Re: [viff-devel] My dissertation
Hi Martin Thanks for your prompt reply. Much appreciated :) If anyone on the list has any further pointers, I will be grateful for your response. Thanks in advance Kind Regards Sriram From: m...@mgsys.dk [...@mgsys.dk] On Behalf Of Martin Geisler [...@lazybytes.net] Sent: 23 April 2010 18:04 To: Srinivasan S Mr (Computing); viff-devel@viff.dk Subject: Re: My dissertation writes: Hi Sriram, I'll CC this message to the VIFF mailinglist as well -- there are many people there who can help you if I cannot. Please keep the list as CC on your replies. > Hi Martin > > I hope you have settled into your new workplace and enjoying yourself. Thanks, it's going well in my new job. > I had a few quick queries if you don't mind. > > I have been trying to get a grasp on MPC ideas over the past few days. > I was wondering if you could point me a recent comprehensive, but not > too heavy survey on MPC. If I could get an idea of how the field has > progressed and various approaches to achieving MPC without getting > bogged down in too much detail, it may be helpful for me at this > stage. I don't know of such a survey off the top of my head. > As I understand, VIFF implements the underlying functionality using > Shamir Secret Sharing. It seems there are other approaches possible as > well and one of my colleagues wanted to know if VIFF also implements > the ideas from the paper "Multiparty Computation from Threshold > Homomorphic Encryption by Ronald Cramer and Ivan Damgard and Jesper > Buus Nielsen"? My guess is No, but I am not not sure. Right, we have not implemented the protocols from that paper. > There are some papers that refer to this latter work, for example > "Efficient Binary Conversion for Paillier Encrypted Values by Berry > Schoenmakers and Pim Tuyls". I was wondering if the application in > this paper can be coded with the functionality provided by VIFF in > principle, or there are some details which may prevent it from being > so. I'm sorry, but I'm not familiar with that paper. But you should think of VIFF as having a number of layers: 1. basic player administration 2. network setup 3. secret sharing and reconstruction (Shamir-based) 4. secure addition and multiplication 5. more complex protocols So depending on how the primitive above fits into the stack, you may be able to reuse more or less code. > Sorry if my queries don't make sense. I can admit I don't have much > knowledge of the details and I am trying pick up stuff as I go along. > > Thanks in advance. > > Best Wishes > > Sriram -- Martin Geisler ___ viff-devel mailing list (http://viff.dk/) viff-devel@viff.dk http://lists.viff.dk/listinfo.cgi/viff-devel-viff.dk
Re: [viff-devel] Query from the University of Surrey
Dear Thomas Yes. It works fine now. Thanks for your help. Kind Regards Sriram From: thomas@gmail.com [thomas@gmail.com] On Behalf Of Thomas P Jakobsen [...@cs.au.dk] Sent: 23 February 2010 12:44 To: Srinivasan S Mr (Computing) Cc: viff-devel Subject: Re: [viff-devel] Query from the University of Surrey > > When I run the Millionaires example (again with --no-ssl option) with the > Path set to point to Python 2.6.4, I get the following error message. > > Traceback (most recent call last): > File "generate-config-files.py", line 57, in >from viff.config import generate_configs > ImportError: No module named viff.config > This exception typically shows up if VIFF is not on the PYTHONPATH. When Python sees an "import" statement, it looks for the needed packages on the PYTHONPATH. The solution would be to either 1) Download and install the VIFF .exe installer from viff.dk or 2) Download the VIFF source code from viff.dk and run "setup.py install" when you are in the root of the source tree. These steps will both install VIFF in C:\Python26\Lib\site-packages which is always on the PYTHONPATH. A third alternative would be to include the path to the VIFF source code in an environment variable called PYTHONPATH. Hope this helps, Thomas ___ viff-devel mailing list (http://viff.dk/) viff-devel@viff.dk http://lists.viff.dk/listinfo.cgi/viff-devel-viff.dk
Re: [viff-devel] Query from the University of Surrey
Hi Thomas I now have the following 2 setups on my machine (Windows XP Professional, SP3) The first is Python 2.4.4 Twisted_NoDocs-8.1.0.win32-py2.4.exe gmpy-1.10-beta.win32-py2.4 Win32OpenSSL_Light-0_9_8l.exe pyOpenSSL-0.7.win32-py2.4.exe This setup works just fine when I run the Millionaires example. (with --no-ssl option) The second setup is identical to the one in your e-mail. When I run the Millionaires example (again with --no-ssl option) with the Path set to point to Python 2.6.4, I get the following error message. Traceback (most recent call last): File "generate-config-files.py", line 57, in from viff.config import generate_configs ImportError: No module named viff.config Hope this helps. Best Wishes --Sriram Sriramkrishnan Srinivasan Research Fellow Trustworthy Voting Systems Group FEPS, 27 BB02 University of Surrey s dot srinivasan at surrey dot ac dot uk + 44 (0)1483 682646 "backwards understood be only can but, forward lived be to has Life" ___ From: thomas@gmail.com [thomas@gmail.com] On Behalf Of Thomas P Jakobsen [...@cs.au.dk] Sent: 22 February 2010 08:27 To: viff-devel Cc: Srinivasan S Mr (Computing) Subject: Re: [viff-devel] Query from the University of Surrey >> I am on a Windows XP machine. I tried with the latest versions of the >> the components from the respective websites, but the apps from the >> VIFF folder threw up errors. I realize I should have taken note of the >> error messages - sorry about that. So I installed slightly older >> versions (Python 2.4 and related) and things worked fine. > > Okay. Perhaps one of the windows-savvy people on this list can help you > more or at least look at the errors. Hi I've just run the millionaires example using Windows XP Pro SP3 Python 2.6.4 Twisted 9.0.0 GMPY 1.11 Win32 OpenSSL v. 0.9.8l Light PyOpenSSL 0.10 It works fine for me when running with --no-ssl. With ssl turned on, I get a Failure: twisted.internet.error.ConnectionLost: Connection to the other side was lost in a non-clean fashion. when connections are closed, but the millinonaires program gets executed anyway. I agree that we should definitely try to make the installation process easier, e.g. as Sriram suggests, by including the various dependencies in the viff packages. Sriram: Can you make it work with the above versions? If not, which error messages do you see? Regards, Thomas ___ viff-devel mailing list (http://viff.dk/) viff-devel@viff.dk http://lists.viff.dk/listinfo.cgi/viff-devel-viff.dk