From drewpc@colorado.edu Fri Jan 26 21:10:30 2001 From: drewpc@colorado.edu (PC Drew) Date: Fri, 26 Jan 2001 14:10:30 -0700 (MST) Subject: [meta-sig] encryptioin/compression sig Message-ID: Okay, so I've been using Python for about a year now and the more I use it the more I really like it (doing without ++ and -- is kinda hard, but I'll get by!). One thing that I've noticed is that there isn't a real big interest in compression or encryption. I notice this because there aren't any modules (or a small amount) that do this! I'm proposing a SIG that will encompass both cryptography and compression libraries in a clean API so that users can encrypt/decrypt or compress/decompress files or strings easily. In addition to creating the APIs, also wrapping C libraries of many of the popular algorithms. Feedback? If people are interested (or not adverse to it), then I'll draft up a formal proposal with a name, description, and the like. -- PC Drew "To understand recursion, we must first understand recursion." From akuchlin@mems-exchange.org Fri Jan 26 21:18:19 2001 From: akuchlin@mems-exchange.org (Andrew M. Kuchling) Date: Fri, 26 Jan 2001 16:18:19 -0500 Subject: [meta-sig] encryption/compression sig In-Reply-To: ; from drewpc@colorado.edu on Fri, Jan 26, 2001 at 02:10:30PM -0700 References: Message-ID: <20010126161819.A2762@amarok.cnri.reston.va.us> On Fri, Jan 26, 2001 at 02:10:30PM -0700, PC Drew wrote: >isn't a real big interest in compression or encryption. I notice this >because there aren't any modules (or a small amount) that do this! Python comes with a zlib module and some lame SSL support in the socket module (though both usually aren't built by default on Unix), and there are various extensions for crytographic algorithms: mxCrypto, my amkCrypto, and m2crypto, which wraps OpenSSL. However, this area has been slow to progress because the APIs are large -- OpenSSL's is huge! -- and various people seem to have little time to work on this. More help would be really appreciated. >Feedback? If people are interested (or not adverse to it), then I'll draft >up a formal proposal with a name, description, and the like. There's a very quiet python-crypto mailing list; see http://groups.yahoo.com/group/python-crypto/ --amk From jeremy@alum.mit.edu Fri Jan 26 21:19:26 2001 From: jeremy@alum.mit.edu (Jeremy Hylton) Date: Fri, 26 Jan 2001 16:19:26 -0500 (EST) Subject: [meta-sig] encryptioin/compression sig In-Reply-To: References: Message-ID: <14961.59870.171089.22545@localhost.localdomain> >>>>> "PCD" == PC Drew writes: PCD> Okay, so I've been using Python for about a year now and the PCD> more I use it the more I really like it (doing without ++ and PCD> -- is kinda hard, but I'll get by!). One thing that I've PCD> noticed is that there isn't a real big interest in compression PCD> or encryption. I notice this because there aren't any modules PCD> (or a small amount) that do this! The number is small but not zero. Some of the packages that exist provide a large collection of encryption algorithms. PCD> I'm proposing a SIG that will encompass both cryptography and PCD> compression libraries in a clean API so that users can PCD> encrypt/decrypt or compress/decompress files or strings easily. PCD> In addition to creating the APIs, also wrapping C libraries of PCD> many of the popular algorithms. I don't see that there is a need for a sig to accomplish. I also wonder if much of what you propose isn't already done. Have you looked at, for instance, mxCrypto/amkCrypto or M2Crypto? (The latter is at http://www.post1.com/home/ngps/m2.) PCD> Feedback? If people are interested (or not adverse to it), PCD> then I'll draft up a formal proposal with a name, description, PCD> and the like. If you have time to do work on crypto stuff, why don't you raise the issue on python-crytpo@egroups.com and see what needs the people there identify? There is no need to reinvent the wheel, but there is lots of opportunity for someone to contribute. Jeremy From PC Drew Wed Jan 31 00:14:27 2001 From: PC Drew (PC Drew) Date: Tue, 30 Jan 2001 17:14:27 -0700 Subject: [meta-sig] Re[2]: [python-crypto] Add AES to amkCrypto? In-Reply-To: <3A771DFA.71056D40@stroeder.com> References: <59317538836.20010130110942@colorado.edu> <3A771DFA.71056D40@stroeder.com> Message-ID: <54339427196.20010130171427@colorado.edu> I totally agree. That's why I suggested an crypto-sig...mostly to develop a standard API. In addition to that, I think that it is important to also develop an abstraction layer to go with the API so that, not only do things look the same, but if you wanted to go from one algorithm to another, it would be very easy. If anyone is interested in defining this with me, please email me and I'll get a list together and we can start hashing this out. I'm also interested in doing the same thing with compression libraries, and I'd like some feedback on that as well. -- PC Drew Tuesday, January 30, 2001, 1:03:06 PM, you wrote: MS> PC Drew wrote: >> >> I just read on Python-URL! that someone has released the new AES in >> python. (http://deja.com/=dnc/getdoc.xp?AN=720935577)  I also >> expressed interest in creating an encryption sig and was told of >> amkCrytpo and m2crypto. MS> IMHO the main problem for application and crypto lib developers is MS> the lack of a standard API. Crypto lib developers are designing new MS> APIs every time and application developers have to deal with the MS> different APIs. =>> there should be a standard class library which abstracts from MS> various algorithms and their implementation etc. MS> This would save time... MS> Ciao, Michael. MS> Yahoo! Groups Sponsor MS> www.    From guido@digicool.com Wed Jan 31 00:32:17 2001 From: guido@digicool.com (Guido van Rossum) Date: Tue, 30 Jan 2001 19:32:17 -0500 Subject: [meta-sig] Re[2]: [python-crypto] Add AES to amkCrypto? In-Reply-To: Your message of "Tue, 30 Jan 2001 17:14:27 MST." <54339427196.20010130171427@colorado.edu> References: <59317538836.20010130110942@colorado.edu> <3A771DFA.71056D40@stroeder.com> <54339427196.20010130171427@colorado.edu> Message-ID: <200101310032.TAA30149@cj20424-a.reston1.va.home.com> > If anyone is interested in defining this with me, please email me and > I'll get a list together and we can start hashing this out. I'm also > interested in doing the same thing with compression libraries, and I'd > like some feedback on that as well. Note that the meta-sig is about the least effective place to announce the desire to form a new SIG -- it is read by only a few people. It's a great place to discuss a proposed SIG once you have the critical mass -- but to get it going, I'd suggest rounding up people on c.l.py or c.l.py.a! --Guido van Rossum (home page: http://www.python.org/~guido/)