summaryrefslogtreecommitdiffhomepage
AgeCommit message (Collapse)Author
2016-11-03move simple-mat from logic to compat to simplify dependenciesStanislaw Halik
2016-11-03logic/tracker, dinput, tracker/joystick: fix rounding errorsStanislaw Halik
Issue: #487 Reported-by: @H-Bear-22 Closes #487 Also use a power of two for joy_axis_size. Floats are base 2. axis_max in tracker/joystick had a fencepost error.
2016-11-02tracker/rift-{042,080}: don't overwrite other rifts' settingsStanislaw Halik
2016-11-02tracker/aruco: don't break build if opencv not foundStanislaw Halik
2016-11-02spline-widget: prevent breakage on values close to zeroStanislaw Halik
2016-11-02tracker/rift-140: addStanislaw Halik
2016-10-29spline-widget: remove aliasing artifactsStanislaw Halik
We're using QPainterPath and its bezier curves. However, we supply it a plethora of intermediate values to mitigate different results in interpolation.
2016-10-29spline-widget: don't compute the same thing over and over againStanislaw Halik
2016-10-29spline-widget: avoid sqrt in global namespaceStanislaw Halik
2016-10-29spline-widget: revert last commitStanislaw Halik
In the last line of the blocks we're reducing the precision to reduce rounding errors. It's actually proper to use round rather than truncate here.
2016-10-28spline-widget: fix rounding bug introduced in a3b890b9b4opentrack-2.3-rc99p14Stanislaw Halik
Since we're already adding snap_x/2, truncation is the correct choice here.
2016-10-28gui/options: rename r[xyz] to yaw/pitch/roll in camera offsetStanislaw Halik
2016-10-28spline-widget: use round, not trunc when converting pixel to valStanislaw Halik
2016-10-28spline-widget: don't recalculate point valueStanislaw Halik
Only apply snap for closeness check.
2016-10-28gui/main-window: round raw/mapped value displayStanislaw Halik
Otherwise infinitesimal differences prevent from reaching max mapped value.
2016-10-26compat/util: split out run-in-thread{,-async}Stanislaw Halik
2016-10-26spline-widget: fix logic for detecting points being too closeStanislaw Halik
It was still possible to reorder points by dragging one of them. Also get rid of the iteration. It was previously introduced to try converge one of the points to another position. Allow for changing Y value while the point is dragged. Take snap into account. Use newly introduced "iround<t>".
2016-10-26spline-widget: use stable sortStanislaw Halik
2016-10-26compat/util: introduce "iround" for round to integerStanislaw Halik
2016-10-26compat/util: move stuff aroundStanislaw Halik
2016-10-26gui/mapping: set snap higher for more precisionStanislaw Halik
Previous snap values were so low that getting the exact value required pixel-perfect accuracy. Fix that.
2016-10-23compat/util: these type params are unusedStanislaw Halik
2016-10-23cmake: change mingw-w64 compiler pathStanislaw Halik
2016-10-20compat/nan: fix isnan in presence of -ffast-mathStanislaw Halik
The trick prevents GCC isnan builtin from getting called.
2016-10-20proto/freetrack: game list updateStanislaw Halik
2016-10-20contrib: output small csv diffsStanislaw Halik
Perl line noise returned arbitrarily messed up diffs when merging contents.
2016-10-19options/slider: check for floating-point division by zeroStanislaw Halik
It's not strictly necessary but a good practice nevertheless.
2016-10-19dinput, tracker/pt: reword comments and fix harmless warningsStanislaw Halik
2016-10-19compat, tracker/pt: add warn_unused_result compat macroStanislaw Halik
2016-10-19filter/accela: simplify QString usageStanislaw Halik
2016-10-19tracker/pt: simplify state displayStanislaw Halik
2016-10-19tracker/pt: poll for tracker info only when it could changeStanislaw Halik
2016-10-19tracker/pt: don't delay camera stop artificiallyStanislaw Halik
If the filter crashes on quick stop/start it's not our fault.
2016-10-19tracker/pt: show tracker state immediatelyStanislaw Halik
2016-10-19tracker/pt: don't call calibration twiceStanislaw Halik
Also simplify timer logic.
2016-10-19options/tie: output slider_value after truncation errorStanislaw Halik
2016-10-19filter/accela: what the hell is this nonsenseStanislaw Halik
2016-10-19gui/options: also rename to "relative translation" in optionsStanislaw Halik
Reported-by: @MathijsG Issue: #461
2016-10-19tracker/pt: minor cleanupStanislaw Halik
2016-10-19compat/util: fix conversion from value<t> for clampStanislaw Halik
2016-10-11tracker/pt: fix brain fart in extractorStanislaw Halik
It's multiplied by 3 just a few lines below. So ~2 is actually a good lower bound.
2016-10-11options/connector: bundle can't be null at all hereStanislaw Halik
It's a shared_ptr<detail::bundle> stored in base_value.
2016-10-11tracker/pt: increase min point radius to sensible valueStanislaw Halik
We can't possibly have LEDs smaller than (6/3=2) each. Especially not (2.5/6).
2016-10-11options/connector: fix pointless check for nullptrStanislaw Halik
2016-10-11pose-widget: make octopus more slimStanislaw Halik
It looked weird as such a fatso forgotso.
2016-10-10tracker/pt: get rid of heap allocationStanislaw Halik
Use stack arrays rather than vectors. std::array may be a better choice though.
2016-10-10tracker/pt: add unperspective for yawStanislaw Halik
It's commented out for the time-being.
2016-10-10tracker/pt: change focal length formulaStanislaw Halik
It looks kind of funky. At the moment it gives me correct Z (with a ruler) given right FOV. Though the pitch value is still skewed given lack of perspective transform for Y pixel value in extractor.
2016-10-07tracker/rs: fix warning on msvc 15 preview 5Stanislaw Halik
2016-10-07tracker/pt: C++11-izeStanislaw Halik
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-30 22:10:20 +0000