Printable View
I vote keeping it the new way. This way the player can target the token and I can give concealment. Previously a token was 25% visible to in the GM map view, but not visible to the player because of the old LOS calculation. I would have to drag the token out to make it visible, have the player target it, then move it back.
I agree that those literal "edge" cases can be too extreme, but overall I prefer the new implementation as it fits Pathfinder 2 (see everything that is not behind a solid barrier) and also more closely resembles Pathfinder 1 (lines from single edge to target edges).
For those edge cases in PF2 the GM can either keep the NPC invisible to the party or define the NPC as "hidden", even if said NPC did not actively try to hide.
So the PCs got a glimpse of a large creature behind a corner, they can (try to) target the creature, but chances to hit anything are low (plus added cover, which I would define as "greater cover").Quote:
A creature that’s hidden is only barely perceptible. You know what space a hidden creature occupies, but little else. Perhaps the creature just moved behind cover and successfully used the Hide action. Your target might be in a deep fogbank or behind a waterfall, where you can see some movement but can’t determine an exact location.
...
When targeting a hidden creature, before you roll to determine your effect, you must attempt a DC*11 flat check. If you fail, you don’t affect the creature, though the actions you used are still expended—as well as any spell slots, costs, and other resources. You remain flat-footed to the creature, whether you successfully target it or not.
Knowing the NPC being there leads to different tactics, of course. But anyone not actively hiding around the corner is detectable, that's how vision works. On one of our encounter maps the change just caused the party of 5 PCs to all detect all three NPCs, whereas before only 2 PCs could do so. It changes the dynamic, especially for spellcasters, but the NPCs still got cover and can try to hide in it once their initiative comes up (the PCs rolled higher Perception than the Stealth DC of the NPCs, so hiding failed at the start of combat).
PS to Players: Prepare your Magic Missiles! ;)
It was kind of mentioned above, but I wanted to get some visual representation. There are times where some tokens can see another token, but not vice versa. This is because the center of one token can see the corner of the other. But on the flipside, the center of the other token cannot see the corner of the 1st one. That is an issue with the new updated version. The included pics show what I mean.
I dont see that as an issue. The bird is on the corner the wizard is 5 feet back from the edge.Quote:
Originally Posted by Paperclipkiller
Again its the colour of the square thats giving away the token not the picture of the aarakocra.
I agree with Jiminimonka on this, especially after testing this in real-life on a corner of my apartment. When I stand in the center of a 5 ft. square that is 5 ft. back from an edge in my apartment then I can even see more of the 5 ft. square around the corner than what the Abjurer example is showing and I cannot see the other way around. The reason for the Abjurer seeing less than me in real-life is that the LoS lines do not align with the grid in the example, but with the protruding wall, else the Abjurer would see more of the Aras space.
To my surprise I even just found out that you cannot snap to grid LoS lines when you draw them.
Map LoS contributors need to watch out more how they draw LoS now, else something like this can happen:
https://i.imgur.com/6yzJ2AW.png
https://i.imgur.com/B5nGWvm.png
There is a projection on wall to the right side, still the PC can easier look around the right corner than the left corner. It should be the other side around.
Two inconveniences I noticed:
- Tokens not on the CT are not LoS visible even when set to "Mask visible".
- When tokens are moved without snap-to-grid (or via CTRL modifier) then LoS is only updated to half-grid during (mouse-)movement until the token is finally released. So only after being release the correct LoS is calculated from the center of the token instead of half-grid steps.
Some bugs I noticed:
- It happened repeatedly that a PC token added to the map from the CT did *not* reveal LoS to the player owning it (myself via localhost). So all the player got was a black window when LoS was enabled.
This happened with a map from a bought AP just the same as with one of the FG Battle Maps. I specifically loaded a test campaign with nothing (!) loaded but the single module including the map and it still happened.
"Revert changes" seems to have helped then. Next time it happens I will check simply removing the token and then adding it again might help, too.
- With different maps it happened that LoS lines were gone and inaccessible. Unlocking the map and opening the LoS tool displayed no lines at all then. Once I noticed that opening the Mask tool showed some crooked lines outside the house shown on the map. I think all pins were also missing.
Using "Revert changes" while the map was open made the crooked mask lines vanish, but and pins appeared at the lower right corner of the map (wrong places). I suspect this happened because I resized the map window to be smaller than in its original state. When I closed the map window and then revert changes again the map window could be opened at original size with all pins and LoS lines intact.
- Early at my testing it happened that my own GM controlled NPCs could not see each other even when they stood right in front of each other. What I mean is that PCs were revealed as being seen, but the other NPCs stays half-translucent even in plain sight.
- I recently tried an extension that allows to place spell tokens on a map. In one instance when I placed a 8x8 squares token on a map the token borders became LoS borders without any LoS lines showing up. Furthermore the PC that placed the token could not leave the token space anymore. Said extension does not (officially) meddle with LoS, so this should not have happened and usually does not happen (only this once yet).
Can we just get an official statement that at the moment the new LoS is reciprocal Any Corner to Any Corner as per 5e RAW, cause PaperClipKiller's example above still seems to show Centre to Any Corner which is less than ideal?
I tested this and LoS starts at the center of the token. You can test this yourself, draw a LoS box aligning with the grid, place the player token on the grid-line instead of the center. LoS will go straight from the center of the token along the grid/LoS line and then still straight where the box ends instead of going a bit around.
I wonder if this is a bug or intentional?Quote:
Originally Posted by Weissrolf
As Moon Wizard wrote, the LoS is just about the corners/center of the occupied square, the token image does not matter for that :) (so, you're right, but as an official emphasis :D)Quote:
Originally Posted by Jiminimonka
About the visibility thing: I like it :) Yes, maybe it is not RAW, but I prefer it that way, and I personally will use FG's LoS "rule" rather than RAW :) (and let's be honest, who of you was rigorously checking LoS when doing that stuff manually? :D I often did that very handwavy, so, for me no hurt when it is not always in alignment with RAW, especially when it is just about such deviations :D) Since it is also not just about 5e, it might be the best fit for among all rulesets :) (but when there can be an option, that would be nice of course)
Moon Wizard already wrote here, that it takes the center of the "seeing token" and checks the visibility of corners of the other tokens :) (or I completely confuse all the statements here now)Quote:
Originally Posted by Ludd_G