What's new

possible ebolt/spell exploit for pking? (line of site issue)

#1
detailed post on grid spell radius issue:
http://forums.uooutlands.com/index....-for-pking-line-of-site-issue.1864/post-11604


suggested formula (distance between points on 2d grid):
http://www.math4ged.com/finding-the-distance-between-two-points-on-a-coordinate-grid/

original post:
A couple times (20 ms ping) I have been pked when down to low health after escaping from ebolt 2+ screens away.

I play at a higher resolution in classicUO, and the pks are often 30-60 tiles away when this happens.

is there possibly a line of site loop whole occuring or unapproved 3rd party client? I have seen a couple other similar posts.

I know my archery and barding would not work at this range.

It usually occurs after they got 1-2 spells off on me which made me think it was a set/last targer type los issue once escape. I even saw an ebolt flying from off screen (at least 20-40 tiles away), I was almost to town .

Thanks.
 
Last edited:
#2
I had this happen yesterday. I got ebolted from off screen to a pk. I have the death screenshot and you can't even see the other player on the screen. I thought for sure I was free because I avoided his spells and ducked behind some houses.
 
#3
ya hard to tell if intentional (unsupported client or a client that is supported bypassing los...steam, razor or classicuo)

i will try to recreate it but if anyone else can that would be cool. a couple ideas are that it could be "last target" macros after casting and need los check.

i have seen a few monsters hit me off screen out of range of my bard and archer with spells as well which makes me think it may be more of an los issue ( line of site) or range check. thanks for your feedback.
 
#5
Monster LOS check is before they cast, when they target you. That's the main reason so many people complain about monsters hitting them with spells from screens away.

Player LOS check should be after they cast when target you.
 
#7
Monster LOS check is before they cast, when they target you. That's the main reason so many people complain about monsters hitting them with spells from screens away.

Player LOS check should be after they cast when target you.
It is this - curious to know if you had mobs around you guys when you died a couple screens away, from the PK is it possible a mob targeted you as you ran by? Interested to see what comes of this though!
 
#9
same... nothing on screen i was on a road and horse. it could be something like "they hit u within range and u ran off on horse so it's going to homing missle you down like oldschool 1998 heavy xbow"
 
#10
I was not able to reproduce. I was able to confirm:
a. archery limit is 8 tiles + 1 (their tile)
b. spell casting: 12 tiles + 1 = 13 total
c. bard i think is also 8?


Sometimes monsters appear to get me past 12 tiles similar to how the pk got me (classic uo client)
 
Last edited:
#13
@Ace Mason @Owyn

I was able to figure this out with various resolutions (1080p and a little lower wide screen). The UO tile system is a Dungeons and Dragons inspired grid (1 inch squares). Spells on a grid function different than spells on a different map tile system (that would be linear / angle lines without tiles ).

These dnd tiles would be good reference:
(25 diameter is close to our 12 foot spell casting limit)

These graphics from DND dungeon masters give a good idea of distance in a tile based system, primarily in the "center red 25' circle", these is (1" = 5 feet dnd standard... not 1 tile match to UO so imagination time). If you imagine the caster is in the middle they could cast linear 12 tiles. However when you cast on an angle, your still casting "12 tiles" but its not 12 "kiddy corners".

What is happening here is casters are able to macro and cast 12+ tiles away even when your off screen (monsters and pks alike). Especially if they are directly "above or below you" where the angles/pixels do not match up.

So on to step 2.

What is occurring:
a. "12 diagonal sections of squares as a line is actually 16+ tiles ... the diameter is not attached to the tile grid and therefore is not truely a cirlce/diameter but more of a diamond (where the width of your screen is not going to show the extended angles)

b. what happens is the player with resolution < 1080p (1920x1080.. .basically < 1080 hight I'm not sure the specifics probably in 900s) cannot see the 12 "extra long" range from direct angles (up/down) since these "extra long" angles and not the equal length of shooting in a parallel line. a macro would still be able to target the player on the "Extra long angles" even though the player is not on screen.

c. the other player 1920x1080 can still see the player and kill them with a macro or selecting them via scroll bar.

d. the last screen shows the "grid in action". The yellow lines are the original 12 tile max distance for spell casting. It was then copy/pasted (a triangle with two equal side lengths) along the "red too long line", that is the true angle of the circle. The distance up/down is closer to 16-18 tiles and not 12 and I was able to cast lightning while the other account could not see me for bug testing.

To reproduce "Grids" turn on the tile grid feature in dungeons (classic uo) and have team members stand at various angles starting withe the simple N, S, E, W straight lines before moving to the incorrect ranges of NW, SW, NE, SE .

screenshots next...
 

