire timeout handling
>
> kstars/ekos/guide/guide.cpp
>
>
> commit c97f2adb4f7060812d8bd591593dd93c07e8e11b
> Author: Eric Dejouhanet <eric.dejouhanet at gmail.com <mailto:eric.dejouhanet at gmail.com>>
> Date: Tue Feb 9 10:50:17 2021 +0000
>
> Fix issues found by Coverity Scan (level medium)
>
> Issues in SEP were set to ignore.
> Some issues set to ignore based on code intent.
>
> CID1445705 - unchecked cast in details table.
> CID1445707 - unchecked fread returned bytes in StarComponent.
> CID1445708 - unchecked cast in guider interface.
> CID1445711 - unchecked cast in observing list.
> CID1445714 - unchecked cast in details dialog.
> CID1445719 - unchecked cast in details dialog.
> CID1445721 - uninit var hfr in Analyze capture session.
> CID1445725 - unchecked cast in EkosLive message.
> CID1445726 - dead duplicated code in FITS packer.
> CID1445728 - possible division by zero in Scheduler job time estimation.
> CID1445729 - uninit var in Linear Focus.
> CID1445730 - unchecked cast in EkosLive message.
> CID1445734 - uninit vars in SchedulerJob.
> CID1445742 - unintended sign extension in FITS star filtering.
> CID1445743 - unchecked cast in EkosLive message.
> CID1445748 - uninit vars in Meridian Flip test.
> CID1445753 - stray semicolon in Capture test helper macro.
> CID1445757 - fix if construct (sorry Akarsh).
>
> Tests/kstars_ui/test_ekos_capture_helper.h
> Tests/kstars_ui/test_ekos_meridianflip_base.h
> kstars/dialogs/detaildialog.cpp
> kstars/ekos/analyze/analyze.h
> kstars/ekos/ekoslive/message.cpp
> kstars/ekos/focus/focusalgorithms.cpp
> kstars/ekos/guide/guide.cpp
> kstars/ekos/scheduler/scheduler.cpp
> kstars/ekos/scheduler/schedulerjob.h
> kstars/fitsviewer/fitsdata.cpp
> kstars/fitsviewer/fpackutil.c
> kstars/printing/detailstable.cpp
> kstars/skycomponents/starcomponent.cpp
> kstars/skyobjects/starobject.cpp
> kstars/tools/observinglist.cpp
>
>
> commit 578da02cb81a2cbb50e39e8412bcf04799cbc2d5
> Author: Jasem Mutlaq <mutlaqja at ikarustech.com <mailto:mutlaqja at ikarustech.com>>
> Date: Mon Feb 8 18:20:50 2021 +0300
>
> Disable placehold test on Windows until fixed later
>
> Tests/capture/CMakeLists.txt
>
> commit 8962f1cc618977b342b3565eb7c04606b7018805
> Author: Eric Dejouhanet <eric.dejouhanet at gmail.com <mailto:eric.dejouhanet at gmail.com>>
> Date: Mon Feb 8 09:56:45 2021 +0000
>
> Fix issues found by Coverity Scan (level High).
>
> https://scan.coverity.com/projects/tallfurryman-kstars/view_defects <https://scan.coverity.com/projects/tallfurryman-kstars/view_defects>
>
> Issues in SEP and Eigen3 were set to ignore.
>
> Fix CID258587 - uninit on key array of HTMesh range.
> Fix CID1342633 - uninit on RangeConvex sign_ in some situations.
> Fix CID1445642 - leak in Printing in the FOV path exporter.
> Fix CID1445593 - overflow in strncpy of FITS header keyword.
> Fix CID1445649 - resource leak in Align closest alignment point finder.
> Fix CID1445678 - overflow in strncpy of argv in infits.
> Fix CID1445692 - uninit on imagestats fields.
> Fix CID1445696 - uninit on fp_unpack hduname.
> Fix CID1445703 - overflow in strncpy of primaryCCD format.
> Fix CID1445706 - uninit on ra/dec pulses per pixel.
> Fix CID1445723 - overflow in strncpy.
> Fix CID1445738 - overlap in sprintf.
> Fix CID1445754 - overlap in sprintf.
> Fix CID1445750 - uninit on dtype.
>
>
>
> commit 8962f1cc618977b342b3565eb7c04606b7018805
> Author: Eric Dejouhanet <eric.dejouhanet at gmail.com <mailto:eric.dejouhanet at gmail.com>>
> Date: Mon Feb 8 09:56:45 2021 +0000
>
> Fix issues found by Coverity Scan (level High).
>
> https://scan.coverity.com/projects/tallfurryman-kstars/view_defects <https://scan.coverity.com/projects/tallfurryman-kstars/view_defects>
>
> Issues in SEP and Eigen3 were set to ignore.
>
> Fix CID258587 - uninit on key array of HTMesh range.
> Fix CID1342633 - uninit on RangeConvex sign_ in some situations.
> Fix CID1445642 - leak in Printing in the FOV path exporter.
> Fix CID1445593 - overflow in strncpy of FITS header keyword.
