📄 pegwit.htm
字号:
<html>
<head><title> Pegwit (v8) </TITLE></head>
<h1> Pegwit (v8) </h1>
<body>
Pegwit is a program for performing public key encryption
and authentication. It uses an elliptic curve over GF(2^255), and
the symmetric block cipher
<a href="http://www.esat.kuleuven.ac.be/~rijmen/square/index.html">
square</a>. It is free, you may do
what you like with it, but if you make changes (which make the
software incompatible with pegwit), please use a different name
to avoid confusion. It is possible that using pegwit may
infringe U.S. patents, I don't know. (But see this
<a href="http://www.base.com/gordoni/thoughts/pkp.html">
essay on PKP's claim to have a monopoly on public key
cryptography </a> and <a href="http://bbs.cruzio.com/~schlafly/">
Roger Schlafly's page </a> ) In any case the offending patent,
Hellman-Merkle 4,218,582 expires Oct. 6, 1997.
<p> Here is some <a href="blurb.txt"> blurb</a> and also
some <a href="design.txt"> design information.</a>
<p> Here is the <a href="manual.txt"> user manual</a> and
here is a <a href="chall.htm"> challenge</a>
<p> If you can read zip files you can download all the sources
for pegwit by pressing <a href="pegwit.zip"> here.</a>
<p> The source containing main is
<a href="pegwit.c"> pegwit.c</a>.
The other sources are
<a href="ec_crypt.c">ec_crypt.c</a>,
<a href="ec_curve.c">ec_curve.c</a>,
<a href="ec_field.c">ec_field.c</a>,
<a href="ec_param.c">ec_param.c</a>,
<a href="ec_vlong.c">ec_vlong.c</a>,
<a href="square.c"> square.c</a>,
<a href="sqcts.c"> sqcts.c</a>,
<a href="sha1.c"> sha1.c</a> and
<a href="binasc.c"> binasc.c</a>.
Included source files are
<a href="ec_crypt.h">ec_crypt.h</a>,
<a href="ec_curve.h">ec_curve.h</a>,
<a href="ec_field.h">ec_field.h</a>,
<a href="ec_param.h">ec_param.h</a>,
<a href="ec_vlong.h">ec_vlong.h</a>,
<a href="sha1.h">sha1.h</a>,
<a href="square.h"> square.h</a>,
<a href="sqcts.h"> sqcts.h</a>,
<a href="square.tab"> square.tab</a> and
<a href="binasc.h"> binasc.h</a>.
<p> The source is intended to be completely portable, please let me
know of any problems so I can fix them. The only special action needed
is that either LITTLE_ENDIAN or BIG_ENDIAN may need to be defined externally.
A compile-time error is reported if neither is defined and the endianness
cannot be determined at compile time. A run-time error is reported if the
wrong one is defined.
<p> <a href="win95/pegwit.exe"> pegwit.exe</a> is a Windows 95 executable.
<a href="win95/pwkey.exe"> pwkey.exe</a> and
<a href="win95/pwjunk.exe"> pwjunk.exe</a> are optional Win95 utilities
for entering passwords and generating random junk.
<a href="pwkey.c">pwkey.c</a> and
<a href="pwjunk.c"> pwjunk.c</a> are (non-portable) sources
for these utilities.
<p> <a href="dos/pegwit.exe"> pegwit.exe</a> is an MSDOS executable.
<p> Test data
<p> <a href="test.bat"> test.bat</a> is a batch file for testing.
<p> <a href="test.pub"> test.pub</a> is the public key for
<a href="test.pri"> test.pri</a>
<p> <a href="test.tx0"> test.tx0</a> is <a href="test.txt"> test.txt</a>
encrypted with test.pub ( but note that -e is not deterministic )
<p> <a href="test.sig"> test.sig</a> is test.txt signed with test.pri (using -s)
<p> <a href="test.tx1"> test.tx1</a> is test.txt encrypted symmetrically
(-E) with test.sig
<p> <a href="test.tx2"> test.tx2</a> is test.txt signed with test.pri (using -S)
<p> <a href="test.tx3"> test.tx3</a> is test.txt encrypted with test.pri (using -fe)
<p> <a href="test.tx4"> test.tx4</a> is test.txt encrypted with test.sig (using -fE)
<p> Changes. Note that a change of first digit of the version number
indicates that the new version is incompatible with earlier versions.
