28

The background text that follows is in this case not part of the puzzle, nor fictitious. The puzzle, which is self-contained, consists entirely of the image below; I have composed it as a genuine (albeit poor) tribute to the work of a great artist.


Today (21 June) marks the eighteenth anniversary of the death of the great Australian Indigenous artist Clifford Possum Tjapaltjarri. He was one of the pioneers in the 1970s of the dot painting technique, now ubiquitous and widely associated with Australian Indigenous art. These works make heavy use of symbolism and other techniques to convey information, often relating to the land. Possessing an extraordinary artistic vision, Clifford Possum produced a great number of remarkable works. One of his masterpieces, Warlugulong, sold for $2.4 million in 2007, the fifth-highest price ever paid for an Australian art work at that time.

The following simple little puzzle is a tribute to this artist, composed in a style reminiscent of his artform. Can you find the meaning it contains?


Dot


Hint:

For any who may still be looking at this puzzle, the best way forward should be to use the numbers @DavidG. has found to guide you in your attempt to read the yellow and red circles.


Hint 2:

For any who may still be interested in this puzzle, note that the red and yellow dots essentially form a set of strings in a two-element alphabet. They may not make sense as written, but @DavidG. has essentially provided a corresponding set of strings that contain (in some order), the numbers from $1$ to the string-length... Note that some data may have to be discarded due to the significant constraints of the puzzle's construction forcing the inclusion of non-meaningful data.


Hint 3:

Just in case anyone is still interested in this puzzle (and because we got annoyingly close to a solution in the comments), here's what has been correctly deduced so far: @DavidG has (correctly) deduced a grid of numbers from $1$ to $9$. @Stiv has then read the red and yellow dots left-to-right using these numbers to produce a (correct) image. I have hinted that there is an obvious way to read this image (i.e. convert the binary sequences to characters, noting that the puzzle's stringent construction forced the inclusion of non-meaningful 'buffer' data), and @Berry has explicitly guessed the required conversion, but there the puzzle has stagnated.

Note that if this simple row-based reordering were all that was required, the puzzle would have been trivial to make. What made the puzzle challenging to make, even with allowing nonsense data, is that the red and yellow dots must also be read another obvious way to get the rest of the message, and then hopefully solve the puzzle.


For those who may have trouble reading the image, I have provided a pastebin which contains as much as is relevant of the image that I could describe in words. Owing to the inherently visual nature of this type of art, however, you will need to use at least some parts of the image to find the solution.
It is worth pointing out that artistic license has been used in the precise details of the artwork. Generally speaking, counting precise numbers of small dots is not important. Apart from the numbers of outer dots surrounding some of the circles, and the numbers of green dots, you do not need to count e.g. the number of small dots in each circle. If it comes to counting small dots, if I haven't mentioned it in the pastebin, it's not relevant. Precise details of the colours (e.g. hex values) are also not necessary for the solution.

