Frank McSherry

@frankmcsherry

Co-founder and Chief Scientist, Materialize, Inc. "Science advances one funeral at a time" -Planck

Vrijeme pridruživanja: listopad 2012.

Medijski sadržaj

  1. 20. sij
    Odgovor korisnicima

    Super interested to hear more about it!

  2. 1. pro 2019.
    Odgovor korisnicima

    Eh, not exactly! I'm still hacking around with it and haven't figured out what I like best. Current status is:

  3. 30. stu 2019.
    Odgovor korisniku/ci
  4. 30. stu 2019.

    After a bit of shopping, waiting, soldering, waiting, ... we finally get our first styled keyboard!

  5. 26. stu 2019.

    Delighted to report that just dropped their NYC2019 videos, including this one featuring yours truly, speaking about the SQL incremental view maintenance engine:

  6. 20. lis 2019.

    Am I confused, or does the ACM not understand the difference between "opt in" and "opt out"?

  7. 27. ruj 2019.

    Moving day for ! Now with a corner office and a view of downtown!

  8. 18. ruj 2019.

    When you use eventually consistent infrastructure for your core business...

    Prikaži ovu nit
  9. 9. ruj 2019.

    American exceptionalism in the heart of NYC.

  10. 24. kol 2019.
    Odgovor korisnicima

    To get with the screenshot-the-juicy-bits program, we have:

  11. 14. kol 2019.
    Odgovor korisnicima i sljedećem broju korisnika:

    This is probably what you want, I think. It's work time here so latency on the interactive debugging may go up.

  12. 14. kol 2019.
    Odgovor korisnicima

    I'm not certain. The ^C there is because tests were hung, which I only know how to fix with either a `confluent stop` or occasionally a `killall -9 java`. The multiple calls to `confluent start connect` are also a thing, I'm not sure what to make of them.

  13. 11. kol 2019.

    We've got another bit of video, this time on "groups, semigroups, and monoids in differential dataflow". It's not a programming video, and may not be a great explainer video. Bit of an experiment, really.

    Prikaži ovu nit
  14. 21. srp 2019.

    More video content, this time developing arrangements for graph processing. Learn how to write graph algorithms that can re-use the same in-memory graph layout across computations!

  15. 14. srp 2019.

    I'm away for a few days, hiking with team , but while I'm gone I have six common graph problems written in differential dataflow, just for you:

  16. 13. srp 2019.

    Further video content on building up trusses in differential dataflow (a graph theory thing). We do a second, independent implementation and eventually get the same answers, even! We also perform an entirely inappropriate time-travel optimization!

    Prikaži ovu nit
  17. 10. srp 2019.

    Some new vlog content where post- I eat my own dogfood by trying (and not succeeding) to achieve the perf of recently presented research, using differential dataflow. We get as far as reproing prior work, but a few orders of magnitude remain...

    Prikaži ovu nit
  18. 4. srp 2019.
    Odgovor korisnicima

    Here's LDBC BI query 25, which might be a good case study for collapsing scopes. :D

  19. 3. srp 2019.

    More mysteriously, their GAP measurements (courtesy ) are hundreds of seconds, versus what my laptop just reported (a few seconds) and which on my work desktop is 0.6s per trial.

    Prikaži ovu nit
  20. 3. srp 2019.

    I am keen to learn why their COST measurements are higher than we reported in our 2015 paper, which still hold on my 2015 laptop as of .. just now.

    Prikaži ovu nit

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:

    ·