MantisBT - Zandronum
View Issue Details
IDProjectCategoryView StatusDate SubmittedLast Update
0001499Zandronum[All Projects] Bugpublic2013-09-10 20:152018-09-30 21:48
Reporterarkore 
Assigned To 
PrioritynormalSeverityminorReproducibilityrandom
StatusclosedResolutionfixed 
PlatformOSOS Version
Product Version1.2 
Target Version3.0Fixed in Version3.0 
Summary0001499: Slight chance that players will spawn on each other (same spawnpoint)
DescriptionSometimes players will spawn at the same spawn point at the start of a new round/match and therefore one players gets telefragged immediately upon round start.

With regular duel maps, the chance is less for this bug to occur.

However, for (a minor "mod" wad) where a duel map is a 1vs1 arena(1 room map) with two (strategically-placed) spawnpoints, the chance is greater for this bug to occur, regardless of "sv_spawnfarthest true" (I understand though why sv_spawnfarthest cannot be relied on when the engine has no idea what "farthest" is when simultaneously spawning two players at the start)

Video:'http://youtu.be/2HiauA7Ek0g [^]' (from Demo(below))
Demo:'https://www.dropbox.com/s/fbiho8c7vc9zt6g/2013.09.08_12.22.29_skulltag_data_126pk3.skulltag_actors_1-1-1pk3.krra1.1.cld [^]' (Long)

Minor IRC Chat log about it:'http://pastebin.com/22687p2D [^]'
Steps To ReproduceI was unable to reproduce this on my local server; but, it was happening on the live server. Please see video(above).
Additional InformationThis video shows the bug occuring on MAP04 of krra1.1.wad. But, the bug was occuring on other maps as well; there is nothing unique about this map, compared to the other maps. Every map is setup exactly the same, with only two deathmatch spawnpoints.

---

Idea(Solution): Perhaps the engine can be updated to index all spawnpoints of the map, and index all players, and specifically assign each indexed player to each indexed spawnpoint for when it's the beginning of a duel game/match?
TagsNo tags attached.
Relationships
related to 0001253closed  Team telefrag issue: G_CheckSpot always returns true if the spawn spot is > player height for the sector it's in 
related to 0001634closed Edward-san Weird spawn telefragging behavior on Duel32 + Transfer Heights 
related to 0001032closed Torr Samaho Z-Height support for player starts 
related to 0002791confirmed  Moving slightly off a spawn makes you vunerable to join/defect telefrag in team games 
Attached Files? krra1.1.wad (801,400) 2013-09-10 20:21
/tracker/file_download.php?file_id=1060&type=bug
Issue History
Date ModifiedUsernameFieldChange
2013-09-10 20:15arkoreNew Issue
2013-09-10 20:21arkoreFile Added: krra1.1.wad
2013-09-11 22:49arkoreNote Added: 0007172
2013-09-13 13:49DuskRelationship addedrelated to 0001253
2013-09-17 18:48WatermelonNote Added: 0007201
2013-09-17 20:11QentNote Added: 0007204
2013-09-18 17:06WatermelonNote Added: 0007219
2013-09-18 17:27QentNote Added: 0007222
2013-10-08 02:22ArcoPriorityhigh => normal
2013-10-08 02:22ArcoSeveritymajor => minor
2013-10-08 02:22ArcoStatusnew => confirmed
2014-06-11 15:48WatermelonRelationship addedrelated to 0001032
2014-06-11 15:49WatermelonRelationship addedrelated to 0001634
2016-07-19 02:53Ru5tK1ngNote Added: 0015383
2016-07-19 03:17Ru5tK1ngStatusconfirmed => feedback
2016-07-19 17:21Ru5tK1ngRelationship addedrelated to 0002791
2017-02-08 02:58Ru5tK1ngNote Added: 0016824
2017-02-08 02:58Ru5tK1ngStatusfeedback => resolved
2017-02-08 02:58Ru5tK1ngResolutionopen => fixed
2017-02-08 02:58Ru5tK1ngFixed in Version => 3.0
2017-02-08 02:58Ru5tK1ngTarget Version => 3.0
2018-09-30 21:48Blzut3Statusresolved => closed

Notes
(0007172)
arkore   
2013-09-11 22:49   
Someone mentioned that this could be occurring due to a "spectator" being idle at a spawnpoint. However, when I did extensive testing of this, it turned out to not be the case.

Actually, the telefragged spawn happened on the first try, and second try. But then it didn't happen the next time.

I also tried it without sv_respawnfarthest. Still happens. I then removed the spectator, and it still happens. But, only sometimes.

So, it still happens "randomly", regardless of the settings or spectators.

Perhaps a dev could maybe just take a quick look at the spawning code of the engine, and you might see something wrong with it? Or might see why this could be happening.

Thanks,
(0007201)
Watermelon   
2013-09-17 18:48   
'http://zandronum.com/tracker/view.php?id=1253 [^]'
(0007204)
Qent   
2013-09-17 20:11   
I'm not sure that they're related. I checked that map, and both player starts are on the ground.
(0007219)
Watermelon   
2013-09-18 17:06   
Ok since this seems to be quite a prevalent issue... and to fix two birds with one stone, could we consider writing our own bounding box checker and skip the spot unless it passes the max iterations before it forces a spawn? Easy math too, I love box math.
(0007222)
Qent   
2013-09-18 17:27   
The engine already does that for spawns that are on the ground, though.
(0015383)
Ru5tK1ng   
2016-07-19 02:53   
I simply could not duplicate the bug as illustrated in the video.

However, this is easily reproduced in medium sized LMS matches. My test case was 7 players on a map with 8 spawns . In 2.1.2 I was able to see the bug happen 3 times in a span of 10-15 minutes with over 30 rounds of LMS on a local server.

I repeated the same test with the build mentioned here:'http://www.zandronum.com/tracker/view.php?id=1634#c15362 [^]'

The bug never occurred.
(0016824)
Ru5tK1ng   
2017-02-08 02:58   
Since this fix: 'https://bitbucket.org/Torr_Samaho/zandronum/commits/3114b070fe24 [^]'

solved this issue I'll mark this a resolved as seeing the remaining telefragging issues pertain specifically to team games.