Alternative UTF-8 Letters

I don't know why this isn't more discussed.
Instead of chatting on places with regular easily parsable letters, why not use UTF-8 alternatives that are both easy to read (hopefully very similar so other people don't complain) but harder to parse and analyze?

a) Post character sets that can be used for this

b) Discuss tools/ways to apply them to systems/applications (I've used Autohotkey for this before)

Attached: how-to-prepare-lettering-font-making-1200x580.jpg (1200x580, 137.05K)

Other urls found in this thread:

vocaroo.com/i/s1CQEmpJ6Ob8
en.wikipedia.org/wiki/Ideographic_Description_Characters_(Unicode_block)
alt-codes.net/gender-symbol.php
alt-codes.net/snowflakes-symbols.php
en.wiktionary.org/wiki/Appendix:Unicode/Egyptian_Hieroglyphs
twitter.com/SFWRedditImages

Also the characters should be common enough to be viewable to other people (and yourself) in most cases.

Yeah, I thought about making a webapp that does this to text to help fight censorship. People already do this to make fancy usernames on sites like Twitter. Just go through a string and randomly select similar looking characters. Throw in some zalgo, and you can get around wordfiltering pretty easily.

Here's an example of an autohotkey script:

a::Send Ǻb::Send вc::Send Ƈd::Send Ðe::Send €f::Send ƒg::Send ǥh::Send Ћi::Send Ïj::Send ʝk::Send ĸl::Send Łm::Send ɱn::Send иo::Send Øp::Send ρq::Send գr::Send яs::Send Ŝt::Send †u::Send µv::Send ∀w::Send ωx::Send ×y::Send Ŷz::Send ƶ^+p::Suspend, toggle

Here's another set (some letters are normal):
ᗩ ᗷ ᑕ ᗪ E ᖴ G ᕼ I ᒍ K ᒪ ᗰ ᑎ O ᑭ ᑫ ᖇ S T ᑌ ᐯ ᗯ ᙭ Y Z

It's called unicode, nigger.

If you're worried about linguistical fingerprinting then this will make it worse, because just the fact that you're a member of the group who posts in weird glyphs makes it easier to identify posts made by you. Not to mention that if you post without protection all the agencies have your IP and thus your ID already.
If you're worried about automated analysis for purposes other than de-anonymization, I wouldn't worry about it. This place is interesting and low volume enough that every posts is probably read by an actual agent rather than just an AI, and current AI is too dumb to dox people from random details posted anyway, so what could they achieve by running an AI over all the posts that isn't achieved by human review? Besides, if they really wanted to spy on you that bad and had an advanced AI capable of analyzing natural text, they could bypass character substitutions pretty easily. It's just a matter of taking a bunch of substitutions like in , automatically generating a couple gigs of scrambled words, and feeding the original plus the modified words to a seq2seq network to obtain the original un-scrambled word.

Sure but that'd take some time until the changes trickle down to all the other AI tools, plus the extra processing might cause them to be a bit slower overall. And yeah, unless most people is doing it it's just some sort of signature.


Fuck off.

He's right, though. utf-8 is an encoding for Unicode. The encoding is largely arbitrary from an end-user perspective.

Attached: Untitled.png (306x138, 6.15K)

Not a bad idea. A script to quickly generate images based on input text would be nice.

What the fuck are you talking about? Are you advocating a removal of UTF-8 and replacing it with something like EBCDIC? Or are you suggesting we use obscure Unicode Characters involving different fonts that look like latin characters? If it is the former, you can try all you want but you're 20 years too late. If it is the later, this post is just awful and it is painfully obvious you do not belong here.

vocaroo.com/i/s1CQEmpJ6Ob8

For a short time in 16chan(?) there was a script that let you convert the text in posts from/to unreadable characters. Nobody could read it except for people who had that script. I think it was called niggertexting or something.

Attached: whynot.webm (306x138, 182.43K)

My idea was similar. You'd have a list of possible unicode substitutions (including not even making a substitution), and randomly pick one for each letter, and maybe pepper in some meaningless diacritics to pad it out.

