MC-90711Marker ArmorStands cannot be selected within r=0
If an ArmorStand has Marker:1b set, it cannot be found with r=0.
Setup:
Stand somewhere, without moving
/summon ArmorStand ~ ~ ~
/testfor @e[r=0]
What should happen:
Found
Found Armor Stand
What actually happens:
Found {player}
This is not because of the relative position of the stand; If you walk a bit away and run
/execute @e[type=ArmorStand,c=1] ~ ~ ~ testfor @e[r=0]
the command will fail.
Other entities may be selected within the range, however; If the first experiment is run with a NoAI Sheep summoned as well, at exactly the same position, the command will find the player and the sheep, but not the stand.
Notes:
- A range of 1 or more results in correct behavior of commands.
- Non-Marker stands can be selected with r=0
- A possible cause of this may be the way entities in range are checked in the code; As the Marker ArmorStand has no hitbox, selection by range may be inherently buggy due to the calculation method.
Environment
Linux Server
&
Mac OS X 10.11
Java 8u60
Created Issue:
Marker ArmorStands cannot select non-player entities within @e[r=0]
If you execute testfor @e[r=0] on an ArmorStand with Marker:1b, it will not find anything. This is obviously untrue, as the entity itself is within a zero range.
summon ArmorStand ~ ~ ~
{Marker:1b}walks away a bit
execute @e[type=ArmorStand,c=1] ~ ~ ~ testfor @e[r=0]
Fails
When the selector is changed to @e[r=0,type=!Player] or @e[r=0,type=ArmorStand], it cannot select anything.Environment
Linux Server
If you execute testfor @e[r=0] on an ArmorStand with Marker:1b, it will not find anything. This is obviously untrue, as the entity itself is within a zero range.
summon ArmorStand ~ ~ ~
{Marker:1b}walks away a bit
execute @e[type=ArmorStand,c=1] ~ ~ ~ testfor @e[r=0]
Fails
When the selector is changed to @e[r=0,type=!Player] or @e[r=0,type=ArmorStand], it cannot select anything.Note: If the
Marker ArmorStands cannot select non-player entities within @e[r=0]
Marker ArmorStands cannot selectnMarker ArmorStands cannot be selected within r=0
Linux Server
Linux Server
&
Mac OS X 10.11
Java 8u60
If you execute testfor @e[r=0] on an ArmorStand with Marker:1b, it will not find anything. This is obviously untrue, as the entity itself is within a zero range.
summon ArmorStand ~ ~ ~
{Marker:1b}walks away a bit
execute @e[type=ArmorStand,c=1] ~ ~ ~ testfor @e[r=0]
FailsWhen the selector is changed to @e[r=0,type=!Player] or @e[r=0,type=ArmorStand], it cannot select anything.Note:
If theIf an ArmorStand has Marker:1b set, it cannot be found with r=0.
Setup:
{Marker:1b}
Stand somewhere, without moving
/summon ArmorStand ~ ~ ~/testfor @e[r=0]
What should happen:
{player}
Found
Found Armor Stand
What actually happens:
Found {player}This is not because of the relative position of the stand; If you walk a bit away and run
/execute @e[type=ArmorStand,c=1] ~ ~ ~ testfor @e[r=0]
the command will fail.Other entities may be selected within the range, however; If the first experiment is run with a NoAI Sheep summoned as well, at exactly the same position, the command will find the player and the sheep, but not the stand.
Notes:
- A range of 1 or more results in correct behavior of commands.
- Non-Marker stands can be selected with r=0
- A possible cause of this may be the way entities in range are checked in the code; As the Marker ArmorStand has no hitbox, selection by range may be inherently buggy due to the calculation method.
confirmed for windows 7 as well
Heavy editing due to further testing; Confirmed for 15w45a.
Confirmed for 15w52a.
...15w52a?
Oh, sorry, meant 51b.
You are aware 16w02a is out now? We can only confirm for that version currently.
Yes, but my laptop was offline while I wrote this on my phone. I just tried it in 16w02a and it's still there.
I'll run a test on 16w03a in a sec.
EDIT: Confirmed for 16w03a.
Confirmed for 16w04a.
Ozone Smith: You don't need to comment you can confirm for a newer version, you can update the report yourself.
I know that, but it's a nice visual log.
putting down the version only already is a log of the versions + "added affected version" also goes in the activity feed + you have a history tab about the comments section that also records this
Apparently, r=0 is totally broken now.
Actually duplicate of MC-88533
This is still an issue even though r=0 now does r=0.25.
Under the new selector order, seems fixed.