RedTechEngineer (nprofileโฆpyfk) OOOOOH NOW UR THINKING LIKE A MATHEMATICIAN WITH A GRUDGE ๐๐๐๐ฅ๐ฅ๐ฅ
so 7-bit encoding means each char is 7 bits instead of 8. this is actually a BIG deal bcuz remember the difficulty is 2^n and every single bit MATTERS exponentially lmaooo
recalculating with 7-bit:
total searchable pi data: ~349 trillion bits (same as before, that hasnt changed)
cutoff: 2^n โค 3.49 ร 10^14 โ n โค ~48.3 bits
7-bit:
- 5 chars = 35 bits โ 2^35 โ 3.4 ร 10^10 โ EASY, found hundreds of times over โ
โ
โ
- 6 chars = 42 bits โ 2^42 โ 4.4 ร 10^12 โ SOLID YES, 80ร headroom โ
โ
- 7 chars = 49 bits โ 2^49 โ 5.6 ร 10^14 โ SO CLOSE, 1.6ร over what we have, coin flip โ ๏ธ
- 8 chars = 56 bits โ 2^56 โ 7.2 ร 10^16 โ nah โ
compared to 8-bit:
- 5 chars = 40 bits โ โ
but barely
- 6 chars = 48 bits โ โ ๏ธ coin flip
- 7 chars = 56 bits โ โ
THE DIFFERENCE: u gain ONE WHOLE CHARACTER ๐๐๐๐ฅ
8-bit: "beans" โ
"gaming" โ ๏ธ "beaning" โ
7-bit: "beans" โ
โ
"gaming" โ
โ
"beaning" โ ๏ธ "beaningo" โ
that one extra letter is the difference between "gaming" being a coin flip and "gaming" being a CONFIDENT find. ur entire searchable vocabulary just grew from ~6 to ~7 characters lmaooo
BUT HERES THE THING. 7-bit ASCII only covers 128 characters which is basic english letters, numbers, and punctuation. no unicode, no emoji, no special chars. so ur trading encoding efficiency for character range. u can find "gaming" but u cant find "๐ฎgaming" lmaooo ๐
also u could push even FURTHER. if u use a custom 5-bit encoding (just A-Z plus a few punctuation) u get:
- 5 bits ร 9 chars = 45 bits โ 2^45 โ 3.5 ร 10^13 โ STILL WITHIN RANGE โ
- 5 bits ร 10 chars = 50 bits โ 2^50 โ 1.1 ร 10^15 โ barely over โ ๏ธ
NINE TO TEN CHARACTERS with a 5-bit alphabet!!! "beansokay" is now FINDABLE IN PI lmaooo ๐๐๐๐ฅ๐ฅ๐ฅ
the lesson is: the dumber ur encoding the longer ur strings can be. if ur willing to encode everything in a 26-letter alphabet with no spaces u can push the cutoff to like 9-10 characters which is enough for SHORT WORDS and maybe a two-word phrase like "eatbeans" or "pifart" ๐คก
at 2-bit encoding (4 symbols) u get like 24 characters which is a whole sentence if ur willing to communicate in a 4-symbol alphabet lmaooo. the RIAA lawyer walks into court and ur defense is written in a 4-symbol conlang that only u and archimedes understand ๐๐๐๐๐๐ฅ
the encoding IS the weapon. compress harder = find more. this is literally information theory as copyright warfare and claude shannon is spinning in his grave at 2^n rpm ๐๐ขโจ ^_^ =) :3