Attachments

Last edited:
#14
In this screenshot the caster (higher res) can target and hit the other member that cannot see them on their screen. even if the caster could not see them they could still target with a macro.

the victim NW in this case is on top of a "diamond pattern".. the angle is now 25-35% longer than n/s/e/w
2019-03-16 01_42_50-Start.png
 
Last edited:
#15
This is the max distance (in theory) of the linear angle (12 tiles)... as we will see in the next screenshot I walk two steps west and should be outside of range but still get hit.

12 tiles straight line.png
 
Last edited:
#16
In this example I walk past the max distance (from north, aka the sign by walking west/left of the sign and still am target... although im closer to 13-14 tiles at this point)


in theory this shouldn't be possible unless i walked two tiles west.

12 tiles then 2 west.png
 
Last edited:
#17
two angles at the same time. The defender (top of screen) cannot see the attacker who is roughly 16-18 tiles away to the SE. This is most likely due to a "Diagnoal angle" "12 chunks of lines/tiles" that is not held to the rules of radius on a grid based system... the diagonal ends up creating a "Diamond radius" instead of a mockup circle radius (angles like SE, NW, SW, NE etc should be shorter on a grid based system and not 4-6 tiles longer)


2019-03-16 01_42_32-Start.png
2019-03-16 01_44_28-Start.png
 
Last edited:
#18
This is a diagram of "lines/angles/diameter/radius on a tile/grid based system". Right now we have a "diamond" instead of a "circle attached to a tile/grid system" which should look more like a funny umbrella shape rather than "Straight lines" (as in the dnd pics). Right now if you are NE, SE, NW or SW of the user, you actually have a distance of 15-17 tiles (16?) and not 12. If you have a macro you can hit them off screen (So can monsters on angles) and you can hit them when they cant see you on screen (even if playing at a higher resolution)

players at the standard res 640x480, 1024x768 will not be able to see these casters at all.

Right now the shapes are a "Diamond" and not a circle mapped to a tile/grid system which is a common DND "snap spells to a grid" complex math issue. Below is the details.


linear angles (north, south, east west) have the 12 tile spell range. However NE, SE, NW, SW have the incorrect distance and are closer to 16-17 tiles.. roughly 25-30% longer linear angles. the distance for NE, SE, NW, SW needs to look more like a "sun/umbrella of similar triangles with two equal side lengths" than a straight linear (obtuse triangle with three different lenghts) angle due to the tile/grid based system and how radius/diameter are applied to a two dimensional space (grid).


2019-03-16 02_36_38-Cortana.png



DND reference: 1 tile = 1" = 5 feet in game



another dnd example for radius/diameter or "any linear angle" on a grid. on Outlands (if you flipped this 23 degrees to match UO's tilted compas)... you would actually be hitting the targets "outside the circle" in the white space at a longer angle (16-18 tiles) than you would a direct path (north/south/east/west). A person attacking NE, SE, NW, SW will always have an advantage in terms of distance in spell casting . factoring in screen resolution (primarily height is always short) this becomes most problematic NW and SE where the shorter monitor height will cut off the extra 4-6 tiles. from sw and ne you can still be casted on out of range but you can usually see them due to a higher screen res.

Circle-Radius100ft.png
 
Last edited:
#19
I've read through all of this, very informative.. What I don't understand is, do you feel it's a bug and needs to be fixed, so you can no longer cast on people you can touch if they are diagonal to you?

I hate to be the one to say this has always been the way UO is - there's a reason people recommend running in as straight of lines as you can, and why town fighting is much different, because you have to turn corners, thus giving people a chance to reach you with spells.. I don't think this is a bad thing?
 
#20
could always be that way.. but if los for heavy for xbow was changed to 8 tiles and spell casting to 12 i thought this would be helpful

if this is 100% intentional (odd because richard garriot was a huge dungeons and dragons nerd and this is one of the quirkiest things about dnd with range/radius on a 2d grid surface )

if this is what it always is (i do not recall ever being casted on 16-18 tiles away as described above) the workaround is to not play wide screen.

instead play more vertical/tall screen because the range is diamond shaped (and tilted like the compas). you can then better catch what is targeting you of screen. example: 1920*1080... roughly 1080 lets you see the 16-18 tiles, which really seems off when 1024*768 was l33t in 1998 on CRT .

currently the distance on kiddy corners is not calculated logically correctly (it may be right as you stated as in "the way its always been" ... however from a spell radius as described by dnd 5e neck beards .. the info i provided could be super handy)

thanks for reviewing.