Forum Discussion
7 years ago
Decided it might be worth doing a quick code post-mortem for anybody curious how we all figured that out.
For me, I started by looking at the gold characters, thinking they might be useful. I grabbed an Aurebesh translator image like the one @RogueJedi posted and converted the second image into a block of letters. Then I looked at the first image, and started translating the gold characters only. I noticed that there were a number of letters, but they were all in the A to F range---no letters after F. I also noticed that there were a lot of 4's and 5's in the image. The character sets posted by @Mstrefe1 and @MistaLeahy also helped a lot for seeing these patterns. @Iy4oy4s also noticed some of this patterning and commented on it.
Those were both consistent with the idea that the image might be a big block of hexadecimal code. The A to F bit is straightforward there, but the 4's and 5's less so. It's a safe bet, though, that all the letters in ASCII are going to come in a contiguous sequence, and since there are 26 letters in the alphabet, they should all fall in a range with just 2-3 distinct (also contiguous) leading digits. I didn't know that 4 and 5 would be the two digits in question, but the format was suspicious enough to make me think a hexadecimal to ASCII conversion was probably the right way to go. @gordonjones also picked up on the apparent hex patterning---I think it was just sort of dumb luck that there weren't any C's in the list. Looking at Reddit, it looks like a few other people also independently came up with the hex step over there.
The next bit owes heavily to insights from @JediKnight_MarkRyan and @APX_919 . When I first translated the characters in the second image from Aurebesh, I didn't notice that each letter was appearing only once. Or that 'J' was missing, which @chionophile caught and which wound up being very important. Knowing that, my first inclination was to try to treat the problem like a substitution cypher---kind of like the cryptoquips that show up in newspapers. That went nowhere fast, though. The obvious substitution schemes I tried resulted in character strings that were obviously not plaintext, so I was able to rule them out pretty quickly. But @APX_919 's comment about the Polybius checkerboard is what got me pointed in the right direction. I checked out the Wikipedia page for it, and found it didn't really help me much, but there was a comment in the Wiki page about the Playfair cypher being paired with the Polybius setup. When I looked that one over, it started to look very promising.
I still screwed it up pretty badly at first, though. I was trying to rush through the decoding, and I wound up running the cypher backwards---double-encoding the bits that fell into the same row/column and decoding the bits that formed rectangles. But I was generating letters that looked more like natural text, so I was pretty sure that I was onto something; I just needed to get my act together on the decoding. Eventually, I got the first bit to come out plain: "The road ahead..." This is where @KausDebonair got to as well---but then I ran into the problem of the 7's and D's. I completely missed this when I did the initial ASCII encoding, using one of the character lists provided in the thread. Without @dsteve , @Kingpin01 , and @JediKnight_MarkRyan pointing that out, I would have been stuck for a while (and might not have posted the bit about the Playfair cypher). Unfortunately, I didn't see @dsteve 's revision of the hex code and ASCII, so I went off and did it by hand from the image again---and then did the deciphering by hand as well. By the time I finished, @newershadow already had the text up for everyone to see. He got exactly what I'd gotten by hand, though; it was nice to be sure this was working exactly as intended. And he got the bit.ly link, which would have flown completely over my head.
The last bit with the 'J' was awesome, I thought. The I/J issue was still floating around in the back of my head from (a) @chionophile 's earlier comment about the missing J and (b) the Wikipedia description of the Playfair cypher where they put the two together in the 5x5 grid. As soon as I watched the clip from @CG_Carrie , I knew exactly what needed to happen---thanks in large part to @Sev74907 , who picked up on the fact that the bit.ly link had Revan backwards in it.
Anyway, for anyone reading this thread who was trying to figure out how on earth we all figured this out, I thought it might be nice to write up a run-through. It was really fun working with all you guys on this and cracking the code (and to go check out CubsFanHan's livestream right after we finished, and get to see him pull up the video right after us). Big props to @CG_SBCrumb for making this Friday night so much fun, too! It was a blast. And here's hoping I didn't miss too many of the people contributing to the decoding when I was writing this up... If I did miss you, I'm very sorry! But everybody participating in the real-time brainstorming over here was invaluable, and if you were in on it, thank you very much!