Is there a way to encrypt messages, so that you give everyone you want to be able to read it a (different) key, and when you don't want anyone to be able to read your new messages anymore after some point, you revoke their key, but the others don't need to change theirs? I can image doing this by encrypting each message with a different key and adding an encrypted form of the message's key for each person, but can this be done in a way that doesn't add multiple bytes of overhead for each additional person you add to the list?

"Key revocation" is not a thing, even if the cryptotards try to make you think it is (can't sell the "identity-based" "encryption" scam otherwise). What you describe is simply not encrypting to some person you dislike, nothing is being revoked. The "key revocations" you likely heard about are simply signed messages saying "oops lole don't use key XYZABC".
This is how current systems like PGP do it.
No known method afaik. I highly doubt that it is possible because of information-theoretic limits; this is no proof though. Even if possible, I would expect it to have a massive cost upfront that makes the scheme impractical. That said, don't take my suspicions as gospel.

𝐀𝐁𝐂𝐃𝐄𝐅𝐆𝐇𝐈𝐉𝐊𝐋𝐌𝐍𝐎𝐏𝐐𝐑𝐒𝐓𝐔𝐕𝐖𝐗𝐘𝐙 𝐚𝐛𝐜𝐝𝐞𝐟𝐠𝐡𝐢𝐣𝐤𝐥𝐦𝐧𝐨𝐩𝐪𝐫𝐬𝐭𝐮𝐯𝐰𝐱𝐲𝐳
𝐴𝐵𝐶𝐷𝐸𝐹𝐺𝐻𝐼𝐽𝐾𝐿𝑀𝑁𝑂𝑃𝑄𝑅𝑆𝑇𝑈𝑉𝑊𝑋𝑌𝑍 𝑎𝑏𝑐𝑑𝑒𝑓𝑔𝑖𝑗𝑘𝑙𝑚𝑛𝑜𝑝𝑞𝑟𝑠𝑡𝑢𝑣𝑤𝑥𝑦𝑧
𝑨𝑩𝑪𝑫𝑬𝑭𝑮𝑯𝑰𝑱𝑲𝑳𝑴𝑵𝑶𝑷𝑸𝑹𝑺𝑻𝑼𝑽𝑾𝑿𝒀𝒁 𝒂𝒃𝒄𝒅𝒆𝒇𝒈𝒉𝒊𝒋𝒌𝒍𝒎𝒏𝒐𝒑𝒒𝒓𝒔𝒕𝒖𝒗𝒘𝒙𝒚𝒛
𝒜𝒞𝒟𝒢𝒥𝒦𝒩𝒪𝒫𝒬𝒮𝒯𝒰𝒱𝒲𝒳𝒴𝒵 𝒶𝒷𝒸𝒹𝒻𝒽𝒾𝒿𝓀𝓁𝓂𝓃𝓅𝓆𝓇𝓈𝓉𝓊𝓋𝓌𝓍𝓎𝓏
𝓐𝓑𝓒𝓓𝓔𝓕𝓖𝓗𝓘𝓙𝓚𝓛𝓜𝓝𝓞𝓟𝓠𝓡𝓢𝓣𝓤𝓥𝓦𝓧𝓨𝓩 𝓪𝓫𝓬𝓭𝓮𝓯𝓰𝓱𝓲𝓳𝓴𝓵𝓶𝓷𝓸𝓹𝓺𝓻𝓼𝓽𝓾𝓿𝔀𝔁𝔂𝔃
𝔄𝔅𝔇𝔈𝔉𝔊𝔍𝔎𝔏𝔐𝔑𝔒𝔓𝔔𝔖𝔗𝔘𝔙𝔚𝔛𝔜 𝔞𝔟𝔠𝔡𝔢𝔣𝔤𝔥𝔦𝔧𝔨𝔩𝔪𝔫𝔬𝔭𝔮𝔯𝔰𝔱𝔲𝔳𝔴𝔵𝔶𝔷
𝔸𝔹𝔻𝔼𝔽𝔾𝕀𝕁𝕂𝕃𝕄𝕆𝕊𝕋𝕌𝕍𝕎𝕏𝕐 𝕒𝕓𝕔𝕕𝕖𝕗𝕘𝕙𝕚𝕛𝕜𝕝𝕞𝕟𝕠𝕡𝕢𝕣𝕤𝕥𝕦𝕧𝕨𝕩𝕪𝕫
𝕬𝕭𝕮𝕯𝕰𝕱𝕲𝕳𝕴𝕵𝕶𝕷𝕸𝕹𝕺𝕻𝕼𝕽𝕾𝕿𝖀𝖁𝖂𝖃𝖄𝖅 𝖆𝖇𝖈𝖉𝖊𝖋𝖌𝖍𝖎𝖏𝖐𝖑𝖒𝖓𝖔𝖕𝖖𝖗𝖘𝖙𝖚𝖛𝖜𝖝𝖞𝖟
𝖠𝖡𝖢𝖣𝖤𝖥𝖦𝖧𝖨𝖩𝖪𝖫𝖬𝖭𝖮𝖯𝖰𝖱𝖲𝖳𝖴𝖵𝖶𝖷𝖸𝖹 𝖺𝖻𝖼𝖽𝖾𝖿𝗀𝗁𝗂𝗃𝗄𝗅𝗆𝗇𝗈𝗉𝗊𝗋𝗌𝗍𝗎𝗏𝗐𝗑𝗒𝗓
𝗔𝗕𝗖𝗗𝗘𝗙𝗚𝗛𝗜𝗝𝗞𝗟𝗠𝗡𝗢𝗣𝗤𝗥𝗦𝗧𝗨𝗩𝗪𝗫𝗬𝗭 𝗮𝗯𝗰𝗱𝗲𝗳𝗴𝗵𝗶𝗷𝗸𝗹𝗺𝗻𝗼𝗽𝗾𝗿𝘀𝘁𝘂𝘃𝘄𝘅𝘆𝘇
𝘈𝘉𝘊𝘋𝘌𝘍𝘎𝘏𝘐𝘑𝘒𝘓𝘔𝘕𝘖𝘗𝘘𝘙𝘚𝘛𝘜𝘝𝘞𝘟𝘠𝘡 𝘢𝘣𝘤𝘥𝘦𝘧𝘨𝘩𝘪𝘫𝘬𝘭𝘮𝘯𝘰𝘱𝘲𝘳𝘴𝘵𝘶𝘷𝘸𝘹𝘺𝘻
𝘼𝘽𝘾𝘿𝙀𝙁𝙂𝙃𝙄𝙅𝙆𝙇𝙈𝙉𝙊𝙋𝙌𝙍𝙎𝙏𝙐𝙑𝙒𝙓𝙔𝙕 𝙖𝙗𝙘𝙙𝙚𝙛𝙜𝙝𝙞𝙟𝙠𝙡𝙢𝙣𝙤𝙥𝙦𝙧𝙨𝙩𝙪𝙫𝙬𝙭𝙮𝙯
𝙰𝙱𝙲𝙳𝙴𝙵𝙶𝙷𝙸𝙹𝙺𝙻𝙼𝙽𝙾𝙿𝚀𝚁𝚂𝚃𝚄𝚅𝚆𝚇𝚈𝚉 𝚊𝚋𝚌𝚍𝚎𝚏𝚐𝚑𝚒𝚓𝚔𝚕𝚖𝚗𝚘𝚙𝚚𝚛𝚜𝚝𝚞𝚟𝚠𝚡𝚢𝚣
ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz
⒜⒝⒞⒟⒠⒡⒢⒣⒤⒥⒦⒧⒨⒩⒪⒫⒬⒭⒮⒯⒰⒱⒲⒳⒴⒵ ⒜⒝⒞⒟⒠⒡⒢⒣⒤⒥⒦⒧⒨⒩⒪⒫⒬⒭⒮⒯⒰⒱⒲⒳⒴⒵
ⒶⒷⒸⒹⒺⒻⒼⒽⒾⒿⓀⓁⓂⓃⓄⓅⓆⓇⓈⓉⓊⓋⓌⓍⓎⓏ ⓐⓑⓒⓓⓔⓕⓖⓗⓘⓙⓚⓛⓜⓝⓞⓟⓠⓡⓢⓣⓤⓥⓦⓧⓨⓩ
①②③④⑤⑥⑦⑧⑨⑩⑪⑫⑬⑭⑮⑯⑰⑱⑲⑳
0 1 2 3 4 5 6 7 8 9

