Combat Ranges

[Jakkal]

Can you change just the model's height in game to test it?

[John Adams]

Nope. drawScale scales the set model. What I might try is putting both at a 1.00 size and seeing how the client UI changes. We'll get it, for now you'll just have to hug your big mobs as you are murdering them.

[zippyzee]

However it works, I would expect in theory to get range = getDistance(player, target) - .5 * target_width;. How you determine target width is one question, but it does follow the same path you are seeing that the width of the object would probably be the priority for targeting.

And that code you posted was not by me, so I am not familiar with it.

[John Adams]

Yeah zippy, the conclusion I came to yesterday is that not all drawScale's are created equal, but there is definitely a missing element to the puzzle that the client uses which we are not, yet. We'll get it. Maybe Xinux or Ratief can take another pass at the ASM since they've now spent another year in it.

[Amnath]

Played with it didn't you?

I now get a range of 5 from average humanoids that's about twice, visually, what it should be.

May not be the final tweak, but, at least you're able to affect it.

[Jakkal]

No, the distance is correct on the compass. That's the way it was in the live game. I just don't think people noticed how far away it was.

[Amnath]

Hmm it still seems pretty far, so does 10 for parley range.

Are you saying the compass range is built in and correct, and the problem is re-creating that same value for the combat engine?

One thing that is true, is that say you are at a range of 3, it should take 2-3 individual walking steps to back out to 4. But, when your distance ticks, if you then move forward a step, it will not drop back to 3, it will again take multiple steps. That is how it worked on live.

Currently I noticed you have to take 4-5 steps to get from 0 to 1. Maybe this is accurate, but close range seems a little "stretchy".

[Jakkal]

The compass measure hasn't changed at all. There is zero difference between Live and what you see on New Telon. I have screenshots of my character "5" away from regular human NPCs, it's staggeringly far. This is why we think the distance is measured in meters. Most people stand 1-2 away from NPCs when they hail them, but think they're closer than that.

So unless something is specifically wrong with your client, what you see on NT is exactly the same as Sunset VG.

Here is a 'for instance'. I used to parley in Leth Nurae for all my characters to quickly gain levels to 45 to do my POTA gear faction parleys. I would stand on this table in the center. From the center, I could hit each and every one of the NPC's and parley them in a circle (Though I often just parleyed between two using the diplo 'trick' so I wouldn't have to wait for a cooldown).

You can see that I'm well within parley range for all of them. This is on New Telon.

[John Adams]

[quote="Amnath"]Are you saying the compass range is built in and correct, and the problem is re-creating that same value for the combat engine?[/quote]
Yes. In all things VGOEmulator, the "client" is the authority, as we cannot change it. It is very likely VGOEmulator will not mimic live in this respect, unless we inadvertently stumble upon a math genius who worked for Sigil who can simply say "do x+y+z, boom".

This only really became obvious as I am building spells and seeing inconsistencies in spell cast ranges (in C++ code, not client. client is correct). And reflecting, the many bug reports of being on top of a small NPC and still being "too far away", which is when zippy gave us our first glimpse at an offset (the 400.0f)

[Amnath]

I took this into consideration. From what I could tell, the range finder might be automatic, but it isn't correct. I put myself on a true square meridian so only my y was changing, and I could, for example, at a range of 2, be at coords that were between 0 and 1. You can move from 7 to 9.

The range finder may be changing its measurement depending on whether you go forwards or backwards, maybe it is switching between CM and hitbox, I dunno. But I think only the coordinates are accurate and it does look like 100 coords = 1 distance unit.

Even though 0 equals 0 to .99 distance, you don't know where 0 is, but at 1 you are 1, or 100 coords, away from it. So starting from center of mass, I moved from 1 on one side, to 1 on the other. So, less 200 units of "range", what's left is his "zero area". On a normal sized mob, their "zero" is about 200 units wide, where for a larger one (in this case, Aluul, at RI) is about 500 units wide. Since 0 equals 0 to 99 coords, 200 for normal sized width sounds like 1:1 scale. Aluul would be 2.5:1.

So that's basically just finding zero by dead reckoning.