ChangeLog 24 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699
  1. 2.6.1
  2. =====
  3. * [CVE-2013-1445] Fix PRNG not correctly reseeded in some situations.
  4. In previous versions of PyCrypto, the Crypto.Random PRNG exhibits a
  5. race condition that may cause forked processes to generate identical
  6. sequences of 'random' numbers.
  7. This is a fairly obscure bug that will (hopefully) not affect many
  8. applications, but the failure scenario is pretty bad. Here is some
  9. sample code that illustrates the problem:
  10. from binascii import hexlify
  11. import multiprocessing, pprint, time
  12. import Crypto.Random
  13. def task_main(arg):
  14. a = Crypto.Random.get_random_bytes(8)
  15. time.sleep(0.1)
  16. b = Crypto.Random.get_random_bytes(8)
  17. rdy, ack = arg
  18. rdy.set()
  19. ack.wait()
  20. return "%s,%s" % (hexlify(a).decode(),
  21. hexlify(b).decode())
  22. n_procs = 4
  23. manager = multiprocessing.Manager()
  24. rdys = [manager.Event() for i in range(n_procs)]
  25. acks = [manager.Event() for i in range(n_procs)]
  26. Crypto.Random.get_random_bytes(1)
  27. pool = multiprocessing.Pool(processes=n_procs,
  28. initializer=Crypto.Random.atfork)
  29. res_async = pool.map_async(task_main, zip(rdys, acks))
  30. pool.close()
  31. [rdy.wait() for rdy in rdys]
  32. [ack.set() for ack in acks]
  33. res = res_async.get()
  34. pprint.pprint(sorted(res))
  35. pool.join()
  36. The output should be random, but it looked like this:
  37. ['c607803ae01aa8c0,2e4de6457a304b34',
  38. 'c607803ae01aa8c0,af80d08942b4c987',
  39. 'c607803ae01aa8c0,b0e4c0853de927c4',
  40. 'c607803ae01aa8c0,f0362585b3fceba4']
  41. This release fixes the problem by resetting the rate-limiter when
  42. Crypto.Random.atfork() is invoked. It also adds some tests and a
  43. few related comments.
  44. 2.6
  45. ===
  46. * [CVE-2012-2417] Fix LP#985164: insecure ElGamal key generation.
  47. (thanks: Legrandin)
  48. In the ElGamal schemes (for both encryption and signatures), g is
  49. supposed to be the generator of the entire Z^*_p group. However, in
  50. PyCrypto 2.5 and earlier, g is more simply the generator of a random
  51. sub-group of Z^*_p.
  52. The result is that the signature space (when the key is used for
  53. signing) or the public key space (when the key is used for encryption)
  54. may be greatly reduced from its expected size of log(p) bits, possibly
  55. down to 1 bit (the worst case if the order of g is 2).
  56. While it has not been confirmed, it has also been suggested that an
  57. attacker might be able to use this fact to determine the private key.
  58. Anyone using ElGamal keys should generate new keys as soon as practical.
  59. Any additional information about this bug will be tracked at
  60. https://bugs.launchpad.net/pycrypto/+bug/985164
  61. * Huge documentation cleanup (thanks: Legrandin).
  62. * Added more tests, including test vectors from NIST 800-38A
  63. (thanks: Legrandin)
  64. * Remove broken MODE_PGP, which never actually worked properly.
  65. A new mode, MODE_OPENPGP, has been added for people wishing to write
  66. OpenPGP implementations. Note that this does not implement the full
  67. OpenPGP specification, only the "OpenPGP CFB mode" part of that
  68. specification.
  69. https://bugs.launchpad.net/pycrypto/+bug/996814
  70. * Fix: getPrime with invalid input causes Python to abort with fatal error
  71. https://bugs.launchpad.net/pycrypto/+bug/988431
  72. * Fix: Segfaults within error-handling paths
  73. (thanks: Paul Howarth & Dave Malcolm)
  74. https://bugs.launchpad.net/pycrypto/+bug/934294
  75. * Fix: Block ciphers allow empty string as IV
  76. https://bugs.launchpad.net/pycrypto/+bug/997464
  77. * Fix DevURandomRNG to work with Python3's new I/O stack.
  78. (thanks: Sebastian Ramacher)
  79. * Remove automagic dependencies on libgmp and libmpir, let the caller
  80. disable them using args.