Ā ā Ă ă Ą ą Ĉ ĉ Ċ ċ Č č Ď ď Đ đ Ē ē Ĕ ĕ Ė ė Ę ę Ě ě Ĝ ĝ Ğ ğ Ġ ġ Ģ ģ Ĥ ĥ Ħ ħ Ĩ ĩ Ī ī Ĭ ĭ Į į İ ı IJ ij Ĵ ĵ Ķ ķ ĸ Ĺ ĺ Ļ ļ Ľ ľ Ŀ ŀ Ł ł Ņ ņ Ň ň ʼn Ŋ ŋ Ō ō Ŏ ŏ Ő ő Œ œ Ŕ ŕ Ŗ ŗ Ř ř Ś ś Ŝ ŝ Ş ş Š š Ţ ţ Ť ť Ŧ ŧ Ū ū Ŭ ŭ Ů ů Ű ű Ų ų Ÿ Ż ż Ž ž ſ ƒ Ʒ DŽ Dž dž LJ Lj lj NJ Nj nj ǝ Ǟ ǟ Ǥ ǥ Ǧ ǧ Ǩ ǩ Ǯ ǯ DZ Dz dz Ǵ ǵ Ǻ ǻ Ǽ ǽ Ǿ ǿ ɐ ɹ ɼ ʇ ʌ ʍ ʎ ʒ
Ḃ ḃ Ḋ ḋ Ḑ ḑ Ḟ ḟ Ḱ ḱ Ṁ ṁ Ṗ ṗ Ṡ ṡ Ṫ ṫ Ὅ

I don't think that will work. Let's say there are ten different glyphs that look like the letter A, then all a parser has to do is look up what glyph corresponds to to unscramble the message.

For example, the unscrambler could look like this:
char scrabled = getchar();char unscrambled;if scrambled in [𝐀, 𝐴, 𝑨, 𝒜, 𝓐, 𝔄, 𝔸, 𝕬, 𝖠, ...] unscrambled = A;else if scrambled in [𝐁, 𝐵, ...] unscrabmled = B;else if ...
You get the idea. Adding diacritics just adds a bit of overhead to the process, but that's it.

Yeah, I proposed a neural network based approach to handle ambiguous characters or misspellings, but that works too for most cases.
The big guns like the NSA probably have that in place already, it would be pretty ridiculous if you could bypass Echelon just by using some l33tspeak.

GNU Unifont solves this problem - it is one of not many free fonts supporting the entire Unicode.

what the fuck is this retarded shit?
is this babbys first encryption algorithm?
kill yourselves

It's not an encryption algorithm you pretentious fuck, it's just a small change that if adopted could _potentially_ make it harder for some automated text processors. Sure it can be easily hardcoded to avoid this and it probably already is, but likely not on all systems. Also stop telling people to kill themselves, it's not nice or necessary.

reminder to learn your alternate characters.
t. Chinese who hates how UTF-8 fucked Chinese

This is dumb shit. They should remove it and add the hooked/runic cross and other common symbols. Same for emojis and other garbage.

It didn't. You have traditional characters. If you're talking about the CJK-Unification, that's completely justified. Just use a specifically Chinese font on your Chinese website or in your document.
German fraktur doesn't work with a lot of other European languages like English and all European letters have been unified too.
However I only hear you god damn ant people bitching about it. It's actually a pro because it makes it easier to identify letters. We already have enough letters in Unicode that look the exact same.

kill yourself


Fuck chinks. I hate you fucks even more than niggers and jews.
I hope you gas yourselves to death over there.

