Tadpole code convert any <4k string to a 'single' character (Unicode magic) • Hacker News Board

Re: Tadpole code convert any <4k string to a 'single' character (Unicode magic)

#11

Post by Quarrelsome » Fri, Jan 08, 2010, 2:41 PM UTC

paraschopra wrote: post #10

JeffJenkins wrote: post #3

Clever. Looks like it's using all of the diacritic marks to encode information onto a / character. (it's easy to see if you use a large enough font)

Can you explain a bit more in detail? This looks fantastic and am interesting in knowing more about it but didn't really get what you mean by diacritic marks.

I'm interested too. I'm guessing the data is actually encoded into the character as they do look ever so slightly different with different content.

This would suggest this means that you can create a dynamic font/character in unicode. I had no idea this was possible, can someone confirm/slap me down for saying silly things?

Re: Tadpole code convert any <4k string to a 'single' character (Unicode magic)

#12

Post by BigZaphod » Fri, Jan 08, 2010, 2:50 PM UTC

I tried posting a long (more than 140 chars) one of these on twitter, but it seems they cut it off because it won't decode. (Checksum error.) Bummer.. would have been a fun loophole. :P

Update: Actually, sending a short one via the twitter client I'm working on seems to mangle it as it won't decode after it's posted. Posted short ones via the web interface will decode correctly, though. There must be a unicode handling bug somewhere...

Re: Tadpole code convert any <4k string to a 'single' character (Unicode magic)

#13

Post by brk » Fri, Jan 08, 2010, 2:54 PM UTC

What am I missing? How do you decode it?

Re: Tadpole code convert any <4k string to a 'single' character (Unicode magic)

#14

Post by afshin » Fri, Jan 08, 2010, 2:54 PM UTC

paraschopra wrote: post #10

JeffJenkins wrote: post #3

Clever. Looks like it's using all of the diacritic marks to encode information onto a / character. (it's easy to see if you use a large enough font)

Can you explain a bit more in detail? This looks fantastic and am interesting in knowing more about it but didn't really get what you mean by diacritic marks.

Characters can have different accents or modifications on them. For example, the letter "e" can be "é" or "è" or "ë", etc. The interesting thing being done here is layering multiple diacritical marks on top of the slash character and then decoding those marks back into plain text.

Re: Tadpole code convert any <4k string to a 'single' character (Unicode magic)

#15

Post by sounddust » Fri, Jan 08, 2010, 2:58 PM UTC

brk wrote: post #13

What am I missing? How do you decode it?

You go to the page linked in the story, paste the tadpole-looking character into the field at the top, and press the "convert (0-4kbyte)" button.

Re: Tadpole code convert any <4k string to a 'single' character (Unicode magic)

#16

Post by est » Fri, Jan 08, 2010, 3:04 PM UTC

brk wrote: post #13

What am I missing? How do you decode it?

paste the 'single' character into the textbox and click convert

example screenshot grabbed from reddit:

http://i.imgur.com/LwCSK.jpg

Re: Tadpole code convert any <4k string to a 'single' character (Unicode magic)

#17

Post by andrewcooke » Fri, Jan 08, 2010, 3:53 PM UTC

blazzerbg wrote: post #5

JeffJenkins wrote: post #3

Clever. Looks like it's using all of the diacritic marks to encode information onto a / character. (it's easy to see if you use a large enough font)