  81. * Many other minor bug fixes and improvements (mostly thanks to Legrandin)
  82. 2.5
  83. ===
  84. * Added PKCS#1 encryption schemes (v1.5 and OAEP). We now have
  85. a decent, easy-to-use non-textbook RSA implementation. Yay!
  86. * Added PKCS#1 signature schemes (v1.5 and PSS). v1.5 required some
  87. extensive changes to Hash modules to contain the algorithm specific
  88. ASN.1 OID. To that end, we now always have a (thin) Python module to
  89. hide the one in pure C.
  90. * Added 2 standard Key Derivation Functions (PBKDF1 and PBKDF2).
  91. * Added export/import of RSA keys in OpenSSH and PKCS#8 formats.
  92. * Added password-protected export/import of RSA keys (one old method
  93. for PKCS#8 PEM only).
  94. * Added ability to generate RSA key pairs with configurable public
  95. exponent e.
  96. * Added ability to construct an RSA key pair even if only the private
  97. exponent d is known, and not p and q.
  98. * Added SHA-2 C source code (fully from Lorenz Quack).
  99. * Unit tests for all the above.
  100. * Updates to documentation (both inline and in Doc/pycrypt.rst)
  101. * All of the above changes were put together by Legrandin (Thanks!)
  102. * Minor bug fixes (setup.py and tests).
  103. 2.4.1
  104. =====
  105. * Fix "error: Setup script exited with error: src/config.h: No such file or
  106. directory" when installing via easy_install. (Sebastian Ramacher)
  107. 2.4
  108. ===
  109. * Python 3 support! (Thorsten E. Behrens, Anders Sundman)
  110. PyCrypto now supports every version of Python from 2.1 through 3.2.
  111. * Timing-attack countermeasures in _fastmath: When built against
  112. libgmp version 5 or later, we use mpz_powm_sec instead of mpz_powm.
  113. This should prevent the timing attack described by Geremy Condra at
  114. PyCon 2011:
  115. http://blip.tv/pycon-us-videos-2009-2010-2011/pycon-2011-through-the-side-channel-timing-and-implementation-attacks-in-python-4897955
  116. * New hash modules (for Python >= 2.5 only): SHA224, SHA384, and
  117. SHA512 (Frédéric Bertolus)
  118. * Configuration using GNU autoconf. This should help fix a bunch of
  119. build issues.
  120. * Support using MPIR as an alternative to GMP.
  121. * Improve the test command in setup.py, by allowing tests to be
  122. performed on a single sub-package or module only. (Legrandin)
  123. You can now do something like this:
  124. python setup.py test -m Hash.SHA256 --skip-slow-tests
  125. * Fix double-decref of "counter" when Cipher object initialisation
  126. fails (Ryan Kelly)
  127. * Apply patches from Debian's python-crypto 2.3-3 package (Jan
  128. Dittberner, Sebastian Ramacher):
  129. - fix-RSA-generate-exception.patch
  130. - epydoc-exclude-introspect.patch
  131. - no-usr-local.patch
  132. * Fix launchpad bug #702835: "Import key code is not compatible with
  133. GMP library" (Legrandin)
  134. * More tests, better documentation, various bugfixes.
  135. 2.3
  136. ===
  137. * Fix NameError when attempting to use deprecated getRandomNumber()
  138. function.
  139. * _slowmath: Compute RSA u parameter when it's not given to
  140. RSA.construct. This makes _slowmath behave the same as _fastmath in
  141. this regard.
  142. * Make RSA.generate raise a more user-friendly exception message when
  143. the user tries to generate a bogus-length key.
  144. 2.2
  145. ===
  146. * Deprecated Crypto.Util.number.getRandomNumber(), which had confusing
  147. semantics. It's been replaced by getRandomNBitInteger and
  148. getRandomInteger. (Thanks: Lorenz Quack)
  149. * Better isPrime() and getPrime() implementations that do a real
  150. Rabin-Miller probabilistic primality test (not the phony test we did
  151. before with fixed bases). (Thanks: Lorenz Quack)
  152. * getStrongPrime() implementation for generating RSA primes.
  153. (Thanks: Lorenz Quack)
  154. * Support for importing and exporting RSA keys in DER and PEM format.
  155. (Thanks: Legrandin)
  156. * Fix PyCrypto when floor division (python -Qnew) is enabled.
  157. * When building using gcc, use -std=c99 for compilation. This should
  158. fix building on FreeBSD and NetBSD.
