Live data from Hacker News

Viewing profile — fred123

fred123

HN member
Joined
Mon, Oct 07, 2013, 9:20 AM UTC
HN karma
56
Public activity
39 items

About fred123

No profile information was provided.

Recent public activity

  1. comment
    Comment #43850756

    It’s just $20, that’s almost free compared to cost of human labor

  2. comment
    Comment #41355491

    With an LLM integrated into your IDE like Cursor or Copilot, oftentimes the LLM autocompletes the correct code faster than I can think about what must be done next. I’ve been codin…

  3. comment
    Comment #41211935

    I feel like text detection is much better covered by the various ML models discussed elsewhere in the comments. Maybe you can combine those with Live Text. I found Tesseract pretty…

  4. comment
    Comment #41207404

    Fair enough. There are some new OCR APIs in the next macOS release. I wonder if the model has been improved.

  5. comment
    Comment #41205604

    You can get them through the Vision API (Swift/Objective-C/AppleScript)

  6. comment
    Comment #41204741

    It seems to struggle with German text a lot (umlauts etc)

  7. comment
    Comment #41204730

    Azure Vison OCR is supposed to be the best commercial OCR model right now and it’s really cheap (same price as Google‘s)

  8. comment
    Comment #41204029

    IIRC Tesseract is trained on 300 DPI

  9. comment
    Comment #41203873

    Something wrong with your setup. It should be less than 30 s per page with your hardware

  10. comment
    Comment #41203657

    macOS Live Text is incredible. Mac only though

  11. comment
    Comment #40896279

    https://en.m.wikipedia.org/wiki/Extraversion_and_introversio...

  12. comment
    Comment #36917655

    > sharing state between threads is such a narow niche use case It is the norm. „Kafka scale“ problems are not the norm.

  13. comment
    Comment #36917640

    Do you think Meta (Instagram) are pushing GIL removal and Cinder for no reason? They clearly have that scale and still benefit from faster single machine performance

  14. comment
    Comment #36917623

    Why? On AWS you can rent a 24 TB, 500 core machine. Almost all problems are smaller than that so don’t need to scale to more than one machine. Building applications that run on mul…

  15. comment
    Comment #36463375

    In most parts of Europe the standard room socket has >= 3.6 kW

  16. comment
    Comment #35457246

    More like 20x actually, eg Germany has a homicide rate of 0.8 (and 0.3 if you don’t take attempted homicides into account)

  17. comment
    Comment #35450051

    I’d consider that dangerous when you compare it to the ~10x lower rate in Western Europe

  18. comment
    Comment #34975420

    SIMD, multithreading, better handling of NaN, maybe other factors.

  19. comment
    Comment #33626708

    Or just use Conda/Mamba.

  20. comment
    Comment #33251843

    You should provide a way for people to reach out to you privately eg email. If you have any public profile/CV that helps as well.

  21. comment
    Comment #32817121

    Somewhat related: I built a library to speed up matching many regexes with mostly mismatches by adding non-regex pre matchers. https://github.com/Quantco/multiregex

  22. comment
    Comment #31984622

    Where are you based?

  23. comment
    Comment #30376043

    SSDs and other storage drives have two layers (or more). The last layer is stable storage (= when you disconnect power no data is lost or corrupted). When you write to such a devic…

  24. comment
    Comment #30311316

    What’s your expertise and technology stack? 100€ is easily possible, I think it is kind of the default rate in my space (Python Data Engineering, cloud stuff, …). From my experienc…

  25. comment
    Comment #30233565

    Actually I think chapters 1–8 are more valuable than the chapters dealing with a concrete implementation because they teach you ideas and principles