• 2 Posts
  • 59 Comments
Joined 1 year ago
cake
Cake day: October 20th, 2025

help-circle
  • Normal Google on the phone, though I am looking at alternatives in the nearish future, my phone will still be good a few more years, but there isn’t much in the way on other options beyond GrapheneOS. Not a bad start to de-google, but I wanna go full swing all in when I decide to do it, hopefully some solid Linux phones will exist by then.

    Dabbled with Ubuntu back in like 2015. Windows fucked me hard this year, wiping 3 years of data because of a failed automatic update. It used a back up 3 years prior without even asking me, and I had done several this year for various reasons. That was my last straw, wiped everything, threw mint on a USB and did a full cinnamon install. Never been happier. Had a few hiccups to start and then it’s just fucking worked. I find random shit of mine not working, but I can always find a fix online.

    Fuck Microsoft, maybe they burn eternal in the AI fire (yes, I know AI is here to stay, doesn’t mean it’s the infinite industry Microsoft thought when they dumped cash into it)


  • The authors of the paper this article is talking about graded their own work. These are the model’s builders testing it with their own code. An outside group checked it against 32 independent expansion-rate measurements and concluded it faces significant challenges…

    https://academic.oup.com/mnras/article/547/3/stag430/8507252

    I get people wanna jump to different explanations for DE and DM, but it needs to fit all the data points already gathered and organized into a theory to be a better explanation than ∆CDM model. No one is saying ∆CDM is perfect or absolutely fully correct or the final answer, just that the models fit it the best at the moment. It most certainly isn’t the correct model, their too much discrepancy in the calculated and apparent DE by a huge order or magintude. But it’s the best model at the moment.

    It’s not tired light folks, sorry 🙃



  • I work in Aerospace that works with models and GD&T (mostly metrology), if there’s any chance it misinterprets a model in any way, that’ll cause huge problems down the line during milling. Not sure what type of milling you’re doing, but the more axis you have, the more complicated all that model based programming math gets, the more room for error. Just what I’ve personally heard from our NC guys, they’re running NX and Vericut, so it could be different, but I doubt the issue would be too different in this case. AI is too good at hallucinating inputs that never happened. Maybe it’ll get that past that, personally I hope it doesn’t as that could put a lot of good guys out of work 😞




  • They’re just such fun reads, and they keep you guessing all the way through m I’m on my 2nd read through of the series, and he sets things up so well to revisit later, you almost never see it coming because of the casual way he writes, it catches you off guard all the time, never gets old. So many context clues I missed on the first go around too














  • Cooking is actually my deepest interest, I cook soooo many things, I’m obsessed with flavor. My only issue is texture, but usually I can find a dried or alternative way of prepping ingredients to avoid anything that will set off my gag reflex. So I do like 30+ different things.

    I also eat ramen all the time, I have my go to when everything else sounds revolting.