Anon
  • 5,342
  • 2
  • 13
  • 45
  • 4
    This is a thing of beauty. Did you place each dot by hand?? – Stiv Jun 21 '20 at 10:49
  • 3
    @Stiv Not really. I wrote a program to do it, which probably took much much longer... – Anon Jun 21 '20 at 10:58
  • 1
    Either way - wow! Impressive. And fantastic visuals :) – Stiv Jun 21 '20 at 11:00
  • 1
    @Stiv I'm glad you find the visual pleasing - I did spend a lot of time on it :) – Anon Jun 21 '20 at 11:13
  • 4
    Thank you for the text version! I'm colorblind, and absolutely never would've noticed those green dots. – Deusovi Jun 21 '20 at 18:14
  • @deus — didn’t know you were colorblind! red/green, green/red, blue/yellow...? – matt Jun 21 '20 at 21:17
  • 1
    @Deusovi I'm glad the pastebin helped! Hopefully you can catch enough of the general details of the ochre, brown, blue, and white regions from the image, as they're kinda hard to describe in words. A side note on the green dots: they're in some sense redundant, which means you could solve the puzzle without noticing them. They clue to something you have to do with the data/what is relevant, but you'd very possibly figure it out without them, so if anyone's feeling they've solved/are solving the puzzle without them, don't get too hung up on them. – Anon Jun 21 '20 at 21:28
  • Well, I'm not sure if anyone's still interested in this puzzle, but just in case, I've added a new hint, which I think will enable a big step towards the solution. Hopefully it doesn't make it too trivial... – Anon Jul 10 '20 at 08:43
  • 2
    @Anon I've been turning my attention to this one regularly and have managed to produce various messages as described in your hint (looking most promising with a 're-ordering'), but none of them yet make any meaningful sense! Will look again... – Stiv Jul 10 '20 at 08:44
  • Okay, I'm struggling to see past this. rot13(Jura lbh erbeqre gur ebjf ol ahzore gur lryybj qbgf nccrne gb cebqhpr Zbefr pbqr. Ernq gbc gb obggbz guvf tvirf HGAZAAHNQ; ernq fdhner ol fdhner guvf vafgrnq tvirf RJA BQRHM. Arvgure bs gurfr frrz gb yrnq naljurer jvgu n Pnrfne fuvsg be na nantenz.) Is the intended answer in English? (I've tried other things too - binary, tertiary, columnar ordering - but none seem to be quite as promising as this one right now...) – Stiv Jul 11 '20 at 23:44
  • @Stiv That's definitely one way of looking at it and you're very close! Have a closer look at the image you supplied... You need to read it - is there another relatively obvious way you could transform it into characters? Don't forget that there's extra 'buffer' data due to the puzzle's constraints. – Anon Jul 12 '20 at 12:26
  • @Stiv Gb chg vg nabgure jnl, lbhe vzntr pbagnvaf jung vf ol qrsvavgvba n ovanel frdhrapr juvpu vf cnefrq anghenyyl ol gur funcr bs gur vzntr vagb fhofrdhraprf. Phggvat bhg gur veeryrinag naq abafrafr qngn, gurer fubhyq or n snveyl boivbhf (orpnhfr vg vf fgnaqneq) jnl bs genafsbezvat gurfr fhofrdhraprf vagb punenpgref. Gb nafjre lbhe frpbaq dhrfgvba, grpuavpnyyl vg vf abg Ratyvfu, ohg vgf zrnavat fubhyq or ernfbanoyl boivbhf. – Anon Jul 13 '20 at 03:12
  • @Stiv Vs guvf ebj-onfrq erbeqrevat jrer nyy lbh arrq gb qb ubjrire, gur chmmyr jbhyq unir orra dhvgr gevivny gb znxr, nf V jbhyq whfg unir arrqrq gb erbeqre n ovanel zngevk ol gur ahzoref va gur fbqhxh. Gur xvpxre (naq jung znqr gur chmmyr punyyratvat gb znxr, rira jvgu nyybjvat abafrafr qngn), vf gung lbh nyfb arrq gb ernq gur erq naq lryybj qbgf nabgure boivbhf jnl (juvpu lbh unir gevrq) gb trg gur erfg bs gur zrffntr... – Anon Jul 13 '20 at 03:12
  • @Anon Gunax lbh sbe guvf neg cvrpr bs n chmmyr. V srry yvxr lbh'er uvagvat jr fubhyq erbeqre gur erq naq lryybj qbgf onfrq ba gur ahzoref cre ertvba bs gur fhqbxh. Jura V qb gung, V trg na vzntr juvpu ybbxf n ovg fvzvyne gb Fgvi'f. V abgvpr gung gurer ner dhvgr n srj frdhraprf gung pbhyq fgneg jvgu 0011, juvpu znxrf zr fhfcrpg jr fvzcyl arrqf gb hfr NFPVV qrpbqvat. V srry guvf svgf lbhe uvagf gbb. Jung qb lbh guvax? Fubhyq V whfg fcraq zber gvzr ba vg, be vf guvf n qrnq genvy? Ner jr rira ybbxvat sbe grkg be vf guvf creuncf jurer gur ivfhny nfcrpg fubhyq or pbafvqrerq ntnva? – Berry Aug 25 '20 at 18:54
  • @Berry You're spot on with that line of thinking... I'm now hopeful we'll see a solution to this puzzle soon! – Anon Aug 26 '20 at 02:32
  • @Berry To be clear though - I'm not exactly sure what method you're describing to generate your images, but Stiv's image is correct, and it is by applying your insight to their image that you'll get the information you need. – Anon Aug 26 '20 at 02:35
  • I've been very busy with work for a while, but have been prompted by interest in another of my puzzles to look at few of my other unsolved puzzles. I thought now would be as good a time as any to post a hint to this puzzle. – Anon Sep 25 '20 at 08:48
  • I'm looking it over again. The problem I hit before (and still do not yet know how to overcome) is working out which part is* the buffer*. I've tried many different exclusions but nothing makes any sense among all the 'messages' I generate. I have a hunch about the 'other way of reading them' that you mention, but until I can get this to generate anything meaningful I feel like energy spent on the next step would be misspent! I'll see what I can come up with... – Stiv Sep 25 '20 at 23:17
  • @Stiv ...It is hard to clue any further for now without giving it away... You've generated a correct image by reading horizontally. You'll also need to generate another image to get the full message (there's a symmetry to the puzzle that should help guessing the alternative way of reading). Once you have one or both of those images, there's a straightforward way of reading them (that has already been mentioned). Then just discard the data you don't need... – Anon Sep 26 '20 at 05:03
  • I just want to say that this puzzle is awesome! True puzzling genius. – Dmitry Kamenetsky Sep 28 '20 at 13:29