For me, I started by looking at the gold characters, thinking they might be useful. I grabbed an Aurebesh translator image like the one @RogueJedi posted and converted the second image into a block of letters. Then I looked at the first image, and started translating the gold characters only. I noticed that there were a number of letters, but they were all in the A to F range---no letters after F. I also noticed that there were a lot of 4's and 5's in the image. The character sets posted by @Mstrefe1 and @MistaLeahy also helped a lot for seeing these patterns. @Iy4oy4s also noticed some of this patterning and commented on it.
Those were both consistent with the idea that the image might be a big block of hexadecimal code. The A to F bit is straightforward there, but the 4's and 5's less so. It's a safe bet, though, that all the letters in ASCII are going to come in a contiguous sequence, and since there are 26 letters in the alphabet, they should all fall in a range with just 2-3 distinct (also contiguous) leading digits. I didn't know that 4 and 5 would be the two digits in question, but the format was suspicious enough to make me think a hexadecimal to ASCII conversion was probably the right way to go. @gordonjones also picked up on the apparent hex patterning---I think it was just sort of dumb luck that there weren't any C's in the list. Looking at Reddit, it looks like a few other people also independently came up with the hex step over there.
The next bit owes heavily to insights from @JediKnight_MarkRyan and @APX_919 . When I first translated the characters in the second image from Aurebesh, I didn't notice that each letter was appearing only once. Or that 'J' was missing, which @chionophile caught and which wound up being very important. Knowing that, my first inclination was to try to treat the problem like a substitution cypher---kind of like the cryptoquips that show up in newspapers. That went nowhere fast, though. The obvious substitution schemes I tried resulted in character strings that were obviously not plaintext, so I was able to rule them out pretty quickly. But @APX_919 's comment about the Polybius checkerboard is what got me pointed in the right direction. I checked out the Wikipedia page for it, and found it didn't really help me much, but there was a comment in the Wiki page about the Playfair cypher being paired with the Polybius setup. When I looked that one over, it started to look very promising.
I still screwed it up pretty badly at first, though. I was trying to rush through the decoding, and I wound up running the cypher backwards---double-encoding the bits that fell into the same row/column and decoding the bits that formed rectangles. But I was generating letters that looked more like natural text, so I was pretty sure that I was onto something; I just needed to get my act together on the decoding. Eventually, I got the first bit to come out plain: "The road ahead..." This is where @KausDebonair got to as well---but then I ran into the problem of the 7's and D's. I completely missed this when I did the initial ASCII encoding, using one of the character lists provided in the thread. Without @dsteve , @Kingpin01 , and @JediKnight_MarkRyan pointing that out, I would have been stuck for a while (and might not have posted the bit about the Playfair cypher). Unfortunately, I didn't see @dsteve 's revision of the hex code and ASCII, so I went off and did it by hand from the image again---and then did the deciphering by hand as well. By the time I finished, @newershadow already had the text up for everyone to see. He got exactly what I'd gotten by hand, though; it was nice to be sure this was working exactly as intended. And he got the bit.ly link, which would have flown completely over my head.
The last bit with the 'J' was awesome, I thought. The I/J issue was still floating around in the back of my head from (a) @chionophile 's earlier comment about the missing J and (b) the Wikipedia description of the Playfair cypher where they put the two together in the 5x5 grid. As soon as I watched the clip from @CG_Carrie , I knew exactly what needed to happen---thanks in large part to @Sev74907 , who picked up on the fact that the bit.ly link had Revan backwards in it.
Anyway, for anyone reading this thread who was trying to figure out how on earth we all figured this out, I thought it might be nice to write up a run-through. It was really fun working with all you guys on this and cracking the code (and to go check out CubsFanHan's livestream right after we finished, and get to see him pull up the video right after us). Big props to @CG_SBCrumb for making this Friday night so much fun, too! It was a blast. And here's hoping I didn't miss too many of the people contributing to the decoding when I was writing this up... If I did miss you, I'm very sorry! But everybody participating in the real-time brainstorming over here was invaluable, and if you were in on it, thank you very much!
About SWGOH General Discussion
Discuss and share your feedback on Star Wars: Galaxy of Heroes with fellow players.77,436 PostsLatest Activity: 10 minutes ago
Related Posts
Recent Discussions
- 12 minutes ago
- 16 minutes ago