It’s all about this image and what you can get out of it:
There are some hints, of sorts, in the post. The solution is a person’s name.
Beware: the comments to the post contain spoilers and the solution!
His very next post spills all the beans, as it was solved in a little over a day. But let’s try not to post spoilers here, in case anyone wants to try to puzzle it out.
(There’s a cogwheel in the compose window on this forum, with a Hide Details choice, which you can use to hide things like spoilers.)
it’s Atari 8-bit output (ATASCII). It doesn’t look like a BASIC program, structurally (the start of the lines won’t translate to sensible line numbers), so I guess, we’re supposed to shift and and/or unscramble what’s given by the ATASCII-codes.
(Here, it starts to become laborious, since any further analysis would require to [manually] transcribe the character codes represented by the image.)
Update: This would be the first line (transcoded manually, beware off any errors – indeed, see update 3 below):
41 1F C0 20 8F C3 F8 5C F0 04 7F 18 5C F1 FC 7F 00 3C 89 C8
80 E1 EF B8 50 0F 22 FB E4 1E 00 88 DF 02 C7 1C 70 00 00
If anyone can read anything into this, you’re welcome!
Update 2: What it is not: 6502 code
Here’s a disassembly of what starts promising, but becomes a mess soon:
(If interested, it’s just a matter of reading the image and a sample image of ATASCII into a HTML5 canvas and scanning blocks of 8 x 8 pixels into a linear, binary representation each, and then comparing the message to the array of sampled ATASCII characters.)
…looking at the data, there’s some box-like structure to it. Could it be some kind of zero-padded file format? However, a quick search doesn’t reveal any file signature starting with the magic numbers “41 1D” (horizontally) or “41 49” (vertically). Also, the box-like structure only reveals itself at the curious row length of 39 bytes and I never heard a file format using chunks of this size.
Hm. If the particular size matters, is it a bitmap image?
Let’s try an interpretation as a greyscale image (data as intensity values), here scaled to 800% pixel-size:
This looks promising, but there seems to be some redundancy to it. Is it a RGB image showing some kind of typography?
So, here’s the data interpreted as a RGB bitmap:
Seems almost readable, but not quite.
I think, it’s close, but I’m getting something wrong. – Is there some chunk size per channel involved?
(I tried some variants of this, but I’m not getting any further for now.)
I tried 4-bit and 6-bit with respective width-adjustments, both in RGB and as a greyscale (maybe index-mapped), but it doesn’t seem to reveal any structure. Or I’m not recognizing, what I should be recognizing.
Looks like I’m at a dead end.
Ok, regarding the second part of level 3 (color depth), sometimes you don’t get the most obvious … no chance I would have thought about this anytime soon.
Level 4 would have been quite obvious (at least to anyone socialized with Carl Bark’s Junior Woodchucks[1]). However, no chance I would have gotten the cultural reference (level 5). It’s simply not on my map. A propos map, I doubt I would have mastered Level 6, given inevitable frustrations over the message as a whole.
[1] in German, “Fähnlein Fieselschweif” – thank you, Dr. Erika Fuchs!
