ROT13

ROT13 replaces each letter by its partner 13 characters further along the alphabet. For example, HELLO becomes URYYB (or, decrypting, URYYB becomes HELLO again).

ROT13 ("rotate by 13 places", usually hyphenated ROT-13) is a simple Caesar cipher used for obscuring text by replacing each letter with the letter thirteen places down the alphabet. A becomes N, B becomes O, and so on up to M, which becomes Z, then the sequence reverses: N becomes A, O becomes B, and so on to Z, which becomes M. The algorithm is used in online forums as a means of hiding joke punchlines, puzzle solutions, movie and story spoilers, and offensive materials from the casual glance. ROT13 has been described as the "Usenet equivalent of a magazine printing the answer to a quiz upside down" [3]. The name "ROT13" originated on Usenet in the early 1980s, and the method has become a de facto standard. As a Caesar cipher (a method of encryption thousands of years old), ROT13 provides no real cryptographic security and is not used for such; in fact it is often used as the canonical example of weak encryption. An additional feature of the cipher is that it is symmetrical; that is, to undo ROT13, the same algorithm is applied, so the same code can be used for encoding and decoding.

Description

Applying ROT13 to a piece of text merely requires examining its alphabetic characters and replacing each one by the letter 13 places further along in the alphabet, wrapping back to the beginning if necessary, preserving case: a becomes n, B becomes O, and so forth, down to Z, which becomes M. Only those letters which occur in the Latin alphabet are affected; numbers, symbols, whitespace, and all other characters are left unchanged. Because there are 26 letters in the Latin alphabet and 26 = 2 × 13, the ROT13 function is its own inverse:

${\mbox{ROT}}_{13}({\mbox{ROT}}_{13}(x))={\mbox{ROT}}_{26}(x)=x$ for any text x.

In other words, two successive applications of ROT13 restore the original text (in mathematics, this is sometimes called an involution; in cryptography, a reciprocal cipher).

The transformation can be done using a lookup table, such as the following:

ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz

NOPQRSTUVWXYZABCDEFGHIJKLMnopqrstuvwxyzabcdefghijklm

For example, in the following joke, the answer (punchline) has been obscured by ROT13:

How can you tell an extrovert from an
introvert at NSA? Va gur ryringbef,
gur rkgebiregf ybbx ng gur BGURE thl'f fubrf. 

Transforming the entire text via ROT13 form, the answer to the joke is revealed:

Ubj pna lbh gryy na rkgebireg sebz na
vagebireg ng AFN? In the elevators,
the extroverts look at the OTHER guy's shoes.

A second application of ROT13 would restore the original.

History

ROT13 originated in the net.jokes newsgroup in the early 1980s, in an effort to provide a voluntary means to hide jokes that some readers might have found offensive, or just to obscure the punchline of a joke to keep it from being read too soon. (Previous attempts to categorize offensive jokes by putting them in separate newsgroups had failed — site managers did not want to be seen as condoning such postings by creating a special place for them.) ROT13 was a convenient solution because of its simplicity.

Since it replaces alphabetic letters with other letters, ROT13 was guaranteed not to cause problems for any newsgroup software which might have been incompatible with unusual kinds of characters. ROT-13 was chosen over ROT-N (for any other value of N, such as 3 as in the original Caesar cipher) because 13 is the value which arranges that encoding and decoding are equivalent, thereby allowing the convenience of a single command for both encryption and decryption. Strictly speaking, the value 13 works in this way only for languages like English which have 26 letters in their alphabets, although since the 26-character Latin alphabet is embedded in the ASCII character set, ROT13 is pretty much universal, at least for those Western languages which can use ASCII or one of its extensions.

While users could plausibly encode and decode messages by hand, automatic decryption is much more convenient. UNIX systems have a standard utility called "tr" (transliterate) that can be used to perform ROT13 encoding:

tr A-Za-z N-ZA-Mn-za-m

Automatic deciphering was soon added as a built-in feature to newsreading software. From the early 1990s, ROT13 was also used in Fidonet forums, to the extent that Fidonet mail reading software often included features for enciphering and deciphering automatically.

ROT13 as encryption

ROT13 is one of a set of encryption algorithms known as Caesar ciphers, which are themselves a particularly simple type of substitution cipher. ROT13 is not intended to be used where secrecy is of any concern—the use of a constant shift means that the encryption effectively has no key, and decryption requires no more knowledge than the fact that ROT13 is in use. Even without this knowledge, like any substitution cipher the algorithm is easily broken through frequency analysis or the exploitation of pattern words. (ROT13 is an example of what Bruce Schneier calls "kid sister encryption" -- at best it might keep a document secret from your kid sister.)

The real effect of ROT13 is merely to make sure that the viewer of a message must consciously choose to decipher it, which typically means invoking the relevant ROT13 command in the message-reading software. Rather than protecting a confidential message from unauthorized readers, ROT13 safeguards authorized readers from material they may not wish to inadvertently read, such as spoilers in book or movie reviews.

