• 2 Posts
  • 39 Comments
Joined 3 months ago
cake
Cake day: July 4th, 2025

help-circle



  • 4 drinks (for women) isn’t a lot. That can be as little as two martinis or three margaritas.

    5 (for men) isn’t really a lot either. That’s less than a 6 pack of beer. Or like 3 cocktails depending on how strong they are.

    Most people I know will likely have at least one day a month where they drink this much. A birthday party, a holiday celebration, a wedding, a big game, etc. I wouldn’t think twice about a couple drinking this much on a date night. Heck, I’ve been to stuffy work functions that gave me 3 “drink tickets” (which would have put me over the limit). Throw in populations like alcoholics, college kids, service industry workers, etc, and I find it really hard to believe that’s only 5% of the population.


  • I’ve seen this stat a lot, and anecdotally there’s no way it can be accurate. It really makes me wonder about the methodology of data collection. First, it seems like the heavier of a drinker or drug user you are, the less likely you are to set aside time to participate in a long survey. Second, regardless of the assurances of confidentiality, I’m not sure people would always be honest about the extent of their drinking. In AA, one of the most important steps is admitting you have a problem…

    For anyone interested, here’s the paper that explains the survey methodology: https://www.samhsa.gov/data/sites/default/files/reports/rpt47098/Methodological Summary and Definitions/2023-nsduh-method-summary-defs.pdf

    And here’s an FAQ: https://nsduhweb.rti.org/respweb/faq.html#q6

    A few interesting notes:

    • Among people who were selected for the interview but did not complete it, the most common reasons for not responding were (1) refusal to participate by the respondent or by the parent or adult guardian of the adolescent respondent (26.1 percent) and (2) did not participate because the residents were not available, never at home, or did not respond to the web survey (18.9 percent)
    • For both data collection procedures, communications with potential respondents stressed confidentiality. Consequently, respondents’ names were not collected with the interview data. For web-based data collection, the website’s https encryption provided sufficient security for information entered from compatible devices via any Internet connection
    • The interview questions will take about an hour to complete
    • Interview respondents who completed the interview received a $30 incentive











  • This is kind of an odd take imo. FOSS is important because it doesn’t matter who the creators or maintainers are. Even if all of the people OP listed were in that room and agreed to write backdoors into their software for the government, others could just fork those projects and the community could move on without the bad actors. (I know that’s easier said than done, but it is feasible.)

    I’m not about to start cheering for Richard Stallman just because he’s not MAGA. He had some pretty bad takes about the Epstein scandal.




  • I’m not sure how well that works if the cluster is only designed to be temporary, since removing a productive node from a cluster is a bit risky

    Good callout. Just did some reading on the concept of maintaining a quorum, which I didn’t know about. Definitely need to be careful if I go with that approach, but it does sound interesting! I’m not entirely opposed to leaving the old laptop as a node and then using it for experimental stuff or maybe running just one specific standalone service on it after moving the critical stuff to the new server.