https://lua-api.factorio.com/latest/classes/LuaCustomChartTag.html#last_user
merging chests seems to lose last user, which causes https://mods.factorio.com/mod/brave-new-oarc to destroy the item.

I don't see a problem with implementing this. Just one question...who should be the last user when I'm merging two chests, one placed by other player and one by someone else else? Me (because I did the merging action) or majority player of merged chests? Does it matter for brave-new-oarc?

I would leave a null in this case

Wait, so what is the problem with how current merging works (I assume it also leaves the last_user set to null)?

you are correct... the problem is last_user set to nil on a merge.

I think the vast majority of users merge their own chests, so they should just be set to whatever the last user is. I suggest that if it is a mixed set of last_user, then it should not be something to worry about, and either choose one, or set nil as is convenient.

But I'll defer to your expertise.

Ok, I just don't understand why is it necessary to set last_user if it would be ok to sometimes set it to null. Do you know what is it used for in brave-new-oarc?

You make a good point.

In BNO, if regrowth is set, it will destroy items slowly if they have no last_user.

BNO also uses last_user to secure items from multiplayer griefing, I believe.

As setting another users chests to your seems problematic, I, in relative ignorance, suggest it might be abusable. However setting them when they are unowned, or explicitly the users, seems safe.

I say unowned, because Ghosts on Water also seems to have the issue, in that items placed by it may not have last_user set.

Ok, I didn't really think about implications of multiplayer because they are like 0.1% of Factorio gameplay. But I'll add setting last_user to player which merges/splits the chest.

Implemented in 5.3.0