- Power item odd scaling is made much more powerful for smaller servers (2P: x1.21 -> x2.5), and slightly more powerful for bigger servers (16P: x0.71 -> x0.78)
- Item distance bracket scaling was halved, so distance brackets are changed significantly less for both 2P and 16P. (2P: x1.21 in v1.5 -> x1.75 after this branch, 16P: x0.71 in v1.5 -> x0.89 after this branch)
- SPB scaling for 2nd place matches 1v1 scaling explicitly, instead of an arbitrary bonus.
- Does not function with controllers that don't produce continuous joystick events on hold, as testing unfortunately made clear.
- The entire thing would need to be rewritten to make up for that, and I don't have the time to invest that kind of energy into version 1.
- An unreliable feature is worse than no feature at all.
- They're binary inputs as processed by the game.
- A mechanism for binary inputs with no deadzone is already used for item and drift.
- Generally bound to triggers, not the stick.
- SDL still has a minimum deadzone, so won't fire endlessly.
* For left stick up/down/left/right input, allow holding a direction down to move continuously in a direction at a rate of 7 steps per second per axis
* Now supports accel axis being used for the Enter key/"accept" in menus, to mirror how an accel BUTTON bind will be used as an Enter event.
* If `kartvoterulechanges` is set to `Always`, the third option is always the opposite gametype, even if Encore is unlocked - that will now be present on the second option instead.
* This is both for consistancy with having Encore inaccessible on your gamedata, and a response to community feedback.
* The `kartencore` cvar being turned on will now be reflected in gametype rule changes from Battle to Race. This plugs one last hole in its area of effect for Race tracks.
This appears to be a long-time bug that exists upstream too, but nobody even knew about it before the MS rules alert made it clear this was actually happening!
This commit brings back the old code for W_CheckNumForNamePwad, using memcmp
instead of strncmp, which is readily inlined by modern C compilers and has
more acceptable performance in P_AddWadFile. Also removes some redundant
strlens, saving the result of one call instead.
This commit increases this branch's distance from SRB2 2.2, per Ashnal
and SteelT's note that I shouldn't worry too much about it, especially
when there's performance on the line.
