Quote:
Originally Posted by Hezzy002
Why don't you run the hit detection/movement simulation with two NPCs while simultaneously using the same input with the default system and see if they go out of sync at any point? That'd be really cool to check out and it would give you definitive results.
|
I've already debugged the 2 strenuously, for one example in the form of an offline and online NPC showing a 32 x 32 pixel polygon to represent boxes, and concluded that the basic hit mechanisms are like-for-like. I've even shown this to complainers and they've either become convinced, or rejected it with some belief that player damage checks are scripted differently to showcharacter() ones.
Quote:
Originally Posted by Imperialistic
Don't you think if we knew exactly what the problem was we would tell him? I mean we obviously want to see it fixed here, and we really don't have the privileges to check out scripts and certain things to troubleshoot.. and that is why we were trying to provide 'symptoms'.
|
Certain people having this gut feeling that there's a problem but without being able to explain it isn't a good enough justification, such gut feelings have an unreliable tendency and often contradict the gut feelings of other people.
A short while ago, one of the most experienced sparrers on Graal said something like "this hit detection is better than UN's", which didn't make any sense, as to my best knowledge Classic's system is emulating default.
More recently, during the lag problems, they changed their mind to "this hit detection is worse than UN's", and were quite vehement in the suggestion that I must have changed the hit detection, when the system has been the same for years.
Quote:
Originally Posted by Imperialistic
What position is Thor in? I think you're over-thinking a little too much Chris, I'm merely just trying to get a point across about a symptom.. not to bash a very helpful developer.
|
The problem with this alleged symptom is that it is perfectly normal behaviour for the default hit detection, and is therefore expected to occur within our scripted system. The suggestion that it does not occur eleswhere simply isn't true (see video), and even if it was, could easily be explained by circumstances such as the recent server lag problems, or being hosted on an inferior machine, rather than the scripting being off.
Unless someone were to provide a
measurable suggestion, or I were to see all of the related client code, experimenting with changes would be like chasing ghosts as Chris said.