A study on oak (UOAM oak map)

My story begins in the most beautiful day for some choppin'!



I set out into the woods with my trusty silver steed, and began to plot all the oak trees as I stumbled upon them. It didn't take long to notice that a pattern was emerging on my oak map.



I decided to whip up a quick program to produce a map file in the scale of all of Britannia, following this pattern. Having built the map based on my oak samples from Yew, I realized I was onto something when I tried this map in Trinsic and the first trees I tested there turned out to be oak! However there was a problem - this only seemed to work in a narrow line between Yew and Trinsic. If I wandered too far east or west from there, the map would become out of sync with reality. But once I'd found a single oak tree in the the region where this map was off, I found the same pattern worked there once it was re-aligned onto the found oak tree. Okay... so the patterns are aligned differently based on some arbitrary regions. A bit more chopping to see where all these region borders were, and I had my hands on a remarkably accurate map for (nearly) every oak tree in Britannia!



And now you can have it too! Save the UO:R Oak Map in your UOAM directory, rename the .txt file type to .map, enable it in Map->Additional settings->Files, and enjoy.

Thanks to @Fireball for helping me with the rather laborious effort of chopping innumerable trees to verify/falsify theories on patterns and regions! Also, this effort has been ongoing for quite a number of months... which also explains why his boats are so cheap

Here's roughly what your UOAM should look like when the map is in use:



It has about 10,5k oak trees marked out of which I estimate around 10k are valid matches. A few trees are missing from the map(mainly yew trees) and as mentioned there are a number of false positives. But if this map pins a tile that has a tree on it, it is very likely to be an actual oak tree. To my knowledge all false positives are tiles that don't actually have a tree on them. This is due to the fact that the map is generated from an image of the world map - all possible oak tiles are checked against a list of known hues for trees, but some of the hues are also used for flower tiles, road tiles (and a very unfortunate match with the hue of a common grass tile.. that hue had to be disabled because 99% of these matches were false positives) etc.

A more accurate map could be built if anyone would be willing to point me to documentation on how to read the actual UO map file to search for trees.

Also, @El Horno is definitely onto something with the UO:R Geological Survey project and his suspicion that it might be a repeating pattern. Keep up the good work!

*bows and begins chanting* "We're not worthy!"

It was a pleasure to be of assistance. I am glad that we have now been able to share it with our fellow Britannians.

FB

PS: Quality Yachts Inc. at Moonglow Atrium (NE of bank) and Yew at [UO] tower at south end of bank

/ShamelessPlug

I posted the pattern in @El Horno s thread last may.

Here it is again.

Awesome!

@Lightshade That's true, mate, but that only works on a particular north/south band. As soon as you step east or west of that band the pattern is lost and you have to search for the next oak to re-establish the new pattern. What Canis did, he did by mathematically analysing the map pixel colours to locate trees and then extrapolating the vertical bands and oak patterns per band to predict where every oak tree is on the map. The guy has an incredible mind and it was an honour and enlightening to work with him on it.

Some people like to think that they are ending the game buying/possessing a 5 million mask or sandal, a fort on grass...

THIS is ending the game. Like a genius.

Congrats, sir! Roll the credits.

No, actually, the pattern east to west is there, also. Ive been doing it for almost 2 years now. I cant remember the last time I bothered to chop a normal tree.

But that's not true.

Note the screenshot below. By your comment, the left hand tree should be oak because it is 2 moves of 8 west, 3 south from the last oak tree. Yet this tree is not oak because this is on the next westerly band of oak spawn where the 8x3 grid is shifted.

That's just North of Moonglow zoo if you care to check it, but you can check it anywhere you find a line of 8x3 oak trees. The line always stops and you have to search for the next batch of 8x3 trees. With Canis' map you don't need to bother with all that. The points are all marked. The whole map has been decoded for you. It is a gift from one of the good guys, just like you!




PS. I say it's a North/South band but actually it is closer to NW/SE but not quite.

WOW

Amazing. Thanks for doing this.

Double Wow.

Adding my thanks and congrats on completing it, not giving up on a project that seems to have been a massive effort.




Marjo Governess of Wispfelt Village

Westra on IRC

Let me try and illustrate @Fireball's point a bit further. Here's an example of the debugging output from the oak map generator, taken from south of Britain, right next to THB-82. The yellow and red markers are oak trees, the white ones depict "would-be" trees based on the pattern but they fall in the edge where that pattern stops working.



@Lightshade is absolutely right that the trees are indeed in the same 8/3/12 pattern that he has documented well before us, however it's only continuous within the region that's marked in one color. Looking at this example with 7 oak trees in a straight line in the center of the image, you'd expect the circled 8th tree to be oak according to the pattern but it isn't. If you map a large enough area you're bound to run into these discontinuations points. But when you cross that boundary to the other side and chop around randomly until you find one oak tree, you'll find the same pattern applies there too, it's just aligned a bit differently. I added the black line in the image to emphasize that. This applies worldwide, and the red/yellow world map in my first post shows where these boundaries are.

I was chopping around in Yew in search of these discontinuation points while Fireball did the same in Moonglow -- these two points are about as far apart east/west wise as possible while being in the same ballpark north/south wise. Well, it seemed that the width I had found in Yew for one pattern to remain continuous appeared to be quite exactly 49 times the distance to Fireball's Moonglow discontinuation point. So I divided the area into 49 equally wide regions assuming each region has the pattern continuously within itself but that the pattern is not continuous across these region boundaries. And after finding alignments for 8 neighboring regions, I found they started to repeat again from the beginning. So I fired up the generator to make a worldwide map that cycled these 8 different alignment options across each of the regions I'd divided the map into.

Then I ran from Yew to Minoc to try it out (I really did run instead of recalling, for dramatic effect). After screens & screens & screens of running, I stopped to test out the map and was thrilled to find that it was indeed tagging oak trees properly, although it was based on measurements in Yew and Moonglow and we'd interpolated the data for all the areas in between.

Fun times!

When i went to download this it sent me to a url with nothing but text, how do i save that to UOAM?

This is cool but remember.. With great shortcuts comes great boredom.
Play the game, don't beat it

Right click the link, select Save As from the drop down and save it into your UOAM folder.

Then run up UOAM, go to the Places drop down menu, select the Filter option. Click the Files tab, put a tick in the box for "uor-oak" and the spots will all light up on your map labelled "oak". Just take the tick out of the box again when you don't need to see it.

FB

Worked, ty

For some, beating the game is playing it! Different strokes for different folks

thank you, it helps alot!!!!