View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0000260 | FSSCP | gameplay | public | 2004-09-12 14:10 | 2005-08-31 05:10 |
| Reporter | Lightspeed | Assigned To | phreak | ||
| Priority | normal | Severity | tweak | Reproducibility | always |
| Status | resolved | Resolution | fixed | ||
| Summary | 0000260: View cone messed up on spawn missiles | ||||
| Description | Spawn missiles (like the piranha) seem to have messed up view cones on the spawned missile babies. If you fired a piranha on retail, ALL missiles turned towards the target (pretty much full view cone) and chased it. Now, if you do the same thing with FS2open, just two or three of the missiles will "see" the target, turn and chase it (the ones pointing in more or less its direction). It looks like the view cone of the spawned missiles is somehow handled in another way (i.e. it's a lot smaller). If this is indeed a bug, I guess it sneaked in with the firing cone feature for cluster weapons. Someone verify this please, and check if the bug can be fixed. | ||||
| Tags | No tags attached. | ||||
|
|
It's smaller, by half in fact, but I'm not sure what the reasoning is there since it does change the original behaviour. It's easy enough to change it so that it's back to a full view but I'd rather ask phreak about it first in case there is another reason it's set where it is. |
|
|
Phreak made a spawn cone table option, so he would indeed be the person to ask. |
|
|
they aren't tracking but the spawn angle code isn't an issue, it deafults to a full sphere which what happened in retail. |
|
|
T: If you haven't already, revert this to the retail-esque behaviour. It messes up multiplayer mission and singleplayer missions alike, and makes certain weapons pretty much useless. |
|
|
i looked at this and rand_vec_in_cone() takes an angle and doubles its FOF (it goes from -angle,angle). So it would make sense just to use 180 to get a sphere. However if i use 360 for some reason, the spawned missiles track better (from what i've seen). The relevant code for the default spawn angle is on line 1770 in weapons.cpp in parse_weapon() |
|
|
Is this back to it's default behaviour? |
|
|
oops forgot aboot this. fixed... hopefully |
|
|
Lightspeed, can you test this? |
|
|
awaiting Lightspeed's feedback, close in a month and reopen and assign back to me if necessary |
|
|
Reminder sent to Lightspeed Feedback required. |
|
|
Would anyone be so kind to link me to a "fixed" build? I'd probably be able to run a quick test. |
|
|
I'm guessing any recent build would have this fixed. Here's one that Phreak posted on 5/22. http://fs2source.warpcore.org/exes/phreak/fs2_open_phr_05222005.rar |
|
|
From what I've seen, it doesn't look as if it were fixed. |
|
|
Assigning it to myself since it's probably fixable with MISSION_FLAG_USE_NEW_AI. |
|
|
no it isn't. here are the relavant code changes http://fs2source.warpcore.org/cgi-bin/cvsweb/cvsweb.cgi/fs2_open/code/weapon/weapons.cpp.diff?r1=2.63&r2=2.64&f=h |
|
|
Oh excellent... thanks for pointing that out, Phreak. :) I don't think that's quite the same thing as retail, though. A 180 degree cone is still oriented along the object's forward axis, whereas retail seemed to shoot them off in all directions. I'll commit this in a sec. |
|
|
the angle argument in vm_vec_random_cone() uses the range from (-angle,angle), so i think 180 would suffice. I want to try putting some debug lines and alt-pause to see if all of the missiles are properly oriented. |
|
|
All the same, I'd prefer to use the original code. Here's what it looks like now: // phreak's code if (The_mission.flags & MISSION_FLAG_USE_NEW_AI) { // for multiplayer, use the static randvec functions based on the network signatures to provide // the randomness so that it is the same on all machines. if ( Game_mode & GM_MULTIPLAYER ) { static_rand_cone(objp->net_signature + i, &tvec,&objp->orient.vec.fvec,wip->spawn_angle,&objp->orient); } else { vm_vec_random_cone(&tvec,&objp->orient.vec.fvec,wip->spawn_angle,&objp->orient); } } // original code else { // for multiplayer, use the static randvec functions based on the network signatures to provide // the randomness so that it is the same on all machines. if ( Game_mode & GM_MULTIPLAYER ) { static_randvec(objp->net_signature + i, &tvec); } else { vm_vec_rand_vec_quick(&tvec); } } vm_vec_scale_add(&pos, &objp->pos, &tvec, objp->radius); |
|
|
use new AI shouldn't be in there since it pertains to ship AI, not how the missiles are spawned. Plus i don't want to set a flag in every mission in the main FS2 campaign so i could use modded infyrno missiles. Besides making USE_NEW_AI mandatory would also allow the AI to actually use trebs and really screw with the balance of the mission. I sorta want to branch out the USE_NEW_AI flag anyway since people may want to use the AI's better selection of primaries, but not need to worry about trebs flying in every direction. |
