Историја ревизија

Аутор SHA1 Порука Датум
  Stan Gurtler ec1a914e9e changes to HBC stuff (not yet debugged) пре 3 година
  Stan Gurtler 1c6a298337 bug has been squashed пре 3 година
  Stan Gurtler e6735c8a97 hunting down source of HBC issue пре 3 година
  tristangurtler c36f40b878 proof batching for 2/3 cases currently in place пре 3 година
  tristangurtler da7f1d6f7f put a lock around the memoizer stuff, removed segfault пре 3 година
  tristangurtler 4b5da7e2d0 works for low numbers of users, but beyond that begins to segfault пре 3 година
  tristangurtler 33e41b4eb5 Cleaning up code to be more legible, and checking it with valgrind пре 3 година
  tristangurtler 6bf2b85137 networking has been debugged. It now works, no issues. пре 3 година
  tristangurtler 8c1f044b7c repo in state consistent with email sent to Ian at ~5:32AM 14-04-2021 пре 3 година
  tristangurtler 3cce4c56e4 server/client now compile, at least пре 3 година
  tristangurtler 0ff6d430fa oversight that this helper function wasn't already present пре 3 година
  tristangurtler 386371ec37 minor addition to server to make behave more appropriately for networking portion пре 3 година
  tristangurtler da32b7a842 minor changes that lead to ~30% reduction in proof time пре 3 година
  tristangurtler 373a1adbae I don't know what it is, but approximately 3% of the time, there is a bug that causes proofs to fail. Otherwise, I'm adding networking code now пре 3 година
  tristangurtler dda01bd883 all proofs work, all issues fixed, no optimizations yet made пре 3 година
  tristangurtler d0dc9b9bac fixed bugs identified in previous meeting (it turned out to just be indexing problems, as I had suspected) пре 3 година
  tristangurtler 0c958cdddb This compiles, but has bugs right now. The current issue I'm looking at is that, the other ways I've tested it, the fresh pseudonyms are generated correctly, but the proofs of their generation are not verifying. Specifically, when proving that a piece of the result is valid, the proof succeeds if it was generated as the product of a 0 from the permutation matrix and the original pseudonym, but fails if it was generated as the product of a 1 from the permutation matrix and the original pseudonym. I am debugging this now. пре 3 година
  tristangurtler 95332ee7d7 changed syntax to be more legible пре 3 година
  tristangurtler 813ea4bc56 Added the server consensus 'proofs' пре 3 година
  tristangurtler ae49c1f1cf Added servers checking each other when they add a new user to not allow funny business пре 3 година
  tristangurtler beb6c7aa14 Added ability for users to verify servers correctly added them пре 3 година
  tristangurtler f7faf80c5e Refactoring: put proof code (and global variables) in one shared class the other things inherit from пре 3 година
  tristangurtler 12b3a8dc96 Make valid vote proof a little more efficient пре 3 година
  tristangurtler d6215063bb Added proofs of valid votes by users, as well as more timing data to be printed when the program is run пре 3 година
  tristangurtler 6fedf2bb85 cleanup of code to be more legible, and some minor changes in how the epoch is handled to be more appropriate for when networking is introduced пре 3 година
  tristangurtler 667cf20ce1 fix for scaling issue applied; code changed to enable the actual structure proofs will take in the future; code added to enable reputation proofs (although this code is currently commented out, as the oracle function is not yet implemented, so attempting compilation with it in will cause a linker error пре 3 година
  tristangurtler 0ed45665a2 more sensible default tally пре 3 година
  tristangurtler a831ddb779 debugging (and makefile support for gdb debugging) complete пре 3 година
  tristangurtler 885c9e40cd there is now something that can actually be compiled; bugfixing remains to get it to run and be useful, though пре 3 година
  tristangurtler 33764d2df2 adding code into stubs, now missing proof stuff, ElGamal is probably not correct, no actual interaction between classes (that is, networking stuff to be added) пре 4 година