Some TOG's mobile puzzles

[TheOnyxGuy]

Well, I see how people on this forum loved to make more and more puzzles. I think I have a choice: make some own puzzles when I'm not able to work on needed levelset or just forget it and play another Android game. I chose first, so. Here we go!
Sorry that I didn't use standard game elements' icons from Level Sketch.


Supapuzzle_0001
Infotrons needed: 11


Emeraldpuzzle_0001
Jewels needed: 13

[TheOnyxGuy]

Here's a little puzzle with trickery,
This is falling up from mistery,
If you wanna play it like a number one,
You must to know what's up with all these ports at all.


Infotrons needed: 10

[ncrecc]

Well, I see how people on this forum loved to make more and more puzzles. I think I have a choice: make some own puzzles when I'm not able to work on needed levelset or just forget it and play another Android game. I chose first, so. Here we go!
Sorry that I didn't use standard game elements' icons from Level Sketch.


Supapuzzle_0001
Infotrons needed: 11


Emeraldpuzzle_0001
Jewels needed: 13

do you still have these levels? the images are dead

[TheOnyxGuy]

At first I thought I wouldn't be able to find them because I completely forgot what these even are after so much time, but fortunately, the still persistent links still helped me find where the images are being linked from. They're from russian social site "VK" (aka VKontakte), which I guess at one moment stopped supporting direct image links for other sites. Anyways, fortunately for me, I hadn't removed the original images and I can reupload them here just fine.

N9XBYFr0bIO4dCjt7x0Q69DcOqiLXW-u000d3WnyIRvItJ9VzuxJjp3Gyu0ja7uY8tZmKw.jpg (382.1 KiB) Viewed 234 times

Emeraldpuzzle_0001
Jewels needed: 13

JFGMxp195aUlM8DGGgZNUrjsO57Ph0FTfBHIi_8ZXCmcdz41FNKhLKF6AuyUAlCyA97vPg.jpg (298.38 KiB) Viewed 234 times

Supapuzzle_0001
Infotrons needed: 11

[filbo]

Unlike the other recently discussed puzzle (viewtopic.php?t=2572) -- these three are presented as images, not backtick-inlined RnD elements.

@TheOnyxGuy, do you have them as saved RnD level files? Could you upload in either that form, or the backtick inlined format?

Otherwise, every viewer has to start by re-creating the levels themselves in the editor (which I'm sure is super easy to get wrong...)

=====

Or, well, alternatively, has anyone else gone through the reconstruction process and can upload these as usable level data?! :)

[filbo]

(or maybe that software recently used to mine out the Kriszz levels could be applied...?!?)

[ncrecc]

Otherwise, every viewer has to start by re-creating the levels themselves in the editor (which I'm sure is super easy to get wrong...)

Or, well, alternatively, has anyone else gone through the reconstruction process and can upload these as usable level data?!

(or maybe that software recently used to mine out the Kriszz levels could be applied...?!?)

here is my attempt at decoding them with an experimental "natural intelligence" algorithm:













(13 gems)















(11 infotrons)























(10 infotrons)

i haven't beaten any of them so i can't attest that they're 100% accurate, except the last one which i compared to the original in an image editor

[BrownSky]

Looks good.

Would you share the "natural intelligence" algorithm?

I think that whatever visual parser is used, be it the "natural intelligence" algorithm or the Krissz approach, it would have a much easier time parsing levels from Supaplex, where there are not (I think) different forms of monsters based on different initial direction, than levels containing elements which do - which is all the other sorts of levels. On the other hand, the Krissz approach must be able to handle that, given it is working with an engine that does have different forms of monsters based on different initial direction.

John

[ncrecc]

Would you share the "natural intelligence" algorithm?

I think that whatever visual parser is used[...] it would have a much easier time parsing levels from Supaplex, where there are not (I think) different forms of monsters based on different initial direction

that should be irrelevant since, as far as i'm aware, monsters facing in different directions all have different appearances in the editor (even yams will appear to be looking in the direction they're set to move) -- except the "random direction" versions of monsters which i believe look the same as the upward version of the monster. maybe there could have a question mark overlay like how the GDash graphics set puts arrows over monsters facing in different directions in the editor