However note that v8.7 is not encryption compatible with v8.1..v8.5.
<p> 8.0 has many changes - for one thing it is in C rather than
C++ ( thanks to Paulo Barreto ). Secret values are now read from
standard input, and a secret value is required for -e ( see user
manual for details ). The generation of multipliers has been
improved (v7 was not secure for > 2^64 signatures) and the MAC used
is now 'double-barreled' SHA1 ( 240 bits = 160 bits + 80 bits ).
Public keys are now partially validated : they must begin with
"pegwit v8 public key ="
<p> v8.1 : I messed up in v8.0 - prng_next was not hashing the
correct data, and while not insecure, it was a mess. So v8.0
should not be used, only use v8.1 (which is not compatible).
Hopefully no-one got a copy of 8.0.
<p> v8.1 : Re-worked prng to cure endian problems. ( No version
number change )
<p> pwjunk v2.0 : printf format string in pwjunk was wrong.
pwjunk revised to use kbhit and a timing loop which gives
better resolution than using clock(). Also now generates a
(rough) estimate of how much entropy has been gathered, and
does not allow termination until estimate reachs 128 bits.
<p> v8.1 : changed C++ style comments to C comments
<p> v8.2 : changed -e to use single barrel hash of plain text.
( signature still uses double barrel for MAC )
<p> v8.2 : deleted unused function cpMakePrivateKey from ec_crypt.c
<p> v8.2 : put conditional compilation round functions which are
not required for pegwit
<p> The code is now known to run OK on a big-endian processor
( HP700 series Unix box). Thanks to Steve Gilham for this.
<p> v8.3 : -S and -V commands added for signing text files.
These make pegwit more convenient when signing email or news.
Thanks to Mr. Tines for this.
<p> v8.4 : -fX 'filter mode' commands added, also ascii
'armoring' of encrypted input/output. Again, thanks to Mr. Tines.
<p> v8.4 : quite a few mainly cosmetic changes to ec_*.*
modules ( by Paulo Barreto ). There is a new source file
ec_param.c. Note that pegwit does not support different
curves (yet) - these are provided for direct uses
of ec_*.* in a non-pegwit context.
<p> v8.5 : signature verification was totally bust in v8.4
There were actually two seperate bugs :
(1) -v was not printing any error message when signature failed.
(2) -V was not doing anything at all (a boolean test was the wrong way round).
THUS PEGWIT v8.4 SHOULD NEVER BE USED TO CHECK SIGNATURES!
Many apologies for this major cock-up.
<p> v8.6 : changed to using new version of square. Unfortunately
I previously had not been doing the proper endian conversion,
so v8.6 is not encryption compatible with earlier versions.
Signatures and public/private keys functionality has not changed
though. v8.6 uses cipher-text stealing to minimise expansion when
encrypting. For -E encrypted files, provided the size is > 15 bytes
and even, there is no change in size. Encryption and decryption is
also faster in this version.
<p> v8.7 : updated to neater cipher-text stealing : not
encryption compatible with v8.6. Probably no-one got v8.6.
<p> v8.71 : Public keys enclosed in braces {} can be split up, and
need not be at the beginning of the file. New ec_*.* code, which
can be compiled to use Colin Plumb's BNLIB rather than ec_vlong.c
if desired (although there is no reason to do this for pegwit).
<p> If you port pegwit to other systems, or have any suggestions
or comments please let me know. Also let me know if you would like
to be notified by email of any major developments, e.g. major bugs,
major new versions, etc.
<p> Send mail to George at
<a href="mailto:george.barwood@dial.pipex.com">
<i>george.barwood@dial.pipex.com</i></a>
<p> <a href=../index.htm> George's home page </a>
<p> <a href=../crypto.htm> More crypto stuff (including more info on pegwit components) </a>
<p> <a href=../ec_faq.txt> My FAQ on elliptic curve cryptography </a>
<a href="../ec_faq.htm">HTML version</a>
<p>
<IMG SRC="/cgi-bin/nph-count?width=5&link=/town/square/abe26/v8/pegwit.htm">
</html>
⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -