I just read this thread and because i had reworked in one of my last works for CF the attack stuff, i found some more problems. Because i don't know what of my work is not deleted, i don't know the status of the server code. But because my new push code is still in, i think the attack code is it too. At last, most monsters don't kill itself with own spells, even when it looks like. In the past monsters kill itself because - arrow/bullets are not syncronized (monster can step on them, etc - and then they get hit) - mutiple arch monsters have synchron problems - monsters aggro false objects - monsters can attack itself (physical) i had removed the last one - this was the killer number one. The bullet problem is, that a monster can drop a shot and then step on it - or get hit in any other way from it. In old code, in this case then the owner of the bullet object is new attacker - bad idea when its the own object. I had removed this, but the arrow/bullet object code is still not 100%. Iam sure i had not removed all the traps in the code - when you want look in it, insert some LOG() in the critical attack routines (there are only 3-4 of it which are triggered every attack/damage a object do). In my test, most times the monsters dies from side effects and not from her own spell damage. > > I agree with the boss point, but might not someone want to create a map > > where the moster is meant to kill itself sort of thing. IE a problem > > like the Avatar of the Fire God, but where instead of solving it that > > way, you have to wait for it to kill itself? (soz no spoiler here ;). > > The A o t F G kills itself? It shouldn't be able to.... Did someone > mess with it? > > But anywya, I hadn't considered a case where the mapmaker WANTED the > monster to kill itself. However, I'd prefer a map solution to this: > a map maker could put somethng else in the map to kill a monster > rather than requiring it to kill itself with its own spells. > > I do not think the complication of another flag is needed for > this case.... > > PM > _______________________________________________ > crossfire-devel mailing list > crossfire-devel at lists.real-time.com > https://mailman.real-time.com/mailman/listinfo/crossfire-devel