Mate Soos

@SoosMate

Unofficial SAT geek and chiptune DJ. Opinions expressed are mine and mine only

Berlin/Singapore
Vrijeme pridruživanja: lipanj 2011.

Tweetovi

Blokirali ste korisnika/cu @SoosMate

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @SoosMate

  1. Prikvačeni tweet
    14. lip 2019.

    I am very happy to finally release CrystalBall, our data gathering and machine learning system for SAT solvers: Blog: Code: Paper: It's been 2 full years of work. And it's a ton of fun :) Enjoy!

    Poništi
  2. proslijedio/la je Tweet
    1. velj

    First generation lover: aka you’re the first person in your family that’s learned how to make healthy external attachments and connections with other people

    Poništi
  3. proslijedio/la je Tweet
    27. sij

    I had to get a background check for my job, and it turns out the report is a 300+ page pdf of every single tweet I’ve ever liked with the work “fuck” in it. Enjoy your dystopian bs! *waves*

    Prikaži ovu nit
    Poništi
  4. proslijedio/la je Tweet
    28. sij

    For our 23rd annual conference we are heading to , Italy! team is taking over... Visit for all information about the event!

    Poništi
  5. 28. sij

    Just a shout-out to the developers of KDevelop, it has saved me countless hours and helped me develop CryptoMiniSat. Thanks to all of you!

    Poništi
  6. 25. sij

    For the proof geeks, a new paper by Kaufmann,, Kauers on DRAT+PAC. Takes proof conversion approach (perf interesting). I wonder what to do for other theories. DRAT(T) checker can be harder to verify/conversion to pure DRAT can make checking slow

    Poništi
  7. proslijedio/la je Tweet
    24. sij

    In one word, what would you like your legacy to be? Trump: “Victory” Bernie: “Compassion”

    Poništi
  8. 20. sij

    I have a confession to make. I have 4 mechanical keyboards. MX Blues, a Brown and a Romer-G. And while the MX Blue drives everyone around me nuts, it's the best fucking switch around. Also, Ducky One is amazing. Ah, and Logitech's G413 keycaps make my fingers hurt. Fuck that.

    Poništi
  9. 20. sij

    So near to Martin Luther King Jr's birthday, let's not forget that putting him only into the "civil rights" bucket would be a mistake, and in some sense is a misdirection of those who profit from increased inequality, human suffering, and endless wars

    Poništi
  10. 19. sij

    If you are academic creating systems that cannot fix the random seed then you have no idea how testing works in a company that needs to actually make money. Stop wasting my tax $. Go and work in industry for a single week. Sulk at all the mistakes you made. Then fix them.

    Poništi
  11. 18. sij

    In academia, if supply correct code, you have gone 1/3rd of the way. You now need to: 1) Make the code readable so others can verify it's correct 2) Make the code extensible so others can build on your work I'm *really* tired of "academia-quality" code meaning it's utter shit.

    Poništi
  12. 17. sij

    I hate that in academia you have to give a new name to each shitty idea every time. Imagine if Signal had a new name every time they added a feature. No, it's not Google now, it's Bargle, we added autocomplete! It's the same old crap, no new name needed for self-promotion, thx.

    Poništi
  13. 15. sij

    Also, making scikit-learn deterministic is not intuitive either. You have to pass a PRNG to every function that accepts it. You forgot one? Bad luck! Am I doing something wrong, trying to make stuff reproducible in ML? Or nobody cares about that in ML anyway? I feel old :S

    Prikaži ovu nit
    Poništi
  14. 15. sij

    I wanna rant a bit. It turns out that *there is no way to control the seed in SQLite's rand()* Like, how do you write tests? And also, why? I want to be able to have reproducible sampling, so "ORDER BY rand() LIMIT n" is fucking nice. I refuse to use MySQL or pandas for this sh*t

    Prikaži ovu nit
    Poništi
  15. 8. sij

    "That’s the birth of responsibility." Alan Watts

    Prikaži ovu nit
    Poništi
  16. 8. sij

    "There was no past. If the universe began in the past, when that happened it was now [..] Things are not explained by the past. They’re explained by what happens now. [..]If therefore we would stop blaming others, it would be very difficult to go about a war with a straight face"

    Prikaži ovu nit
    Poništi
  17. 7. sij

    If you were always confused about ISO 14443 vs ISO 15693 vs EPC class 1 gen 2, take a look at my 10 year old slides: They are actually quite OK! (ah, the sweet old times when I had the time&patience to read through standards!)

    Poništi
  18. 7. sij

    So, SHA-1 seems to be fully broken now... Cool work by Gaëtan Leurent and Thomas Peyrin:

    Poništi
  19. proslijedio/la je Tweet
    26. stu 2019.

    THE SEVEN SECRETS OF HIGHLY SUCCESSFUL PEOPLE 1. Private school 2. Legacy Ivy admission 3. Nepotism hire 4. Seed capital from family 5. Club memberships 6. Personal assistant, nanny, ghost writer 7. Journalists who ask, "What's your secret?" and uncritically publish the answer

    Poništi
  20. proslijedio/la je Tweet
    27. pro 2019.

    'Learning To Skateboard In A Warzone (If You're A Girl)' is currently on the short list for an Oscar!. It follows a class of girls at , a nonprofit that began as a skate school in Kabul in 2007 & grew into a multinational educational initiative:

    Poništi
  21. 4. sij

    Me: Watching "Alan Watts - Spectrum Of Love" on Youtube Youtube: Wanna watch "Jordan B. Peterson on 12 Rules for Life" right next? Tuning algorithms only to maximize viewership can lead to sad outcomes. Google can change minds (& through it, the world) why only aim for profit?

    Poništi

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:

    ·