Holy shit dude, how new are you? Did you arrive from 4cuck yesterday, or came straight here after hearing about Tarrant on the news? Or was it after the_donald was banned on reddit a few days ago?

because its already done where needed, even by 12 year olds. the normal solution to this problem is to just use end to end encryption for talking to people

KILL YOURSELF

Blame your language lad, something with 50k primitives is not an alphabet. Hebrew and Arabic deserve the gas too for the right-to-left bullshit.

Second Exodus. You?
I don't care. Kill yourself.


based


based


based

You've just described the weakest encryption scheme I've ever heard of.

Then you should go back to schizoposting wherever else and let the non retarded non mentally ill adults speak.

sad

This

Attached: Roadmap_to_Unicode.png (750x500, 73.55K)

Everyone knows RTL is superior fag

‮‮Stop being a nigger, please.

...

use a Chinese font on all documents because Chinese characters are supposed to be written as Chinese characters.

back to cuckchan with you


based


unbased


based


based


unbased

I'm based and I'm telling you to fuck off.
You're sequential based/unbased rating of posts just shows your high inclination to rating systems.
Go to reddit or something and don't come back!

Aww, the newfag is now repeating words he doesn't understand. How cute.

retroactive unbased


Not anymore :^)

I've been using the word LARPing on this board for quite some time.
You have just outed yourself as a newfag.
unbased btw

Yeah, no, more like we have a constant stream of newfags from cuckchan Zig Forums like yourself who use "LARP" as a general insult along with retarded shit like replying to everyone telling us if you agree with them or not. Pro tip: nobody cares. If you have something interesting to say say it, otherwise stfu.

Wrong. I'm responsible for at least 80% of the usage of the the word LARPer on Zig Forums.
Stop LARPing as an oldfag.

Character combiners... en.wikipedia.org/wiki/Ideographic_Description_Characters_(Unicode_block)
Fuck you hwite motherfuckers already made the solution, I can't even...

I was thinking of using it more for bypassing social media censorship, not so much going against the NSA.

Lol. Zig Forums has been using the term "larper" fucking constantly for ages. Stop doubling down and admit you got busted.

Attached: a927bc4feff5dcd035f837f4974d86648c2b10a5077087ead680d8adba8bc121.jpg (700x1002, 116.16K)

Yes, when accusing somebody of pretending to be something he isn't. I wasn't claiming anything about myself in the post he replied to. Using LARPer as a general "no u" is more of a thing done by 4cuck Zig Forums schizoids. There are plenty of schizos on here too, but they have other mannerisms.

Your also responsible for 100% of the based spam you faggot.

correct

it will be fixed quickly if you use it to say those "nazi" things

"Social media" is designed to extract as much info about yourself as possible. I wouldn't use it at all.

Pretty silly idea. Some search engines already do things like find words with an "ä" in them when you typed something with an "a" (and have been doing that for years). Maybe you can find some other substitutions than one letter -> one glyph, so that the text can only be recognized by looking at whole words.

Better transform the output into an image and add mild distortions. Feature idea: When making a text-pic you can set a key word, put some letters in brackets when typing, these letters disappear in the image generated, people looking at the picture in the same program can enter the key word to reveal the full text. The information is hidden in image noise. Though, if you are more lazy and don't want to do picture generation and you just go with UTF-8 and each letter being replaced with one glyph, given that there are so many replacement candidates for each English letter, you can just use that to encode the hidden letters.

𓏎𓏎𓏎𓏎𓏎𓏎𓏎𓏎
alt-codes.net/gender-symbol.php
alt-codes.net/snowflakes-symbols.php

⚦ What is this even supposed to mean?

The snowflakes would be good for whether reporting but the gender symbols?
⚤ is also fucking gay. Why is there a bisexuality symbol and not one symbol with the male arrow going through the female circle?
Fucking niggercattle.

𓂺
en.wiktionary.org/wiki/Appendix:Unicode/Egyptian_Hieroglyphs

Have no font that contains that symbol or the other two penis symbols. I'm on Windows Server 2019. I call censorship!

I did one a year or two ago, but no one was interested. Hard disk was shoahed since so lost it.
IIRC it was a one liner using a graphic library.