/ۤۛۗۜۚۛ۫ۗۢۜۛ۠ۤۨۜ۟۟۫ۗۢۚۘۖ۠۟ۜۗۛۚۜ۠ۡۚ۫ۙۧ۫ۘۧ۬ۤۖ۫ۛ۬ۛۚۧۚۢۧۡۖ۫۟ۧۚۡ۠۫ۜۙۜۡۙۢۡۖۚۡۚۛ۠ۡۗۗ۠ۢۗۢۤۖۢۜۡۧۙۡۜۗۙۛۖۧۖۨۡۢۙ۫ۤۙۙۙۢ۫ۘۡ۟ۗۧۧۚۨۗۤۢۚۚۙۙ۟ۜ۬ۧۧۛ۫ۨ۟ۤۛ۟ۘۛۡۤۨۚ۠ۘۘ۠۬۟ۨۤ۟ۖۚۤۚۗۡ۟ۖ۫ۜ۠ۤ۬ۖۚۧۨ۫ۘۨ۟۬ۢ۟ۗۚۡۜۜۙۗ۟۫ۚۧۜۡۗۧۗۚ۠ۙۖۢۡۙ۫ۤ۠ۛۨۖۤ۠ۗۡۡ۟ۡ۠۠ۚ۫۠ۛۡۘۤ۠ۛۨ۫ۧ۟ۤۨۜۙۚ۫ۘ۫ۡۤۗۢۜۖ۬ۛۜۤۚۛۨۨ۟ۜۙۜ۫ۡۨۛ۠ۗۘۨۙ۫ۚ۟ۗۢۡ۫ۖۖۢۤۚ۫ۛۖۙ۠ۖۗۛۧۛۧۖ۟ۧۘ۬ۘۜ۫ۨۢۛۙۛ۟۬ۤۧۧۖۡۜۗۤ

sorry - i downvoted you by accident while trying to select for cut + paste :(

Re: Tadpole code convert any <4k string to a 'single' character (Unicode magic)

#18

Post by patio11 » Fri, Jan 08, 2010, 3:55 PM UTC

est wrote: post #9

Here's the funny story: this was invent by a Chinese developer (rinick of www.goozo.net) in a filesharing discussion board (TLF). Then it was popularized to avoid keyword censorship systems everywhere in China.

If Unicode causes the Chinese regime as much suffering as it has caused me, I take back everything bad I have ever said about it. Bring on the canonicalization issues!

Re: Tadpole code convert any <4k string to a 'single' character (Unicode magic)

#19

Post by dysth » Fri, Jan 08, 2010, 4:04 PM UTC

dzorz wrote: post #6

/۟ۢۘۙۨۙۖ۫ۜۜۖ۟ۡ۬ۗۡ۠۬ۧۤۧ۬ۚ۫ۖۤ۠ۢۚ۫ۡ۬۫ۢۛۗۡۙۡ۟ۧۤۡۘ۠ۘۜ۠ۜۛۙ۬ۙۖۗۨۙۤۛ۬ۘۛۜۢۧۧۧۗۖۡ۟ۛ۫ۤۡۖۘۚۙۤ۠ۡۧۜۤۖۡۨۜۨ۠۠ۙۚ۫۟ۢۛ۫ۛۧۚ۬ۚۘ۟ۤۙۨۤ۠ۘۤۘ۬ۤۜۤ۠ۨۨۤۤ۬ۡۤۧۘۛ۟۟۫ۗۢۗۙۢۚۨۖ۠ۤۖ۬۟ۢۤۙ۟۟ۢ۟۟ۘ۫ۜۙ۟۫۟ۗۨ۬ۘۘ۟۟ۜۛ۫ۧۧۧۗۢۤۢۜۧۖ۟ۧۜ۠ۢۘۖۤۗۙۛۧ

"I guess the trick is in using combining characters? I guess the first codepoint is always /."

Re: Tadpole code convert any <4k string to a 'single' character (Unicode magic)

#20

Post by nirmal » Fri, Jan 08, 2010, 4:04 PM UTC

Quarrelsome wrote: post #11

paraschopra wrote: post #10

Earlier quoted context omitted.

Can you explain a bit more in detail? This looks fantastic and am interesting in knowing more about it but didn't really get what you mean by diacritic marks.

I'm interested too. I'm guessing the data is actually encoded into the character as they do look ever so slightly different with different content. This would suggest this means that you can create a dynamic font/character in unicode. I had no idea this was possible, can someone confirm/slap me down for saying silly things?

I'd like to see the algorithm used to do this sort of compression.