

me when i am in a “build an unusuable standard” competition and my opponent is “literally any consumer electronics manufacturer”
A Reddit Refugee. Zero ragrets.
Engineer, permanent pirate, lover of all things mechanical and on wheels
moved here from lemmy.one because there are no active admins on that instance.


me when i am in a “build an unusuable standard” competition and my opponent is “literally any consumer electronics manufacturer”


Ethernet over HDMI does exist as a standard, but iirc it requires the device manufacturer on both ends of the cable to have a special implementation, and also requires a special cable that has the Ethernet data lanes included. I’m not sure any modern displays implement it anymore, it kinda died because it sucked and wasn’t that useful.


And at significantly lower transmit power too. Ubiquiti 5ac ptp rigs use like 8w, 802.11ah can make a link with under a watt. Sure it won’t be fast at all but if you’re doing a remote embedded device on a solar panel, it makes a huge difference.
That shelf sag scares me, sir. At least reinforce each layer with a slab of plywood or something.
Whatever is cheapest. When youre first starting out basically any hardware will do, it just needs to boot Linux. As you progress and find more stuff to put on the servers, you’ll discover what you’re real hardware needs are.
When I first started, it was a hand me down single core AMD Sempron machine (socket 754!) that I later upgraded to an Athlon64 and 4gb of DDR. I managed to bodge that poor thing into running a Minecraft 1.5.2 server.
Personally I would stick with the i3 machine since I am assuming it’s an office PC that can be had for cheaper than a Pi 5 (which is quite inflated in price IMO). x86 still retains better software support vs ARM and they are significantly easier to attach large cheap storage to via SATA. Power cost will be greater but I doubt an office i3 pulls more than 70w wall power at full load.


For russia. These sanctions are basically toothless, Russian corporations have already decoupled from the Western banking system and are shadow-fleeting all their oil to willing buyers in India and China without much reduction in flow. This is a theater play.


They don’t catch on fire anymore, just fill with diarrhea from an incontinent asshole.


Your comment is surprisingly spot on. One of my professors from community college that I remain in contact with is a retired AF pilot and long time Boeing manufacturing engineer from well before the M-D merger. He quit only a year or two after the merger because he saw the huge writing on the wall with how suddenly Boeing’s corporate behavior changed. He had contacts at M-D who warned him too and he got out while the getting was good.


service still up = no problem
Can’t access service = problem, better ssh in
Simple as


A. Run a batch transcode with Handbrake and make all your stored files compatible with your end players.
It sounds like the more recent things you are downloading are in a codec that is not compatible with your playback devices.
E.g, older torrents are frequently an H.264 stream in an MP4 container, which practically every device can play now. Many modern releases are being distributed in H.265 or AV1, as they have significant size and quality benefits, but many older devices don’t support them natively. so it is forcing Jellyfin to live transcode to h.264.
Find out what older titles play without any buffer or playback lag/high CPU usage and check what codec those files are in. That is what you’ll need to batch encode everything over to.
B. Sounds like you are still relying on CPU transcoding which is absolute dog. What mini pc specs do you have? If it’s an AMD or Intel CPU/APU then it should have hardware encode/decode included in it’s integrated GPU. When using hardware transcoding the CPU load is generally minimal for 1 to 2 streams. See the Jellyfin docs on hardware acceleration here.


My default goto with any stability issue is to first force a new drive self test
smartctl -l selftest /dev/nvme0
And then I would also run a complete extended memory test (memtest86) to ensure bad ram isn’t doing something dumb like corrupting the part of the kernel that handles disk IO. The number of times I’ve had unsolvable issues that traced to an unstable stick of memory is… Surprisingly high.
If the memtest passes try fsck’ing nvme0, if there are corrupted blocks yeah it’s possible the SSD is dying but the controller isn’t reporting it.


Ejectrael?
Isragone?


Another unsurprising TACO Tuesday.


India will never buy American oil because American producers cannot meet the costs of Russian oil. It costs too much to lift oil here and too much to ship that far, and Russian crude is still at a discount overall due to sanctions.


US only denouncing it because it happened to hit one of their fossil fuel allies.


Oracle just announced a huge $300bn deal with openAI to provide cloud computing services.


all it took was a slight disruption to the circus supply to cause complete societal upheaval. Crazy


sorry best we can do is hate brown and trans people
They were doomed from the second they tied their company success to those terrible awful JATCO steel belt CVT’s.