Paul Harvey

@csirac2

Technical debt collector making defendable things. Software|security|linux|electronics, generalist. Helper. try/catch-as-a-service

Vrijeme pridruživanja: srpanj 2009.

Tweetovi

Blokirali ste korisnika/cu @csirac2

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @csirac2

  1. proslijedio/la je Tweet
    Poništi
  2. prije 12 sati

    Are these really sane states for a bug tracker? - New, Approved, Committed, Done Never seen this brand of madness before, is this what being old does to you

    Poništi
  3. proslijedio/la je Tweet
    5. velj

    You now get TLA+ from pip! It includes all the tooling, so you don't even need to download the .jar or IDE first. Just `pip install tlacli` and you're good to go. More on the project here:

    Prikaži ovu nit
    Poništi
  4. proslijedio/la je Tweet
    2. velj

    One of the nice things about Andrew Gelman's blog is how benign posts routinely produce deep debates about foundations of statistics. This time, the concept of "calibration" (see comments): My attitude (new box in 2nd ed of my book):

    Prikaži ovu nit
    Poništi
  5. proslijedio/la je Tweet
    2. velj

    Giving a demo which contains rebooting the demo system and carrying on is fearless! Talk about reeling in the audience through suspense :D It’s astonishing how far has pushed Genode. It is hard to appreciate how much hard work was put into making it look so seamless!

    Prikaži ovu nit
    Poništi
  6. proslijedio/la je Tweet
    2. velj

    Pointer Compression in V8 and what it means for browser exploitation by

    Poništi
  7. 2. velj

    It turns out I'm just an idiot. Of course these sources don't discuss this; they seem to come from worlds of passively processing harvested data w/little influence over collection subject or how/what is collected. They get the samples they get, why worry about sampling effects..

    Prikaži ovu nit
    Poništi
  8. 2. velj

    I've had to go back to 2016/15 in 's backlog to "Auditing Algorithms", "Too good to be true", "Bias Variance Tradeoff", "Sample Sizes", "The Curse of Dimensionality", "Data Provenance" eps touching on these but it's no "Hierarchical modeling and inference in ecology"

    Prikaži ovu nit
    Poništi
  9. 2. velj

    This is such a painful manifestation of the "telescopes vs astronomy" comparison,

    Prikaži ovu nit
    Poništi
  10. 2. velj

    Status: trawling yrs of data science/ML/statistics podcast episodes to see who is talking about experiment/research design, data quality/context/threats-to-validity problems, how to reason about limitations on analyses thereof bc I lied to once that this was a thing

    Prikaži ovu nit
    Poništi
  11. proslijedio/la je Tweet
    1. velj

    Excited to see this come together - great team, great RB, great volunteers, great trainers, great speakers, great mentors - blessed with awesome sponsors 🙏🙏 after 7 months of work this is happening in 6 weeks 🥳👏

    Poništi
  12. 2. velj

    Drucker & Deming are fascinating. "Turn the Ship Around", turning around a submarine crew, enjoyed this a lot. High performing teams: I haven't read a good text yet, but "Peopleware: Productive Teams and Projects" was highly recommended to me by someone I trust on this topic

    Prikaži ovu nit
    Poništi
  13. 2. velj

    I'm just saying... we like to think we're good at systems thinking, but we rarely use it for the systems we use to build our systems :-) Literature on Quality Management, Safety Engineering, and High performing teams seem to have a lot of cross-over between each other.

    Prikaži ovu nit
    Poništi
  14. 2. velj

    Which brings me to software & tech debt. Every legacy system was legacy the day it was born. So.. how does your team make decisions? Most SCRUM practice isn't sufficient; do you actually know where you need to be in 3/6/12/24 months? 5, 10yrs? Do you truly know your constraints?

    Prikaži ovu nit
    Poništi
  15. 2. velj

    (The health industry has other things going on that are bigger than surgeons and mere cultural problems, but I digress)

    Prikaži ovu nit
    Poništi
  16. 2. velj

    .. and of course aviation is intrinsically more repeatable than surgery, I mean: having systems of fearless accountability/consensus-building. What little exposure I have to the medical system suggests at least some aversion to peer review; or just unsolicited input in general.

    Prikaži ovu nit
    Poništi
  17. 2. velj

    "On a wing and a prayer: surgeons learning from the aviation industry" highlights checklists, teamwork, simulations, "black box" (recordings). Both started 20th century at similar maturity levels, but av seems way ahead on accountability/repeatability imho

    Prikaži ovu nit
    Poništi
  18. 2. velj

    Contractors holding each other accountable worked. Enabling folks to feel safe(r) in confronting and resolving discrepancies, I.e consensus-based decisions worked. Tasking, planning that considers human endurance limits generally but also of specific workers also worked.

    Prikaži ovu nit
    Poništi
  19. 2. velj

    I even saw cultural problem solving; when we worked w/South Asian colleagues, they were often reluctant to disagree with (unfortunately stereotype-reinforcing) unprofessional .au staff; so they had them work in pairs, or at least a buddy system, to help them have disagreements

    Prikaži ovu nit
    Poništi
  20. 2. velj

    .. it turns out “efficiency” in this context was a bad idea: now a single assertion by one worker could be the basis for a whole series of wrong decisions. So they tried doubling up: but they’re all on the same team, so they always ended up “forcing” their answers to be the same

    Prikaži ovu nit
    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:

    ·