  159. 2.1.0
  160. =====
  161. * Fix building PyCrypto on Win64 using MS Visual Studio 9.
  162. (Thanks: Nevins Bartolomeo.)
  163. 2.1.0beta1
  164. ==========
  165. * Modified RSA.generate() to ensure that e is coprime to p-1 and q-1.
  166. Apparently, RSA.generate was capable of generating unusable keys.
  167. 2.1.0alpha2
  168. ===========
  169. * Modified isPrime() to release the global interpreter lock while
  170. performing computations. (patch from Lorenz Quack)
  171. * Release the GIL while encrypting, decrypting, and hashing (but not
  172. during initialization or finalization).
  173. * API changes:
  174. - Removed RandomPoolCompat and made Crypto.Util.randpool.RandomPool
  175. a wrapper around Crypto.Random that emits a DeprecationWarning.
  176. This is to discourage developers from attempting to provide
  177. backwards compatibility for systems where there are NO strong
  178. entropy sources available.
  179. - Added Crypto.Random.get_random_bytes(). This should allow people
  180. to use something like this if they want backwards-compatibility:
  181. try:
  182. from Crypto.Random import get_random_bytes
  183. except ImportError:
  184. try:
  185. from os import urandom as get_random_bytes
  186. except ImportError:
  187. get_random_bytes = open("/dev/urandom", "rb").read
  188. - Implemented __ne__() on pubkey, which fixes the following broken
  189. behaviour:
  190. >>> pk.publickey() == pk.publickey()
  191. True
  192. >>> pk.publickey() != pk.publickey()
  193. True
  194. (patch from Lorenz Quack)
  195. - Block ciphers created with MODE_CTR can now operate on strings of
  196. any size, rather than just multiples of the underlying cipher's
  197. block size.
  198. - Crypto.Util.Counter objects now raise OverflowError when they wrap
  199. around to zero. You can override this new behaviour by passing
  200. allow_wraparound=True to Counter.new()
  201. 2.1.0alpha1
  202. ===========
  203. * This version supports Python versions 2.1 through 2.6.
  204. * Clarified copyright status of much of the existing code by tracking
  205. down Andrew M. Kuchling, Barry A. Warsaw, Jeethu Rao, Joris Bontje,
  206. Mark Moraes, Paul Swartz, Robey Pointer, and Wim Lewis and getting
  207. their permission to clarify the license/public-domain status of their
  208. contributions. Many thanks to all involved!
  209. * Replaced the test suite with a new, comprehensive package
  210. (Crypto.SelfTest) that includes documentation about where its test
  211. vectors came from, or how they were derived.
  212. Use "python setup.py test" to run the tests after building.
  213. * API changes:
  214. - Added Crypto.version_info, which from now on will contain version
  215. information in a format similar to Python's sys.version_info.
  216. - Added a new random numbers API (Crypto.Random), and deprecated the
  217. old one (Crypto.Util.randpool.RandomPool), which was misused more
  218. often than not.
  219. The new API is used by invoking Crypto.Random.new() and then just
  220. reading from the file-like object that is returned.
  221. CAVEAT: To maintain the security of the PRNG, you must call
  222. Crypto.Random.atfork() in both the parent and the child processes
  223. whenever you use os.fork(). Otherwise, the parent and child will
  224. share copies of the same entropy pool, causing them to return the
  225. same results! This is a limitation of Python, which does not
  226. provide readily-accessible hooks to os.fork(). It's also a
  227. limitation caused by the failure of operating systems to provide
  228. sufficiently fast, trustworthy sources of cryptographically-strong
  229. random numbers.
  230. - Crypto.PublicKey now raises ValueError/TypeError/RuntimeError
  231. instead of the various custom "error" exceptions
  232. - Removed the IDEA and RC5 modules due to software patents. Debian
  233. has been doing this for a while
  234. - Added Crypto.Random.random, a strong version of the standard Python
  235. 'random' module.
  236. - Added Crypto.Util.Counter, providing fast counter implementations
  237. for use with CTR-mode ciphers.
  238. * Bug fixes:
  239. - Fixed padding bug in SHA256; this resulted in bad digests whenever
  240. (the number of bytes hashed) mod 64 == 55.
  241. - Fixed a 32-bit limitation on the length of messages the SHA256 module
  242. could hash.
