Age | Commit message (Collapse) | Author |
|
|
|
We don't process untrusted user data but this should be the default
anyway.
|
|
It stays centered this way. Sadly yaw and pitch are still somewhat not
fixed in the user's reference frame.
|
|
|
|
|
|
|
|
|
|
On my Saitek X65F there's erratic panel button logspam. The panel's been
thrown out since it's useless, and the connector is detached from
everything.
In any case don't do stderr io overhead for button presses.
|
|
- Introduce rounding. Before, slider pos didn't correspond to the saved
setting until it "converged" several saves later.
- Move copy-pasted code to .cpp file.
|
|
value<t> needs to persist over the checkbox's lifetime, otherwise the
slot gets disconnected.
|
|
It makes sense to default for opentrack only. Depends can be messy and
they will.
|
|
|
|
|
|
|
|
Qt destructors deregister one component from another.
|
|
|
|
dinput8 has separate logic for pov hats. Before pov hat state was
totally ignored.
Don't support 8-way buttons. Only one button gets registered after a
diagonal pov hat press.
|
|
It's pointless and can lead to leaks.
|
|
The other settings are modified by the options dialog. These are in the
main ui. We need to be able to save modules without saving the options
dialog.
|
|
|
|
|
|
Now that "struct main_settings" doesn't implicitly reload in dtor, we no
longer have to pass it around from the user interface.
Only reload it where it's modified, i.e. in the options dialog. Changing
the filter/dialog/mapping comboboxen implicitly saves the main options
however.
|
|
|
|
The "struct opts" dtor doesn't run at dialog close time so invoke
"reload" on the bundle explicitly.
|
|
- the enumerator and the handle list are now static; joysticks were
erroneously destructed during refresh
- setup dinput only once; creating LPDIRECTINPUT8 many times causes
Release() to invalidate all handles
- don't create joystick handles many times; same thing applies
- refresh joy list on failed acquire; usually joystick is unplugged at
that moment leading to endless stderr spam
- use a static mutex as to avoid dinput races
Issue: #315
|
|
|
|
|
|
|
|
|
|
The height was zero on the test video. The ROI check only saw if width
is at least zero. Check for both to be greater than one.
Video provided by: @kblomster
Issue: #375
Also, fix minor issues:
- nix vars that can be const static in function scope
- don't call solvepnp twice where obj_points shift will do
- don't do bounds checking on vector elt access
- respect sprintf varargs type size; change to snprintf for no reason
- fix clamp-to-image logic
- set proper alpha for fps estimation
|
|
Issue: #375
|
|
|
|
|
|
Some new matrix element type requirements came up after opencv update
Also,
- switch to matrices of known sizes wherever possible
- split into functions for readability
- use member variables rather than locals to ensure heap allocation
There was also breakage wrt "unknown element type" deep in opencv that
only happens on release mode with no symbols. It's unknown to me
whether the issue got fixed or variable reordering made it corrupt
something else. It appears to work however. -fstack-protector-all
doesn't show any errors at all.
@kblomster says it worked in rc49p2. Looks like -fipa-pta has a
miscompilation despite finally working to begin with.
Reported-by: @kblomster
Issue: #375
|
|
Need link dynamic, otherwise symbol lookup failure.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Issue: #318
cf. http://forum.il2sturmovik.ru/topic/2291-open-track/page-26
|
|
|
|
|
|
- switch to our timer since it has more than millisecond precision
- use slider_value for settings
- sanitize allowed scaling curve bounds
- ensure min <= max in dialog
- ensure min <= max at filter runtime
- check for NaN taking -ffast-math into account
- simplify dialog logic
- change delta smoothing constant
|
|
|
|
It needs to filter less for new users.
|
|
|
|
|