2 Answers2

10

Partial Answer

The dots around the rings are a

Sudoku.

It solves as (assuming my transcriptions are right):

\begin{array}{|c c c|c c c|c c c|} \hline 3 & 9 & 5 & 1 & 2 & 6 & 4 & 7 & 8 \\ 6 & 4 & 7 & 8 & 3 & 9 & 1 & 2 & 5 \\ 1 & 2 & 8 & 5 & 4 & 7 & 3 & 6 & 9 \\\hline 2 & 1 & 4 & 6 & 5 & 3 & 8 & 9 & 7 \\ 9 & 5 & 6 & 7 & 8 & 1 & 2 & 3 & 4 \\ 8 & 7 & 3 & 2 & 9 & 4 & 5 & 1 & 6 \\\hline 4 & 3 & 1 & 9 & 6 & 5 & 7 & 8 & 2 \\ 5 & 6 & 2 & 3 & 7 & 8 & 9 & 4 & 1 \\ 7 & 8 & 9 & 4 & 1 & 2 & 6 & 5 & 3 \\\hline \end{array}

I have no ideas yet on the green dots, yellow and red centers, the bug, or the X.

David G.
  • 1,606
  • 1
  • 7
  • 12
9

If you take David G.'s partial solution, and:

Add the red and yellow coloring from the dots to the sudoku solution, you obtain:

enter image description here

Then, as per Stiv's partial solution in the comments:

Sort the numbers per row, in ascending order, retaining their red/yellow coloring, you obtain this table:

enter image description here

If you convert the colors of these rows to ASCII, ignoring the first value per row, you obtain the string "S02311�&5"

Similarly if you:

Sort the numbers per column, in ascending order, retaining their red/yellow coloring, you obtain this table:

enter image description here

If you convert the colors of these columns to ASCII, ignoring the first value per column, you obtain the string "E13142�3�"

Looking at both these partial answers:

It is clear that in both cases the last three characters don't make sense. So, as hinted by Anon, we simply discard these. We are left with: "S02311" and "E13142"

Combining these:

We realize they are coordinates! To be exact they convert to: 23° 11' South, 131° 42' East. The coordinates refer to a location in Australia in the Northern Territory's Western Desert area, to the west of Papunya. According to the Wikipedia page Anon included, this area is where Clifford Possum Tjapaltjarri lived and worked as a Tula artist.

Now we also realize:

The puzzle's visuals do not simply depict a bug, but it's actually a map of Papunya! The X marks the location of Papunya Tjupi Arts the Aboriginal art gallery!
Using Google Maps makes this clear:

enter image description here

Berry
  • 612
  • 4
  • 10
  • 2
    Nice! +1 Essentially done, but you still need to make sense of the visual aspect of the rest of the image... – Anon Sep 26 '20 at 21:51
  • 2
    Ah, well done - I had arrived at the same letter-number strings as you but assumed that I had already done the 'buffer removal', and didn't spot the significance of what was produced. Nice spot on the final spoiler too! Another clever puzzle from this OP - nice to see it wrapped up :) – Stiv Sep 27 '20 at 08:41
  • 3
    Awesome answer! It's nice to see this puzzle solved finally. It's the key step of reading the numbers vertically as well that spiced up the process of making it - designing a latin square and binary matrix pair such that the one could be reordered by the other both horizontally and vertically to produce independent messages was a big trial-and-error challenge even with the inclusion of buffer data. Making the dot painting on the other hand was a very fun programming task - I'm glad some people found it visually pleasing! – Anon Sep 27 '20 at 11:42
  • Finally, another one of @Anon 's puzzles has been put to rest after such a long time! :0 – oAlt Sep 27 '20 at 12:05
  • 1
    @Anon, Thank you for this awesome puzzle. Before I realized what the strings were they just felt like gibberish, but once they made sense I was so amazed you were able to make this encoding work! And it's such a pretty image! – Berry Sep 27 '20 at 14:34