| 02:45.02 | macsforme | JeffM: so what happens when Y > X and we end up taking the square root of a negative number? |
| 02:46.24 | JeffM | you probbly want the fabs of it |
| 02:46.36 | JeffM | an use the component signs for direction |
| 02:48.13 | *** join/#bzflag jnoob22 (~jnoob@unaffiliated/jnoob22) |
| 02:50.21 | the_map | macsforme: I think that's just an imaginary problem |
| 02:51.14 | JeffM | what you are really doing is maping a vector into a spherical coordinate system |
| 02:53.40 | macsforme | ~lart the_map |
| 02:53.40 | infobot | frags the_map with his BFG9000 |
| 02:54.13 | the_map | great |
| 02:54.20 | the_map | I'm going to have to switch to windows so I can defrag |
| 02:55.20 | JeffM | windows has been doing that automaitcly in the background for years |
| 03:00.38 | macsforme | so that seems to give a value of pi/2 at each straight up/down/left/right angle, and 0 at each in-between angle |
| 03:01.42 | macsforme | when the stick is fully over in that direction |
| 03:02.05 | *** join/#bzflag The_Hand_of_Mike (~anonymous@cpe-72-180-51-26.tx.res.rr.com) |
| 03:02.08 | JeffM | 0? nothing in between? |
| 03:06.17 | JeffM | the more I think about it the more I wonder if it'll even work |
| 03:07.21 | JeffM | we want 45 to be 100% turn and 100% speed, but what do we want 30degrees to be? |
| 03:07.52 | JeffM | on square stick you can hold a 30degree angle and still have one axis be at 100% |
| 03:08.07 | JeffM | that's just not possible on a round area |
| 03:08.20 | JeffM | I can't see a way to combine rotary and linear units on the same stick |
| 03:10.56 | macsforme | so if the stick is at 100% in any direction then at least one of +X, -X, +Y, or -Y would be at maximum |
| 03:11.18 | JeffM | not at 45 |
| 03:11.35 | macsforme | at 45 two of them would be at 100% |
| 03:11.44 | macsforme | it may not work well |
| 03:11.57 | JeffM | then what is 30 degrees? |
| 03:12.30 | JeffM | i've got the stick jammed to the edge at 30 degrees, is one maxium? |
| 03:12.36 | JeffM | on a square stick I can do that |
| 03:13.01 | JeffM | are you really just saying that any vector with a magnitude of 1 has some axis at 100% |
| 03:13.22 | macsforme | dropping from 45 degrees to 0 degrees would have +X at maximum and +Y would drop to 0 |
| 03:13.59 | macsforme | yeah, that's what I was trying to achieve... I'm open to other ideas, though |
| 03:14.03 | macsforme | we could also just multiply the ratios by sqrt(2), and lose some of the precision toward the edges |
| 03:15.35 | macsforme | but I can already tell playing with a standard round stick with the original values isn't going to be viable (at least the way mine feels) |
| 03:16.05 | JeffM | round stick Already has lower percision at the edges, there a are just less units |
| 03:18.15 | JeffM | no mater what it's scaling, so some precision is going to be lost |
| 03:20.50 | JeffM | just make sure no axis computes to over 1 |
| 03:47.01 | BulletCatcher | To make a joystick with a circular motion area work almost like one with a square motion area, use a (smaller) square with its corners on the circle as the 100% position. |
| 03:47.41 | JeffM | so 0.707 from center |
| 03:47.53 | macsforme | that would be multiplying the inputs by sqrt(2) |
| 03:49.05 | macsforme | anyway, I can't really find a good multiplier, but part of the problem is my stick... at 45 degrees full over, the actual magnitude of the values is about 0.93 rather than 1.0 |
| 03:50.09 | BulletCatcher | The math would be more complicated than that. The joystick at full deflection 45 degrees would be 100%, but straight ahead would achieve 100% at sqrt(2). |
| 03:51.22 | BulletCatcher | You'd have to map the deflection vector onto the square equivalent of the mousebox. |
| 03:51.36 | macsforme | sqrt(2)/2, rather |
| 03:51.48 | macsforme | (for straight ahead) |
| 03:52.02 | BulletCatcher | Yeah. Math is hard. :-) |
| 03:53.46 | BulletCatcher | Basically, there would be a mousebox inside the area covered by the joystick the same as our standard unconfined mousebox is inside the area (screen) in which the mouse can move. |
| 03:54.37 | BulletCatcher | Make the stickbox smaller for quicker acceleration, or larger for finer control. |
| 04:21.00 | macsforme | we already kind of have that, the way the joystick control area is tied to the mousebox... but it's not very helpful, since there's no visual indication of the extents of the control area |
| 04:22.32 | macsforme | with my regular mousebox setting, my game controller stick reaches maximum at about 9% in X and about 15% in Y |
| 04:25.51 | macsforme | I may just stash this away for now... I think to do it properly, you would have to make the user do a full circle with the stick full over so you can detect the least magnitude, and possibly also allow the user to modify the sensitivity |
| 04:27.09 | macsforme | for now, the options are to use dual sticks, one for each axis, or to figure out the best mousebox size and maybe set up a toggle for different mousebox sizes if you play with both |
| 04:28.23 | macsforme | or use the D-pad :P |
| 04:41.40 | BulletCatcher | It would be reasonable, although not necessarily in 2.4, to draw the joystick circle and stickbox on the screen and display the position within. |
| 04:47.10 | JeffM | at one point I had mocked up a new input GUI that was consistant accross all input modes |
| 12:49.14 | *** join/#bzflag nadir (uid134094@gateway/web/irccloud.com/x-beqcnqihbeuwheuv) |
| 14:54.17 | *** join/#bzflag DTRemenak|RDP (~DTRemenak@cpe-172-112-10-183.socal.res.rr.com) |
| 14:54.17 | *** join/#bzflag DTRemenak|RDP (~DTRemenak@about/essy/CrazyCoder/DTRemenak) |
| 14:54.17 | *** mode/#bzflag [+v DTRemenak|RDP] by ChanServ |
| 15:48.49 | blast007 | anyone care enough to tell openSUSE to use our github releases instead of the old 2.4.2 sourceforge release? https://build.opensuse.org/package/view_file/games/bzflag/bzflag.spec?expand=1 |
| 15:49.59 | blast007 | and also that we're LGPL 2.1 not GPL 2.0+, and that we can build against SDL2 |
| 20:50.29 | *** join/#bzflag Zehra (~Zehra@unaffiliated/zehra) |
| 22:10.02 | *** join/#bzflag Zehra (~Zehra@2601:58b:c401:df8:b12b:138b:f52:3f76) |
| 22:10.02 | *** join/#bzflag Zehra (~Zehra@unaffiliated/zehra) |