shadowpho

7 ms·

61.
by shadowpho 1y ago
>Honestly I would much rather handle that sort of project in an embedded environment and not something running linux. That’s because you haven’t done it. There’s a lot more code to write in embedded case and lots of pitfalls to fall into
62.
by shadowpho 1y ago
It’s easy to do something that will work for weeks. It’s hard to do something that will last years. For example, how does it handle wifi going down? Does it recover gracefully? Does it manage to save the data during downtime and reupload it
63.
by shadowpho 1y ago
SBC are easier to program for, you can trivially ssh in later and fix/update things. Meanwhile mcu are… tricky. For an example I wanted to measure co2 level. It took me about a week to write it in raspberry pi zero w2. It captures da
64.
by shadowpho 1y ago
Because then it’ll become unrepairable. It’s a tricky trade-off that cell phones/tablets-laptops have to deal with. Easy to repair vs durable vs cost/time. And Apple (for example) cares about repairability because they also need&
65.
by shadowpho 1y ago
>There's already credible evidence that we came perilously close to the use of tactical nuclear weapons in Ukraine which likely would've spiraled out of control Do you want to share any of this credible evidence?
66.
by shadowpho 1y ago
For which band? I run 160/160 on 5/6ghz and it’s nice. They are short range enough to work. For 2.4 yeah 20mhz only
67.
by shadowpho 1y ago
>Multi-slope is the gold standard and no one will sell you one. Why not?
68.
by shadowpho 1y ago
Good callout on the gross vs operating margin. I’m not sure on the pricing power. If they had more leeway on making car more expensive why not set it to that point in the first place?
69.
by shadowpho 1y ago
Two points to add: 1) Total cost of the vehicle does not matter. What does matter is the operating margin. Half a percent of the total cost of the vehicle will move them from 2% margin to 1.5% margin. (Ford has operating margin of 2% as an
70.
by shadowpho 1y ago
Same here. With openvpn my somewhat modern cpu takes out a whole core @100% at like 200 megabits/s. With WireGuard I instead max out the internet bandwidth (400 megabits/s) with like 20% cpu usage if that. I really don’t understa
71.
by shadowpho 1y ago
>but it's also not exactly a herculean task. To me it sounds like maybe 10 software engineers for a year. You gotta do a bootloader, test it out thoroughly, setup compiler to work, glue all the pieces together, write the missing pie
72.
by shadowpho 1y ago
It’s not quite that easy, and besides the hard part is the SW. arduino spent years writing SW code and still does to make it easy to run, debug issues and provide support. Also a million dev kits is unrealistic for vast majority of companie
73.
by shadowpho 1y ago
>They were constantly taking my blood, constantly running tests >My body and mind were absolutely shattered, especially after being woken up every 30min all night long for more blood draws, Would you rather they don’t take your blood
74.
by shadowpho 1y ago
True, but there might be more problems — like if you drop support their run time will be slow because they rely on this flag and they are unhappy
75.
by shadowpho 1y ago
AES-ni is enabled and Linux confirms it’s enabled and openSSL has it enabled, but I found no easy way to check if it’s actually being used (I found a link long ago but lost it :( ) I was using either AES-256-GCM or AES-256-CBC. It could als
76.
by shadowpho 1y ago
It’s surprising to me how bad the scaling is too. A 200megabits/s i what I was able to get on a fairly modern 10900k!
77.
by shadowpho 1y ago
Does it actually work well? I tried something similar and quit in disgust after realizing hdr 4k requires terrible hoops to go through
78.
by shadowpho 1y ago
If you remove the hint flag peoples build will break
79.
by shadowpho 1y ago
Nope it’s still busted and worse of all it’s single core bound!
80.
by shadowpho 1y ago
>Mediatek Pump Express Introduction Yep read it. It does even more terrible things (modulating current to change voltage), but no monitoring for voltage drops. It’s also rarely supported, and looks like they just switched to pd3 (the san
81.
by shadowpho 1y ago
>Smartphones already do this. They essentially measure the resistance of the cable by measuring the voltage and correlating it with the current drawn, and regulate the current accordingly. No! Where did you hear this? Smartphones usually
82.
by shadowpho 1y ago
https://a.co/d/2OMfGrg Here you go, I got this guy and it’s fantastic. Tells me pd capabilities, voltage/current and resistance.
83.
by shadowpho 1y ago
Been running it and it’s a beauty! Runs great on a weak-ish mini pc with pictures via nfs on a NAS
84.
by shadowpho 1y ago
> the Pi idles at 2-4W (for similar RAM spec) At the wall? I measured mine and it was higher with gpu disabled. Real world benchmarks/usage is going to be dominated by sd card speed vs m2 perf. That’s why upgrading system is so much
85.
by shadowpho 1y ago
Jeez even less pcie lanes then a n100
86.
by shadowpho 1y ago
And sadly they’ll still be right :<
87.
by shadowpho 1y ago
>because they're private and can get away with it. Also because it’s highest cost due to less populated area
88.
by shadowpho 1y ago
N100/n97/n150 are all perfect (about the same perf). 6-10w idle (similar to rasp pi) but 4-10x more perf.
89.
by shadowpho 1y ago
https://www.jeffgeerling.com/blog/2025/almost-perfect-mini-n...
90.
by shadowpho 1y ago
If you’ve got a choice get something else lol. The m2 hat blocks the cpu heatsink, and doesn’t fit into most cases. It blows my mind why was m2 not just built in already… You can get an n100 mini pc for $100 with more ram and like 5x perf

More ›