|
|
Well part of the purpose of putting USE_NEW_AI in all these places is to flag them so we can come back later with a more flexible customization system. However since spawn_angle is an optional flag anyway, howabout this instead: wip->spawn_angle = -1; if (optional_string("$Spawn Angle:")) stuff_float(&wip->spawn_angle); . . . // phreak's code if (wip->spawn_angle > 0) edited on: 08-30-05 19:19 |
|
|
well im curious as to why having 180 as a default doesnt seem to work. It may mean that another bug may exist. |
|
|
It's because vm_vec_random_cone spins the cone around orient.vec.fvec, whereas vm_vec_rand_vec_quick points the cone in all directions. It's not the same random number generator. |
|
|
Committed. I'm resolving this. :) |
|
|
No its not fixed, its just hiding the bug. If someone has the spawn angle set to 180 then its going to display a different behavior then having it not set, even if they're supposed to be the same. Something is going on and its probably effecting behavior when the spawn angle is less than 180 as well |
|
|
ok fixed the right way. the matrix argument in vm_vec_random_cone() was objp->orient and should have been the matrix, m, formed by vm_vector_2_matrix(&m, &objp->orient.vec.fvec, NULL, NULL); and guess what? the tracking was also effective if the spawn angle was smaller than 180 edited on: 08-31-05 00:55 |
|
|
Oh. Hm. Interesting. Well does that mean if you remove the &objp->orient argument it works properly? I'm still skeptical whether vm_vec_random_cone(&tvec, &objp->orient.vec.fvec, 180) would provide the same result as vm_vec_rand_vec_quick(&tvec). |
|
|
yea it does now. however i don't want to muck around with special cases |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2004-09-12 14:10 | Lightspeed | New Issue | |
| 2004-11-20 12:06 | taylor | Note Added: 0001363 | |
| 2004-11-21 12:02 | kasperl | Note Added: 0001367 | |
| 2004-11-21 18:22 | phreak | Note Added: 0001368 | |
| 2005-01-15 19:30 | Lightspeed | Note Added: 0001555 | |
| 2005-01-17 04:30 | phreak | Note Added: 0001558 | |
| 2005-03-12 20:08 | kasperl | Note Added: 0001759 | |
| 2005-03-12 20:56 | phreak | Note Added: 0001784 | |
| 2005-03-12 20:58 | kasperl | Note Added: 0001786 | |
| 2005-03-12 21:02 | phreak | Status | new => assigned |
| 2005-03-12 21:02 | phreak | Assigned To | => phreak |
| 2005-04-04 17:00 | phreak | Note Added: 0002089 | |
| 2005-04-04 17:00 | phreak | Assigned To | phreak => kasperl |
| 2005-04-04 17:00 | phreak | Status | assigned => feedback |
| 2005-05-25 02:02 | Goober5000 | Note Added: 0002484 | |
| 2005-05-26 02:14 | Lightspeed | Note Added: 0002520 | |
| 2005-05-26 03:10 | Goober5000 | Note Added: 0002521 | |
| 2005-05-26 22:32 | Lightspeed | Note Added: 0002524 | |
| 2005-08-18 20:04 | Goober5000 | Status | feedback => assigned |
| 2005-08-18 20:04 | Goober5000 | Assigned To | kasperl => Goober5000 |
| 2005-08-18 20:07 | Goober5000 | Note Added: 0003021 | |
| 2005-08-30 14:45 | phreak | Note Added: 0003202 | |
| 2005-08-30 15:15 | Goober5000 | Note Added: 0003206 | |
| 2005-08-30 15:19 | phreak | Note Added: 0003207 | |
| 2005-08-30 15:21 | Goober5000 | Note Added: 0003208 | |
| 2005-08-30 22:04 | phreak | Note Added: 0003212 | |
| 2005-08-30 23:17 | Goober5000 | Note Added: 0003213 | |
| 2005-08-30 23:18 | Goober5000 | Note Edited: 0003213 | |
| 2005-08-30 23:19 | Goober5000 | Note Edited: 0003213 | |
| 2005-08-30 23:25 | phreak | Note Added: 0003214 | |
| 2005-08-30 23:56 | Goober5000 | Note Added: 0003216 | |
| 2005-08-31 00:02 | Goober5000 | Status | assigned => resolved |
| 2005-08-31 00:02 | Goober5000 | Resolution | open => fixed |
| 2005-08-31 00:02 | Goober5000 | Note Added: 0003217 | |
| 2005-08-31 04:23 | phreak | Status | resolved => feedback |
| 2005-08-31 04:23 | phreak | Resolution | fixed => reopened |
| 2005-08-31 04:23 | phreak | Note Added: 0003226 | |
| 2005-08-31 04:23 | phreak | Status | feedback => assigned |
| 2005-08-31 04:23 | phreak | Assigned To | Goober5000 => phreak |
| 2005-08-31 04:53 | phreak | Note Added: 0003227 | |
| 2005-08-31 04:55 | phreak | Note Edited: 0003227 | |
| 2005-08-31 05:04 | Goober5000 | Note Added: 0003228 | |
| 2005-08-31 05:09 | phreak | Status | assigned => resolved |
| 2005-08-31 05:09 | phreak | Resolution | reopened => fixed |
| 2005-08-31 05:10 | phreak | Status | resolved => feedback |
| 2005-08-31 05:10 | phreak | Resolution | fixed => reopened |
| 2005-08-31 05:10 | phreak | Note Added: 0003229 | |
| 2005-08-31 05:10 | phreak | Status | feedback => resolved |
| 2005-08-31 05:10 | phreak | Resolution | reopened => fixed |