Because of its utter unsuitability for real secrecy, ROT13 has become a catchphrase to refer to any conspicuously weak encryption scheme; a critic might claim that "56-bit DES is little better than ROT13 these days." Also, in a play on real terms like "double DES", the terms double ROT13, ROT26 or 2ROT13 crop up with humorous intent, including a spoof academic paper "On the 2ROT13 Encryption Algorithm" (PDF). As explained above, applying ROT13 to an already ROT13-encrypted text restores the original plaintext, so ROT26 is equivalent to no encryption at all. The notional ROT26 has been used as a satirical jibe against the United States' Digital Millennium Copyright Act (DMCA); some online users append phrases such as "Encoded with ROT26 — circumvention will be prosecuted!" to their posts to online forums. (The DMCA introduced a broad ban on the circumvention of copy prevention systems, systems which are often found to internally employ woefully insecure methods of cryptography.) "Triple ROT13" is also sometimes seen - which is, of course, equivalent to regular ROT13. (The term is most probably a take on Triple DES.)

In 2001, Russian cryptanalyst Dmitry Sklyarov was arrested after detailing weaknesses in ebook copy prevention systems. One ebook vendor, New Paradigm Research Group (NPRG), actually used ROT13 to encrypt their documents. It has been speculated that NPRG may have mistaken the ROT13 toy example — provided with the Adobe eBook software development kit — for a serious encryption scheme [4].

Windows XP uses ROT13 on some of its registry keys (UserAssist). Didier Stevens wrote a tool to decrypt them. These keys track recently used programs.

ROT13 letter games and net culture

abcdefghijklmnopqrstuvwxyz NOPQRSTUVWXYZABCDEFGHIJKLM
aha ↔ nun	ant ↔ nag
balk ↔ onyx	bar ↔ one
barf ↔ ones	be ↔ or
bin ↔ ova	ebbs ↔ roof
envy ↔ rail	er ↔ re
errs ↔ reef	flap ↔ sync
fur ↔ she	gel ↔ try
gnat ↔ tang	irk ↔ vex
clerk ↔ pyrex	purely ↔ cheryl
PNG ↔ cat	SHA ↔ fun

ROT13 provides an opportunity for letter games. Some words will, when transformed with ROT13, produce another word. The longest example in the English language is the pair of 7-letter words abjurer and nowhere; there is also the 7-letter pair chechen and purpura. Other examples of words like these are shown in the table.

The word ravine is transformed under ROT13 to enivar — its reversal, although not an English word. The only known pair of common English words which are each other's ROT13 and reversal are gnat and tang.

vex is transformed to irk under ROT13; the words are synonyms. Also, terra can be transformed to green to give a satisfying pair.

The 1989 International Obfuscated C Code Contest (IOCCC) included an entry by Brian Westley. Westley's computer program can be ROT13'd or reversed and still compiles correctly. Its operation, when executed, is to either perform ROT13 encoding on or to reverse its input [5],[6].

The newsgroup alt.folklore.urban coined a word — furrfu — that was the ROT13 encoding of the frequently encoded utterance "sheesh". "Furrfu" evolved in mid-1992 as a response to postings repeating urban myths on alt.folklore.urban, after some posters complained that "Sheesh!" as a response to newcomers was being overused [7].

Variants

There are other little-used obfuscation schemes with a similar purpose to ROT13. ROT13 only handles letters, and leaves other characters, such as punctuation, numbers and whitespace, untouched. Because of this, ROT13 may be unsuitable for some purposes, for example, hiding the numeric answer to a riddle, or handling arbitrary binary data.

ROT47

ROT47 is a variant on ROT13 which, in addition to scrambling the basic letters, also treats numbers and many other characters. Instead of using the sequence A-Z as the alphabet, ROT47 uses a larger alphabet, derived from a common character encoding known as ASCII. ASCII maps letters, digits, punctuation and other special characters to numbers in the range 0–127. Viewed in ASCII, ROT13 covers the codes 65–90 and 97–122 — the upper and lower case letters respectively. ROT47 uses 94 characters instead, from ! (the exclamation mark, ASCII code 33) to ~ (the tilde, ASCII code 126), rotating them by an offset of 47. The use of a larger alphabet is intended to produce a more thorough obfuscation than that of ROT13, but ROT47 is far less widely supported.

Applying ROT47 to the earlier example (entirely in plaintext) gives:

w@H 42? J@F E6== 2? 6IEC@G6CE 7C@> 2?
:?EC@G6CE 2E }$pn x? E96 6=6G2E@CD[
E96 6IEC@G6CED =@@< 2E E96 ~%wt# 8FJVD D9@6D]

The Unix command to apply ROT47 is:

tr '!-~' 'P-~!-O'

memfrob()

The GNU C library (a set of standard routines available for use in computer programming) contains a function — memfrob()[8] — which has a similar purpose to ROT13, although it is intended for use with arbitrary binary data. The function operates (humorously termed "frobnicates") by combining each byte with the binary pattern 00101010 (42 decimal) using the exclusive or (XOR) operation. This effects a simple XOR cipher. Like ROT13, memfrob() is self-reciprocal; it is also just about as insecure (that is, just as trivial to decrypt if the key is not known).

See also

• Cryptanalysis

References

• Eric S. Raymond (Editor), the Jargon File, [9].
• Bruce Schneier, Applied Cryptography, 2nd edition, Wiley, 1996, p11. ISBN 0-471-11709-9.

External links

• A description of ROT13 — on the Everything2 website.
• Words with interesting properties under ROT13
• Software for ROT13 in a large number of languages — includes a patch to SSH to add support for ROT13, and a cryptanalysis tool to automatically distinguish ROT13 text from plaintext.
• Online ROT13 encoding
• Another online ROT13 encoder
• Online ROT47 encoder
• Browse the web in ROT13
• Didier Stevens Registry Decrypter