SafeTargeting does not always/properly function

[fuzzlecutter]

A common issue I run into is my character will engage an npc that is already fighting with another player.

Would it be possible to check for this situation when pulling and when assisting so as not to annoy other players? Maybe it is possible to check the target of target in this situation? Or a possible fallback if target of target isn't supported by the server?

[AlgarDude]

We should have a setting for this already.
It is a safety check.

Please provide a lot more detail about when it fails. Thanks!

Focusing on steps to reproduce the issue is best.

[DerpleDude]

if Config:GetSetting('SafeTargeting') and Targeting.IsSpawnFightingStranger(target, 100) then
    Logger.log_verbose("\ayOkToEngage check for %s(ID: %d) - Fighting Stranger --> Not Engaging", targetName, targetId)
    return false
end

[fuzzlecutter]

Hmm.. I have SafeTargeting enabled, and I've yet to not attack another nearby player's mob. Maybe something in Targeting.IsSpawnFightingStranger isn't working on my emu server. I'll have to investigate further.

[AlgarDude]

If I have time this week I will do some limited testing on The server I play on to make sure it isn't something on our end

…

On Mon, Jan 19, 2026, 12:45 fuzzlecutter ***@***.***> wrote: *fuzzlecutter* left a comment (DerpleDude/rgmercs#966) <#966 (comment)> Hmm.. I have SafeTargeting enabled, and I've yet to *not* attack another nearby player's mob. Maybe something in Targeting.IsSpawnFightingStranger isn't working on my emu server. I'll have to investigate further. — Reply to this email directly, view it on GitHub <#966 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AKJIGPQJJZV5EERQB6HEZDD4HUJ5BAVCNFSM6AAAAACSFW6X6KVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTONRZGUZDMNZSGI> . You are receiving this because you commented.Message ID: ***@***.***>

[AlgarDude]

It appears there may be a bug or gap in the logic here, will examine further

[AlgarDude]

Derple and I have spent a couple hours playing around with this... there is actually an issue with the TLO and data being returned. We will revisit soon to try to figure out an alternate method to get the exact desired behavior, but I do not think it will be easy.

[fuzzlecutter]

Derple and I have spent a couple hours playing around with this... there is actually an issue with the TLO and data being returned. We will revisit soon to try to figure out an alternate method to get the exact desired behavior, but I do not think it will be easy.

Sounds like perhaps a fix to the TLO implementation in macroquest is appropriate here, rather than workarounds in rgmercs. I don't know the details, but something to consider.