Talk:The Plain of Atlantis

Thank you very much

Four and a half years ago, Nintendo closed its WFC doors. And yet authors always make us, the community, happy with such beautiful tracks that still work. The last surprising ones were the tracks New Moon Manor, The Plain of Atlantis and Warp Pipe Island. Many thanks to you, Jiyuu, for your great tracks! Keep it up!!

-- Wiimm (talk) 19:45, 19 November 2018 (UTC)

Oh thank you very much for this wonderful message ! That's really so sweet ! I truely appreciate it and of course I'm going to continue to create new tracks~ \(^ω^)/

--Jiyuu (talk) 21:02, 19 November 2018 (UTC)

Glitch report

Greetings. We tested the track yesterday and found some issues that you should be aware of.

• On the part with the trickable pillars on the road, one pillar has a wall effect: video
• If you are having a mini-turbo before the entrance of the cave, or a fast vehicle, you hit the ceiling sometimes: video
• The last cannon sometimes sends players the wrong way around, a 180 degree spin: video
• Throwable items in a majority of the track disappear immediately. This problem occurs, when a track is too big and the item is thrown in an area with negative measurement values (it's hard to explain for me, I had the same problem in Cyberstate and Wiimm fixed it for me).
• Position Jump Bug in the spiral before the cave and also in the ice cave/crystal section.
• The ice cave is nearly unplayable for some outside-drifters. Maybe make it normal road instead of slippery?

--Buschkling (talk) 10:53, 20 November 2018 (UTC)

About item bug and positions

I'm sorry to say this (please don't kill the carrier): Items work online only, if they are dropped/thrown in the position range from -131071 to +131071 for all 3 axis (and y must be >0; and never tested larger y). This is because of the limited online protocol (131072 == 2^17, and the exponent of this number is too large).

The x-coordinate are OK, but the y coordinates (for ITPT) are 17850 .. 175955 and the z coordinates are -152526 .. 196184. This can't be fixed by simply shifting the track into other coordinates. A re-design is necessary (Really a pity!!).

-- Wiimm (talk) 13:14, 20 November 2018 (UTC) :)

Idea: Rotate the track by 45 degrees and center it (move it into the range of ±131071). Then reduce the height of the upper plane. The will test the first point tomorrow. -- Wiimm (talk) 13:28, 20 November 2018 (UTC)

I see :'c I was actually expecting something similar to happen, but not for the items. So when I tested the track offline, I thought it would work similarily online. But seemingly not... :(
Is this a problem related to the server, or how the MKW system codes the online game ? If it's a code matter, maybe it would be possible to create a code that uses more bits to store the coordinates of the items ? (which would obviousely allow much larger floats and fix the problem)

--Jiyuu (talk) 13:39, 20 November 2018 (UTC)

About the rotating idea : I'll try to do this. But I'll also try to calculate what is the volume of the possible drop/throw area for items to see if I can shift the track to fit x and z. For Y I'll try to find another solution :(

--Jiyuu (talk) 13:39, 20 November 2018 (UTC)

I have calculated it for horizontal (X+Z): Without rotating, a scale down to ~65% is needed. With rotating by 45 degrees, a scale down to ~80% should work. Anyway, the small caves becomes too small. (all calculations based on ITPT)

I used: --hrot 45 --scale 0.8 --translate -39565,0,10366

Question: There are 2 points of ITPT with Y=0: Really or bug? If bug, Y can be lowered.

And I have another idea, I'll tell later if I @home again.

-- Wiimm (talk) 14:06, 20 November 2018 (UTC)

Bug. But I don't think that I can lower Y enough so that it matches the limitation (the giant waterfall is the problem)

EDIT : I'll still try to lower Y once I'm at home and can load the project in the 3d editor. :D And thank you for the interest you have to help fixing the bug ! (^.^)

--Jiyuu (talk) 14:09, 20 November 2018 (UTC)

Moving the cave part only

Before i start the tests with rotating+scaling I'll tell you my other 2 ideas: They work without scaling the track: Move the cave part. Only the connection must be changed.

The first image shows the track from the top. Green is the overworld and red the underworld. Blue are the routes. On the left side you see the current layout, and on the right side my idea.

The second image shows the map from the side. Here I moved the lower world below the upper world.

In both models you have to lower the upperworld below 120000 and to redesign the connections.

-- Wiimm (talk) 15:52, 20 November 2018 (UTC)

If using the second idea, you can move the downworld ±X, so that the minimap shows 4 different lanes. -- Wiimm (talk) 16:34, 20 November 2018 (UTC)

That are nice ideas ! Thank you for the your images :D As I said, I'll first try my idea, then try to combine your both ideas to try the make everything match ! I still want to keep the layout the closest as it is now. But for sure what you said is useful. Thank you !!

--Jiyuu (talk) 20:25, 20 November 2018 (UTC)

Fixing the position bug

We are on the way to fix the position bug, so that this track be used online. Therefor I have to edit the track to add an additional info (LEX file). I want to release it as v1.1. Ok? -- Wiimm (talk) 18:52, 2 February 2020 (UTC)

I can also make the cannons faster; factor 3-5 seems to be good -- Wiimm (talk) 16:33, 3 February 2020 (UTC)

That's good news ! You can release it as v1.1 then ! Also, can you please tell me what you exactly do to fix the bug ? So that I can keep your LEX file if I ever update the track. Thank you ! --Jiyuu (talk) 00:28, 4 February 2020 (UTC)

My only edit is the addition of LEX file course.lex. It contains:

	# Online, items are only transmitted correctly, if the positions of all 3 axis
	# are between -131071 and +131071. The following vector defines 3 factors to
	# increase the range at the expense of accuracy. Factors <1.0 are set to 1.0.

	@ITEM-POS-FACTOR = v( 1.000, 1.600, 1.800 )

	.....

	#CANN#
	#---------------------------------------------------
	#                          Deceleration    End
	# Idx    Speed     Height     Factor   Deceleration
	#---------------------------------------------------
	   0   2000.000      0.000   6000.000     -1.000

It works only for distributions that will use the coming update of the LE-CODE executor. The executor transforms the position on sending and receiving.

Then you have (accidental) removed my comment about the cannon. I try 4x speed.

-- Wiimm (talk) 06:30, 4 February 2020 (UTC)

Update released. It will run only with the coming LE-CODE, which is included in the Intermezzo. I will release Intermezzo in some hours. So you can test it by yourself. -- Wiimm (talk) 19:12, 4 February 2020 (UTC)