MC-101249Target selectors ignore invalid arguments
"@e[customName=Xephos]" works EXACTLY the same as "@e". This causes extreme problems, like when using the "kill" or "tp" command. This caused me to teleport all entities on my friend's realm to one spot. It will take a while to fix this.
What I expected: The command fails due to an invalid argument.
What happened: The command completely ignored the argument, and teleported every entity.
Environment
Win 10
Attachments1
Created Issue:
Severe target selector bug
"@e[customName=Xephos]" works EXACTLY the same as "@e". This causes extreme problems, like when using the "kill" or "tp" command. This caused me to teleport all entities on my friend's realm to one spot. It will take a while to fix this.
Environment
Win 10
Attachments1
![[Bot] Arisa](../../../data/assets/gravatar/mustek2.png){kind=small}
![[Mod] CubeTheThird](../../../data/assets/gravatar/cubethethird.jpeg){kind=small}
"@e [customName=Xephos] " works EXACTLY the same as "@e". This causes extreme problems, like when using the "kill" or "tp" command. This caused me to teleport all entities on my friend's realm to one spot. It will take a while to fix this.
What I expected: The command fails due to an invalid argument.
What happened: The command completely ignored the argument, and teleported every entity.
Severe target selectorbugTarget selectors ignore invalid arguments
![[Mojang] Grum (Erik Broes)](../../../data/assets/gravatar/grum.png){kind=small}
[Mojang] Grum (Erik Broes)- Put or summon at least two armor stands
- /say @e[type=armor_stand,argument=anyvalue,c=1] (where "anyvalue" can be even "") will find 1 armor stand (invalid argument was ignored, which is
MC-101249) - /say @e[type=armor_stand,=anyvalue,c=1] (no argument)
/say @e[type=armor_stand,anyvalue=anyvalue=anyvalue,c=1] (multiple equals signs)
/say @e[type=armor_stand,anyvalue,c=1] (no equals signs)
These commands will find all armor stands - this means that everything after an invalid entry was ignored
Please do not mark unreleased versions as affected.
You don't have access to them yet.
There is no "customName" argument. Simply use "name" when specifying a mob name.
@CubeTheThird
customName is an argument, but not for target selectors. It's a summon argument. It converts itself to "name" after the entity is summoned.
I would expect that any invalid argument would return as negative, but what happened is it returned as all entities. This is why I never use the kill command with @e, cause if I mistype one character, it kills all entities.
Using custom arguments is possible, so I see no reason that invalid arguments should be ignored. "customName" could be a custom argument, if it didn't exist already.
You realise this means that misspelled target selector data tags are ignored? If I did:
/kill @e[naem="foo"]
it would kill every entity. If I did:
/kill @e[name="foo"]
it would kill only entities named "foo". One letter difference and there goes all of that work. No entities have the tag "customName", but it runs as if they do. Why isn't that considered a bug? It's scary enough that I never use the @e selector, unless the glitch result wouldn't be that bad. This makes the @e tag pretty much useless to me.
Sorry if I'm bothering you.
Re-opening to allow the developers to decide what to do with this.
Still present in 16w39c
I also consider this to be a bug
WAI: https://bugs.mojang.com/browse/MC-106842?focusedCommentId=329385&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-329385
See [Mojang] Grum (Erik Broes)'s comment on
MC-106842:It is way too risky to use the @e selector with commands like kill and tp. I never use it. People with dyslexia might consider the @e selector unusable and therefore pointless, because one mistake ruins all entities. It would be easy to change the parser, and a lot safer.
If a command doesn't work as you want it to, and you don't realize that, would you want the command to fail and give an error message so that it doesn't mess up anything, and let you know you messed up, or would you rather mess up the world?
I'm sorry if this is wasting your time, but this is really important, at least to me.
Are you guys (including Grum) willing to reconsider if this is WAI or not? Or should I give up, even if I bring up an entirely new point?
Remind them again in a year or when the first 1.12 snapshot comes out or something like that.
This will now make the command fail telling you which argument caused it.