  243. - AllOrNothing: Fixed padding bug in digest()
  244. - Fixed a bad behaviour of the XOR cipher module: It would silently
  245. truncate all keys to 32 bytes. Now it raises ValueError when the
  246. key is too long.
  247. - DSA: Added code to enforce FIPS 186-2 requirements on the size of
  248. the prime p
  249. - Fixed the winrandom module, which had been omitted from the build
  250. process, causing security problems for programs that misuse RandomPool.
  251. - Fixed infinite loop when attempting to generate RSA keys with an
  252. odd number of bits in the modulus. (Not that you should do that.)
  253. * Clarified the documentation for Crypto.Util.number.getRandomNumber.
  254. Confusingly, this function does NOT return N random bits; It returns
  255. a random N-bit number, i.e. a random number between 2**(N-1) and (2**N)-1.
  256. Note that getRandomNumber is for internal use only and may be
  257. renamed or removed in future releases.
  258. * Replaced RIPEMD.c with a new implementation (RIPEMD160.c) to
  259. alleviate copyright concerns.
  260. * Replaced the DES/DES3 modules with ones based on libtomcrypt-1.16 to
  261. alleviate copyright concerns.
  262. * Replaced Blowfish.c with a new implementation to alleviate copyright
  263. concerns.
  264. * Added a string-XOR implementation written in C (Crypto.Util.strxor)
  265. and used it to speed up Crypto.Hash.HMAC
  266. * Converted documentation to reStructured Text.
  267. * Added epydoc configuration Doc/epydoc-config
  268. * setup.py now emits a warning when building without GMP.
  269. * Added pct-speedtest.py to the source tree for doing performance
  270. testing on the new code.
  271. * Cleaned up the code in several places.
  272. 2.0.1
  273. =====
  274. * Fix SHA256 and RIPEMD on AMD64 platform.
  275. * Deleted Demo/ directory.
  276. * Add PublicKey to Crypto.__all__
  277. 2.0
  278. ===
  279. * Added SHA256 module contributed by Jeethu Rao, with test data
  280. from Taylor Boon.
  281. * Fixed AES.c compilation problems with Borland C.
  282. (Contributed by Jeethu Rao.)
  283. * Fix ZeroDivisionErrors on Windows, caused by the system clock
  284. not having enough resolution.
  285. * Fix 2.1/2.2-incompatible use of (key not in dict),
  286. pointed out by Ian Bicking.
  287. * Fix FutureWarning in Crypto.Util.randpool, noted by James P Rutledge.
  288. 1.9alpha6
  289. =========
  290. * Util.number.getPrime() would inadvertently round off the bit
  291. size; if you asked for a 129-bit prime or 135-bit prime, you
  292. got a 128-bit prime.
  293. * Added Util/test/prime_speed.py to measure the speed of prime
  294. generation, and PublicKey/test/rsa_speed.py to measure
  295. the speed of RSA operations.
  296. * Merged the _rsa.c and _dsa.c files into a single accelerator
  297. module, _fastmath.c.
  298. * Speed improvements: Added fast isPrime() function to _fastmath,
  299. cutting the time to generate a 1024-bit prime by a factor of 10.
  300. Optimized the C version of RSA decryption to use a longer series
  301. of operations that's roughly 3x faster than a single
  302. exponentiation. (Contributed by Joris Bontje.)
  303. * Added support to RSA key objects for blinding and unblinding
  304. data. (Contributed by Joris Bontje.)
  305. * Simplified RSA key generation: hard-wired the encryption
  306. exponent to 65537 instead of generating a random prime;
  307. generate prime factors in a loop until the product
  308. is large enough.
  309. * Renamed cansign(), canencrypt(), hasprivate(), to
  310. can_sign, can_encrypt, has_private. If people shriek about
  311. this change very loudly, I'll add aliases for the old method
  312. names that log a warning and call the new method.
  313. 1.9alpha5
  314. =========
  315. * Many randpool changes. RandomPool now has a
  316. randomize(N:int) method that can be called to get N
  317. bytes of entropy for the pool (N defaults to 0,
  318. which 'fills up' the pool's entropy) KeyboardRandom
  319. overloads this method.
  320. * Added src/winrand.c for Crypto.Util.winrandom and
  321. now use winrandom for _randomize if possible.
  322. (Calls Windows CryptoAPI CryptGenRandom)
  323. * Several additional places for stirring the pool,
  324. capturing inter-event entropy when reading/writing,
  325. stirring before and after saves.
