User talk:Ashghan

Boiled River Pearl Mussel

The suggested formula hunch seemed just incorrect to me. So I reverted it. I don't mind if you add your idea's to the discussion page(s), but preferably with some additional info and data that lead you to a given conclusion. Also added a additional quality related link, in case you still like to dig into the subject. --.MvGulik. 18:06, 14 May 2015 (EDT)

"I'm putting in the formula found on the cauldron page."

I have no real objections to that. Although maybe a additional link to the cauldron page in the Boiled..mussel Quality section suffices. --.MvGulik. 11:37, 15 May 2015 (EDT)

"I assume it's correct."

I have not see any objection to it, so I think its relative save to assume that. --.MvGulik. 11:37, 15 May 2015 (EDT)

Thanks and sorry for the formula mix-up. I thought having the formula on the Boiled Mussel page would be nice, since not everyone would think to check the cauldron page for it. Ashghan (talk) 05:47, 19 May 2015 (EDT)

Fish node note

""Fish node note - the jumping fish animation reduces when the node is reduced. Probably to match the reduced radius that fish can be caught. In a q56 node fish were jumping all over 4 map blocks. With q reduced to 15 the only jump in about 10 square radius from the centre. This did not happen instantly though (noticed it the next day).""

Nice. I had heard a similar thing, but never managed to spot it myself. Considering this confirmed. PS: Just to be sure, are those 4 blocks(minimaps?) from q56 a diameter or radius measurement? --.MvGulik. 02:57, 5 June 2015 (EDT)

Unfortunately I didn't actually measure it at the time, I might do it once the node refreshes (though it will be a few weeks to reach 56 again). By 4 blocks I meant the ones that actually share the corner with the node. So it'd be around 100 squares radius from the node center. It may be difficult to measure exactly, due to the fact that the fish jump animations are pretty far in between. I might look at the furthest one and do a measure from it. I will post once I am able to take measurements. If not - remind me. ;) Ashghan (talk) 07:28, 5 June 2015 (EDT)

Aha, around 100 squares/tiles radius makes sense, as that would fit the general known node features.

Tip: Not sure if this can, or how, be duplicated with other clients. But with the Anemone client, if you switch to a session showing a fish node. It shows all the jumping fish-animations that where send while that session was inactive. Which if a nice way, when screenshooted, to get some easy range and generally distribution of the fish-animations. (Do take note of the fact that the characters position might also play a role here. I'n not to sure about this. But if it is, it would skew the data. Showing up as inconsistent data between significant different character positions.). --.MvGulik. 00:31, 6 June 2015 (EDT)

Character position does play a role in that the animations are limited to your view range, which is 44 squares. For a good measurement of a node of such high quality (ie. range) I'd most likely have to do a few screenshots around the entire circumference and then somehow paste all the images together. I do have Anemone, so I'll look into this animation display, didn't even know about that feature. Ashghan (talk) 04:25, 6 June 2015 (EDT)

Ok, did the preliminary research. Node q is 58, no adjacent fish nodes. The fish animation reached roughly 81-83 straight line squares from the center (hard to judge since the animations are not inline with server lines. First fish (q10) was caught 83 squares from the center (sqr 84 produced no results) - this would show, that the q falloff is roughly 0,66 points per square. The animation range seems to be equal in all directions, diagonally animation was measured as a corner of a 58x58 squares square, so it'd be around range 82 with the formula of a square's diagonal being 58 * sqrt 2. Conclusions - animations roughly mark the usable edge of the node and the distance from the furthest animation to the center can be used to gauge the nodes current q. Ashghan (talk) 17:29, 20 June 2015 (EDT)

Fish node q36, animation range 70, first fish caught at range 74. Seems the range vs. quality calculation is not that simple. Ashghan (talk) 12:08, 21 June 2015 (EDT)

Figured out (or so I think) the range/quality thing. Each square away from the center is a 1% reduction in quality. As long as quality is 10 or higher - you get fish. The exact calculation might be in map units (1/11th of a tile) so the numbers may be a little off in some cases, but rough estimates (+/-2 points are possible).Ashghan (talk) 14:49, 22 June 2015 (EDT)

Fish node quality dropoff

For your [44..26] data, the general drop-off factor seems to be 0.9795. Those doubles are natural. The lower the value you multiply with 0.9795, the smaller the change. As you can see in your [17..15] range (lower value, more dupes). misinterpreted that [17..15] range as being from the same spot, but I think it still would if it was.

For quality dropoff based on distance. I'm pretty sure its using Bilinear_interpolation.

Experimenting with your 36q fish node data. With bilinear_interpolation along the edge (which behaves close to linear btw) I get that the default node value for the fish node-map seems to be 1 (or near 1), instead of the expected zero. :)

For node point values: [36, 1, 1, 1]
[[Coord(x,y)], Q-result]
[[73.5, 0.5], 10.228625] <- farthest expected tile(74*) to still give fish.
[[74.0, 0.5], 10.0545]
[[74.5, 0.5], 9.8803749]

For node point values: [58, 1, 1, 1]
[[83.5, 0.5], 10.357975] <- farthest expected tile(84*) to still give fish.
[[84.0, 0.5], 10.0744]
[[84.5, 0.5], 9.790825]

*) When counting tiles as 1..100.

Note: The X.5 point-values are at the center of the tiles.
Although the character fishing graphics shows different fishing points on a single tile. I think there just graphic variations, and the fish is virtual pulled from the tile center-point (just like with other static resources, except for water of course.)
PS: If your Python capable, I can send you some bilinear_interpolation code.
PS2: When fishing, it was possible that the node degraded even when you did not catch a fish due to low fish-Q. It still might. Although with the last Dev's node adjustment, that might have changed.
--.MvGulik. 15:20, 28 June 2015 (EDT)