What do you expect of me?

In null, @fatherbushido wrote:

What do you expect of me?

No expectations, as contribution is entirely voluntary!
We (as in team) do have to either fix, bandage or revert this thing though before releasing, as the game should really attempt to not segfault.
I'll take a look at how to heal this too as soon as I get to it.

In null, @elexis wrote:

In null, @fatherbushido wrote:

What do you expect of me?

No expectations, as contribution is entirely voluntary!
We (as in team) do have to either fix, bandage or revert this thing though before releasing, as the game should really attempt to not segfault.
I'll take a look at how to heal this too as soon as I get to it.

I can revert that, I can also add some check irrelevant to that commit, but it would perhaps hide other stuff :/ So it seems I should not adress straightly the (commit) concern, but I can if needed. (Else, I was looking today at the said bug).

What happens with the same action before that commit?

As mentioned in the ticket, it is not clear whether this commit is responsible for the segfault, it only triggers it. I just r

In null, @fatherbushido wrote:

What happens with the same action before that commit?

At least doesn't segfault and appears to work correctly (as in alpha 21).
There is also a command to revert a specific commit, svn merge -c -REV or something.

ok I cheked, it indeed occurs in that said revision.
The strange stuff is that we don't get the null cmpPlayer warning for the packing stuff.

If I don't understand until tommorow, I will just revert all that and the following patch.

See D143.
It sets things back to the previous state.
(Wich doesn't mean closing #4480)

rP19350 rP19423 rP19424 rP19429 rP19451

(The author of the patch (edit: and of this line) should have done the unit tests he promised to do and never completed.)
Regression : team bonus (at least for 'template' change) are messed with diplomacy changes (with the proper amount of players)

edit: in fact not only team bonus, but all template changes (triggered when there are different players affected by the auras)

if one would want to start its dev from this rev, here is the fix:

Index: binaries/data/mods/public/simulation/components/Auras.js
===================================================================
--- binaries/data/mods/public/simulation/components/Auras.js	(révision 19092)
+++ binaries/data/mods/public/simulation/components/Auras.js	(copie de travail)
@@ -201,8 +201,6 @@
 
 		if (this.IsGlobalAura(name))
 		{
-			for (let player of affectedPlayers)
-			{
 				this.ApplyTemplateBonus(name, affectedPlayers);
 				// When only Player class affected, we do not need a rangeQuery
 				// as applicable only to player entity and templates
@@ -214,8 +212,8 @@
 					this.ApplyBonus(name, playerEnts);
 				}
 				else
-					this.ApplyBonus(name, cmpRangeManager.GetEntitiesByPlayer(player));
-			}
+					affectedPlayers.forEach(player => this.ApplyBonus(name, cmpRangeManager.GetEntitiesByPlayer(player)));
+
 			continue;
 		}

(with wrong identation)