  326. * RandomPool.add_event now returns the number of
  327. estimated bits of added entropy, rather than the
  328. pool entropy itself (since the pool entropy is
  329. capped at the number of bits in the pool)
  330. * Moved termios code from KeyboardRandomPool into a
  331. KeyboardEntry class, provided a version for Windows
  332. using msvcrt.
  333. * Fix randpool.py crash on machines with poor timer resolution.
  334. (Reported by Mark Moraes and others.)
  335. * If the GNU GMP library is available, two C extensions will be
  336. compiled to speed up RSA and DSA operations. (Contributed by
  337. Paul Swartz.)
  338. * DES3 with a 24-byte key was broken; now fixed.
  339. (Patch by Philippe Frycia.)
  340. 1.9alpha4
  341. =========
  342. * Fix compilation problem on Windows.
  343. * HMAC.py fixed to work with pre-2.2 Pythons
  344. * setup.py now dies if built with Python 1.x
  345. 1.9alpha3
  346. =========
  347. * Fix a ref-counting bug that caused core dumps.
  348. (Reported by Piers Lauder and an anonymous SF poster.)
  349. 1.9alpha2
  350. =========
  351. * (Backwards incompatible) The old Crypto.Hash.HMAC module is
  352. gone, replaced by a copy of hmac.py from Python 2.2's standard
  353. library. It will display a warning on interpreter versions
  354. older than 2.2.
  355. * (Backwards incompatible) Restored the Crypto.Protocol package,
  356. and modernized and tidied up the two modules in it,
  357. AllOrNothing.py and Chaffing.py, renaming various methods
  358. and changing the interface.
  359. * (Backwards incompatible) Changed the function names in
  360. Crypto.Util.RFC1751.
  361. * Restored the Crypto.PublicKey package at user request. I
  362. think I'll leave it in the package and warn about it in the
  363. documentation. I hope that eventually I can point to
  364. someone else's better public-key code, and at that point I
  365. may insert warnings and begin the process of deprecating
  366. this code.
  367. * Fix use of a Python 2.2 C function, replacing it with a
  368. 2.1-compatible equivalent. (Bug report and patch by Andrew
  369. Eland.)
  370. * Fix endianness bugs that caused test case failures on Sparc,
  371. PPC, and doubtless other platforms.
  372. * Fixed compilation problem on FreeBSD and MacOS X.
  373. * Expanded the test suite (requires Sancho, from
  374. http://www.mems-exchange.org/software/sancho/)
  375. * Added lots of docstrings, so 'pydoc Crypto' now produces
  376. helpful output. (Open question: maybe *all* of the documentation
  377. should be moved into docstrings?)
  378. * Make test.py automatically add the build/* directory to sys.path.
  379. * Removed 'inline' declaration from C functions. Some compilers
  380. don't support it, and Python's pyconfig.h no longer tells you whether
  381. it's supported or not. After this change, some ciphers got slower,
  382. but others got faster.
  383. * The C-level API has been changed to reduce the amount of
  384. memory-to-memory copying. This makes the code neater, but
  385. had ambiguous performance effects; again, some ciphers got slower
  386. and others became faster. Probably this is due to my compiler
  387. optimizing slightly worse or better as a result.
  388. * Moved C source implementations into src/ from block/, hash/,
  389. and stream/. Having Hash/ and hash/ directories causes problems
  390. on case-insensitive filesystems such as Mac OS.
  391. * Cleaned up the C code for the extensions.
  392. 1.9alpha1
  393. =========
  394. * Added Crypto.Cipher.AES.
  395. * Added the CTR mode and the variable-sized CFB mode from the
  396. NIST standard on feedback modes.
  397. * Removed Diamond, HAVAL, MD5, Sapphire, SHA, and Skipjack. MD5
  398. and SHA are included with Python; the others are all of marginal
  399. usefulness in the real world.
  400. * Renamed the module-level constants ECB, CFB, &c., to MODE_ECB,
  401. MODE_CFB, as part of making the block encryption modules
  402. compliant with PEP 272. (I'm not sure about this change;
  403. if enough users complain about it, I might back it out.)
  404. * Made the hashing modules compliant with PEP 247 (not backward
  405. compatible -- the major changes are that the constructor is now
  406. MD2.new and not MD2.MD2, and the size of the digest is now
  407. given as 'digest_size', not 'digestsize'.
