Closed trampyjoe closed 8 months ago
Thank you for the log and the details. It seems to be flooded with errors from the LNM web server. What happens if you run LNM without the web server? Do you use a MSFS tool which accesses the LNM web-server?
I'll try here with web enabled.
Yes, I use the webserver occasionally. Will test without the webserver running tomorrow. Cheers :)
@albar965 the log entry i guess which is the flooding is a warning about distance
not delivered by the web ui.
It is triggered by this code:
https://github.com/albar965/littlenavmap/blob/6e1e71cf45b376407041bd39cab5056ffe38af8d/src/web/requesthandler.cpp#L212
The web ui by default does not deliver a distance
. The distance is inferred from the starting distance and the number of zoom in / out commands on the mapPaintWidget
.
The web ui CAN deliver a distance
to force overriding mapPaintWidget's "internal zoom state(s) (corrected zoom and requested zoom)" (which are saved to the session), this is made use of in the web plugin autozoom
.
Log entries timestamp indicates auto refresh
has been enabled (at the default refresh interval of 1s).
I would be intrigued whether a compile of LNM / atools with Debugging (output) disabled results in a faster / smoother globe rendering. (I am too wary of setting up LNM for compilation again to test this myself now.) (How about releasing a non-debug build as well as a debug-build for those who like to pinpoint an issue? (You could offer to download the debug build from the non-debug build when the log file is requested from the toolbar menu)).
@albar965 I omitted the contains
check in asFloat
in favor of catching this case by QByteArray
s toFloat
.
If the !ok
case tests for the wrong value in question being the QMultiMap<QByteArray, QByteArray> params
default constructed value returned when the key requested with is not contained (the dcv appears to be the empty string), the case of !contains
can be distinguished from a "real" invalid float and debug text output only in the latter case. This approach can be applied to the other asXXX
methods as well.
Yes, I use the webserver occasionally. Will test without the webserver running tomorrow. Cheers :)
@trampyjoe : Not sure why this happens. I remember to have a similar bug fixed some time ago. I'll do more testing. The only reason to write to a database (the error is caused by this) is either modification of an userpoint or a logbook entry. The latter one can happen on landing or takeoff. Did it happen while landing or taking off? I suppose you run only one LNM instance, right?
@u-an-i : Thank you for the hint. I'll reduce logging and have a look at the distance conversion. There should be no debug log for each request. Debug is on per default as configured in logging.cfg. I know debugging needs some cleanup. Especially regarding info/debug levels.
Just loaded up exactly as yesterday and tried to replicate the error before trying without webserver running. Could not replicate the error.
Error happened taking off, and every time before has been during take off. I've only just noticed, now that I'm paying attention, that the flights have not been logged (obviously!). I only run one instance of LNM
Will monitor during future flights and will report back with logs next time I get the error.
@trampyjoe One thing you can try: Go to the LNM logbook, right click and select Add logbook entry
. Leave all as it is in the dialog and click Ok
. You have a permission problem on the database if this causes the exception. Means, that the file is probably not writable by LNM.
Go to the LNM logbook, right click and select
Add logbook entry
. Leave all as it is in the dialog and clickOk
. You have a permission problem on the database if this causes the exception. Means, that the file is probably not writable by LNM.
No exception when creating a new logbook entry.
No exception when creating a new logbook entry.
Good. This means the file is writeable. I'll do more testing here but could not reproduce it yet.
Closing for now. Feel free to reopen if it appears again.
Have had a search through and not found the exact same problem from other bug reports (a few similar).
Randomly on take off I get the following pop-up error
Caught exception in file "..\atools\src\gui\application.cpp" line 83. database is locked Unable to commit transaction. SqlDatabase::transaction() on invalid database. If the problem persists or occurs during startup delete all settings and database files of Little Navmap and try again. If you wish to report this error attach the log and configuration files to your report, add all other available information and send it to the contact address below.
Can only duplicate on certain flights (this evenings one was duplicateable so am report it now).
No scenery addons Aircraft livery addon (https://flightsim.to/file/29684/xcub-bush-divers-series-liveries) Windows 10 64 MSFS 1.35.21.0 LNM 2.8.12 64-bit
abarthel-little_navmap.log little_navmap.txt