Medijski sadržaj

  1. prije 19 sati

    Experimented with and threw away 600 lines of code for storing consensus-related data to LevelDB. Atomic transactions in syndtr/goleveldb are slow 😢.

  2. 4. velj

    Abstracting away the BFT consensus protocol I gave a talk on in SFBW 2019 into a general-purpose Go package.

  3. 3. velj

    The full interface:

    Prikaži ovu nit
  4. 3. velj

    Neat and fun little abstraction pattern for creating atomic database helper methods with syndtr/goleveldb.

    Prikaži ovu nit
  5. 3. velj

    Finished up an initial gossip protocol implementation on Noise with tests and documentation.

    Prikaži ovu nit
  6. 1. velj

    Writing a gossip protocol plugin over a Kademlia overlay network for Noise. This test demonstrates how the XOR distance metric over peer IDs asserts that peers are well-connected (with a lot of assumptions of course!).

  7. 1. velj
  8. 31. sij

    Now I need to find an alternative to Ristretto .

  9. 29. sij

    Released Noise v0.1: a decentralized p2p networking lib in . Tons of performance/security fixes + features (docs, examples, connection pooling, search for peers by public key, etc.). Will write up a post tmrw; gonna sleep it's 1AM.

  10. 7. pro 2019.

    "So, how do you efficiently count how many non-zero bytes there are in a byte slice?"

    Prikaži ovu nit
  11. 7. pro 2019.

    Anyone know any articles that dives deep into unsafe hacks like converting arbitrary values into a slice of bytes?

  12. 17. stu 2019.

    😴 key exchange/aead/handshake...

  13. 17. stu 2019.

    Overlay network routing stuff for S/Kademlia.

  14. 13. stu 2019.

    Smiling writing code like this :').

  15. 5. stu 2019.
  16. 27. lip 2019.
    Odgovor korisnicima

    We explicitly made the critical transaction puzzle difficult to grind/mine against to prevent such attacks from happening. I attached some photos from the paper that show some measures we made to account for this.

  17. 27. lip 2019.
    Odgovor korisnicima

    You got it. I will share it either tomorrow in the next three days max; will request the auditor to write up a bit of documentation for it.

  18. 27. lip 2019.
    Odgovor korisniku/ci

    Hold up: the proof is correct. More correct so than what is specified in the Avalanche whitepaper. I specifically defined the lemma of strong liveness as _eventual_ termination. I have emphasized this in the whitepaper below the proofs.

  19. 14. lip 2019.
    Odgovor korisnicima i sljedećem broju korisnika:

    (just to clarify, this is with 1000-way concurrency)

  20. 14. lip 2019.
    Odgovor korisnicima i sljedećem broju korisnika:

    Is there any sort of explicit flow control in Cap'n Proto? Trying it 1000-way there are improvements, though huge spikes of latency in between messages sent. I can only infer that there is some batching mechanism in Cap'n Proto, and not necessarily any flow control mechanism.

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·