  408. * The Crypto.PublicKey package is no longer installed; the
  409. interfaces are all wrong, and I have no idea what the right
  410. interfaces should be.
  411. 1.1alpha2
  412. =========
  413. * Most importantly, the distribution has been broken into two
  414. parts: exportable, and export-controlled. The exportable part
  415. contains all the hashing algorithms, signature-only public key
  416. algorithms, chaffing & winnowing, random number generation, various
  417. utility modules, and the documentation.
  418. The export-controlled part contains public-key encryption
  419. algorithms such as RSA and ElGamal, and bulk encryption algorithms
  420. like DES, IDEA, or Skipjack. Getting this code still requires that
  421. you go through an access control CGI script, and denies you access if
  422. you're outside the US or Canada.
  423. * Added the RIPEMD hashing algorithm. (Contributed by
  424. Hirendra Hindocha.)
  425. * Implemented the recently declassified Skipjack block
  426. encryption algorithm. My implementation runs at 864 K/sec on a
  427. PII/266, which isn't particularly fast, but you're probably better off
  428. using another algorithm anyway. :)
  429. * A simple XOR cipher has been added, mostly for use by the
  430. chaffing/winnowing code. (Contributed by Barry Warsaw.)
  431. * Added Protocol.Chaffing and Hash.HMAC.py. (Contributed by
  432. Barry Warsaw.)
  433. Protocol.Chaffing implements chaffing and winnowing, recently
  434. proposed by R. Rivest, which hides a message (the wheat) by adding
  435. many noise messages to it (the chaff). The chaff can be discarded by
  436. the receiver through a message authentication code. The neat thing
  437. about this is that it allows secret communication without actually
  438. having an encryption algorithm, and therefore this falls within the
  439. exportable subset.
  440. * Tidied up randpool.py, and removed its use of a block
  441. cipher; this makes it work with only the export-controlled subset
  442. available.
  443. * Various renamings and reorganizations, mostly internal.
  444. 1.0.2
  445. =====
  446. * Changed files to work with Python 1.5; everything has been
  447. re-arranged into a hierarchical package. (Not backward compatible.)
  448. The package organization is:
  449. Crypto.
  450. Hash.
  451. MD2, MD4, MD5, SHA, HAVAL
  452. Cipher.
  453. ARC2, ARC4, Blowfish, CAST, DES, DES3, Diamond,
  454. IDEA, RC5, Sapphire
  455. PublicKey.
  456. DSA, ElGamal, qNEW, RSA
  457. Util.
  458. number, randpool, RFC1751
  459. Since this is backward-incompatible anyway, I also changed
  460. module names from all lower-case to mixed-case: diamond -> Diamond,
  461. rc5 -> RC5, etc. That had been an annoying inconsistency for a while.
  462. * Added CAST5 module contributed by <wiml@hhhh.org>.
  463. * Added qNEW digital signature algorithm (from the digisign.py
  464. I advertised a while back). (If anyone would like to suggest new
  465. algorithms that should be implemented, please do; I think I've got
  466. everything that's really useful at the moment, but...)
  467. * Support for keyword arguments has been added. This allowed
  468. removing the obnoxious key handling for Diamond and RC5, where the
  469. first few bytes of the key indicated the number of rounds to use, and
  470. various other parameters. Now you need only do something like:
  471. from Crypto.Cipher import RC5
  472. obj = RC5.new(key, RC5.ECB, rounds=8)
  473. (Not backward compatible.)
  474. * Various function names have been changed, and parameter
  475. names altered. None of these were part of the public interface, so it
  476. shouldn't really matter much.
  477. * Various bugs fixed, the test suite has been expanded, and
  478. the build process simplified.
  479. * Updated the documentation accordingly.
  480. 1.0.1
  481. =====
  482. * Changed files to work with Python 1.4 .
  483. * The DES and DES3 modules now automatically correct the
  484. parity of their keys.
  485. * Added R. Rivest's DES test (see http://theory.lcs.mit.edu/~rivest/destest.txt)
  486. 1.0.0
  487. =====
  488. * REDOC III succumbed to differential cryptanalysis, and has
  489. been removed.
  490. * The crypt and rotor modules have been dropped; they're still
  491. available in the standard Python distribution.
  492. * The Ultra-Fast crypt() module has been placed in a separate
  493. distribution.
  494. * Various bugs fixed.