* Make VR mode more seamless
Works to make the `-vr` command flag more seamlessly work (which means the new KNet button, `Play in VR` will also now actually work). Fixes#6430.
First, creates a new in-game option called VR Resolution and have the default be 2500x2500, with a range that goes from 1000x1000 to 6000x6000.
Second, if `-vr` flag is used, then it auto sets flags of `-capture_mouse` `-stretch_menu` and `-window_res 1000x1000` (or anything that's for sure smaller than the screen but not too small, to maintain mouse precision).
It also overrides settings of `-vysnc` and `-window` to off, by ignoring the in-game values (and does not show these options in menu while in VR mode).
Finally, it reads the in-game value of `VR Resolution` instead of `Resolution`. It also shows the `VR Resolution` in-game option in the menu, while hiding the options of `Resolution`, `Windowed Mode` and `Vsync`.
If the game is run in standard mode then FSO does what it current does, including using the `Resolution` value for screen resolution. It also does not display the new in-game option for `VR Resolution`.
* actually use the vr_enumerator
* use emplace_back instead
* update res vr importance
* Fixes for VR one-button mode
* try reordering removes for clang
* fix string ids
* test to see if clang is being weird
* revert clang test
* strings index update
* cleanup options manager
---------
Co-authored-by: Birk Magnussen <[email protected]>
This refactors FontManager to improve it in several ways:
1. Change all the dynamic memory management to use `new`, rather than mixing `new` and `vm_malloc`
2. Change the current font to use an index rather than a pointer
3. Take advantage of more C++11 features to avoid redundant lookups
4. Use case-insensitive name comparison
5. Avoid overflowing the `fnt->filename` field
6. Use `Warning` rather than `Error` in a non-critical error situation
* base working version of unlimited debris types
* Proper FRED support
* make qtFRED handle asteroids properly
* fix qtFRED assertion
* further clean out MAX_RETAIL_DEBRIS_TYPES
* These should be sizes not POFs
* Move this to it's own method
* convert asteroid subtypes to a vector
* unlimit asteroid subtypes
* fix some global shadowing
* signed/unsigned mismatch
* Address feedback
This refactors the object_h struct to store objects as indexes rather than pointers. This is a prerequisite for making the collection of objects dynamic in the future. See also #6100.
Follow-up to #5800 and #5829.
Allow the resize mode to be specified to `gr_set_clip()` without changing the active graphics-wide resize mode. Along with `drawStringResized`, this provides a shorthand to allow text to be drawn at different sizes, with the appropriate clipping regions, without requiring access to the `CurrentResizeMode` virtvar.
Since `gr_string` can accept a length parameter, there's no need to copy each line before it is drawn.
Related to #5495 but not one of the two options mentioned.
* initial API for hud config elements
* get rendering, highlighting, and selecting working
* Add a way to create color objects
* some cleanup and fixes
* go away we dont need you
* allow arbitrarily scaling the hud config draw
* last few config options and innitial preset handling
* vectorize hud presets and setup handling
* some last minute fixes
* iiiiiiiiiiiiiiiiiiiiiiint
* really
* documentation and address feedback
* completely revert this as well
* more documentation
* document all the things
* comment out unused parameter
* peach emoji retentive
* whatever
* return types
* color
Scripts that render models in the On Object Render hook were effectively doing immediate renders in the middle of queued renders. This caused some data, such as model orientation, to 'leak' between renders. It might also have something to do with why the bug in #4637 had the particular graphical side effects that it did.
Anyway, changing the immediate renders to queued renders fixes the leaky orientations, and is probably a good upgrade for other reasons as well.
Follow-up to #4602. Overall the PR was good, but there are a few extra small changes that should go in. In order of priority:
1) The `exitLoop`, `skipMission`, and `skipTraining` functions should check for campaign mode.
2) The `isVisible` virtvar should be changed to a function, because calling it may have side effects. In standard usage, the briefing formula is evaluated exactly once per stage on briefing initialization, so scripts should follow this convention.
3) The `hasGoalsSlide` function should be called `hasGoalsStage`, because "stage" is the common terminology in FSO.
4) Paging out mission messages is already done in `game_level_close` and should be removed from not only the scripting method but the method it was based on.
5) For consistency, unused parameters should use the `SCP_UNUSED` macro.
6) The `isTraining` flag check can be slightly streamlined using the double-not trick.
Mike Nelson