> Fix CID1445649 - resource leak in Align closest alignment point finder.
> Fix CID1445678 - overflow in strncpy of argv in infits.
> Fix CID1445692 - uninit on imagestats fields.
> Fix CID1445696 - uninit on fp_unpack hduname.
> Fix CID1445703 - overflow in strncpy of primaryCCD format.
> Fix CID1445706 - uninit on ra/dec pulses per pixel.
> Fix CID1445723 - overflow in strncpy.
> Fix CID1445738 - overlap in sprintf.
> Fix CID1445754 - overlap in sprintf.
> Fix CID1445750 - uninit on dtype.
>
> kstars/ekos/align/align.cpp
> kstars/ekos/guide/internalguide/calibration.h
> kstars/fitsviewer/fitsdata.cpp
> kstars/fitsviewer/fpack.c
> kstars/fitsviewer/fpackutil.c
> kstars/htmesh/HtmRangeIterator.h
> kstars/htmesh/RangeConvex.h
> kstars/htmesh/SpatialException.cpp
> kstars/indi/indiccd.cpp
> kstars/printing/shfovexporter.cpp
>
>
> commit c5c04f2e9cd77a095550f364aa4bda3f18927416
> Author: Hy Murveit <hy-2 at murveit.com <mailto:hy-2 at murveit.com>>
> Date: Wed Feb 3 10:41:03 2021 +0000
>
> New UI to help user correct polar alignment.
>
> Tests/polaralign/test_polaralign.cpp
> Tests/polaralign/test_poleaxis.cpp
> Tests/polaralign/test_poleaxis.h
> kstars/CMakeLists.txt
> kstars/ekos/align/align.cpp
> kstars/ekos/align/align.h
> kstars/ekos/align/align.ui
> kstars/ekos/align/alignview.cpp
> kstars/ekos/align/alignview.h
> kstars/ekos/align/polaralign.cpp
> kstars/ekos/align/polaralign.h
> kstars/ekos/align/poleaxis.cpp
> kstars/ekos/align/poleaxis.h
> kstars/ekos/align/rotations.cpp
> kstars/ekos/align/rotations.h
> kstars/ekos/guide/internalguide/starcorrespondence.cpp
> kstars/ekos/guide/internalguide/starcorrespondence.h
> kstars/fitsviewer/fitsdata.cpp
> kstars/fitsviewer/fitssepdetector.cpp
> kstars/fitsviewer/fitsview.cpp
> kstars/fitsviewer/fitsview.h
> kstars/kstars.kcfg
>
>
>
>
>
> On Mon, Mar 1, 2021 at 1:23 PM Robert Lancaster <rlancaste at gmail.com <mailto:rlancaste at gmail.com>> wrote:
> Hi Hy,
>
> I have been following this on the forum. I don’t know, since I don’t know what changed since 3.5.1. I looked briefly through the commits, but saw nothing that should have changed guiding, from the descriptions. But of course, something else could have been done in a commit. I was certainly able to help him get set up to test to see if the issue existed in 3.5.1, so I helped, but I don’t know what could have caused this and haven’t had lots of time, so I was very happy to see your debugging progress.
>
> As for testing, I could always build him a new custom DMG with a change to test, but we would need to know exactly what should be changed. If he doesn’t want to set up a development environment or test on Linux, then we would need to build him a full DMG every time we would want him to test something. This is certainly doable, but it could take awhile.
>
> I think what we really need is to know if somebody has made a change that would affect guiding calibration in the last few weeks. Otherwise it is like a needle in a haystack.
>
> Thanks,
>
> Rob
>
>> On Mar 1, 2021, at 4:10 PM, Hy Murveit <murveit at gmail.com <mailto:murveit at gmail.com>> wrote:
>>
>> Jasem, Rob, et al
>>
>> Jean-Claude (@jcj) has been having issues with the mac 3.5.2 DMG where he can't calibrate at all. He retries the tests on 3.5.1 and everything works fine. For instance see this forum post
>> https://indilib.org/forum/development/8882-new-kstars-3-5-2-dmg-to-test.html?start=36#67881 <https://indilib.org/forum/development/8882-new-kstars-3-5-2-dmg-to-test.html?start=36#67881>
>> and some others on pages 4,5,6 of that thread.
>>
>> I looked into his logs and saw that the calibration star detection and logic is all working, and in the "move RA out" phase, things work (though it moves out very quickly, 15-20 pixels, in the first step, so that calibration only requires one step before it exceeds its pixel limit), but in the "move RA back" phase, the RA keeps moving out until the calibration fails.
>>
>> I'm not sure whether the signals to the mount are not getting there or what, and am at a loss on how to help him. JCJ will be sending one more set of logs with Indi Mount debug logging enabled.
>>
>> It is likely a software issue (since 3.5.1 works). Does anyone have any idea how to proceed to debug this?
>>
>> Any ideas of things that changed between 3.5.1 and 3.5.2 that might be related to this?
>>
>> Hy
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kstars-devel/attachments/20210301/26b609ee/attachment-0001.htm>
More information about the Kstars-devel
mailing list