I’m a technical kinda guy, doing technical kinda stuff.
Try “lspci -vv” first to see the devices on the bus and to figure out which device is causing this.
Secondly, check all your BIOS’ “performance” settings, such as memory timings, bus speeds, and etc, and set them to default.
See how things go after that.
Off the top of my head I’d say:
And holy shit does their algorithm latch onto any minor interest in their content.
Accidentally tapped on a floor tiling video the other day, three days of tiling and handyman videos jammed into my feed and me pressing the “not interested” button on every single one.
Facebook, I am there for the rare post from my 150 or so friends and family. That’s it. Nothing else.
The reason we don’t use it anymore is because actual posts from real humans we know are buried under a torrent of shit. Sometimes their posts take days to surface leading to all sorts of chain-mail posts on how to “get your feed back”. None of which work because the whole business model is about jamming sponsored shit down your throat.
Trying to, because there is no more money to continue development.
Hopefully they can pull it off and do the same as Pebble did when they released a last firmware update for their watches that allowed third party servers to be used.
Starlink sats have enough transmit power and receive gain to use normal cellular frequencies with a normal antenna on the phone side.
You might think it’s a long way to space, but a few hundred kilometres of direct line of sight to your cellphone antenna isn’t that much more to overcome compared to say, 25 km to a cell tower on the ground.
The biggest hurdle was getting a few thousand satellites into orbit so that coverage and availability is there.
Mainly when you are building a single-purpose , “appliance” device and you have the bare minimum of RAM/storage available. You just want to get the board powered up and initialised and then jump to your application.
So you build a kernel with only the correct drivers you need, you skip initrd, you skip initscripts and (lord forbid) systemd, you just jump straight to your program, with possibly busybox available if you need debugging.
Edit: I’m talking more about building it from scratch here, not LFS. Regarding security issues, you then “only” have to deal with kernel exploits, with a limited surface as you have limited modules linked, and exploits in your application.
It’s designed and implemented for copy protection. Otherwise you can design a esp32 device that includes software you’ve written and 15 minutes later a clone device with exactly the same software will appear on <insert Chinese electronics website here>
The doors aren’t violent, they’re resolute
New sign on doors:
"The door WILL close.
Puny human flesh is not an obstacle worthy of stopping the inevitable closure of the door."
In certain countries they fall under quasi-bank regulations eg. “PayPal Australia Pty Ltd (PayPal) is a limited Authorised Deposit-Taking Institution (ADI) with authority to provide purchased payment facilities (PPFs).”
That gives some measure of protection on how they handle your funds, but holy shit I would not keep any money in a PayPal account for any longer than absolutely necessary. I use it as a convenient intermediary between my actual card and sellers, no more than that.
You can just use a soulseek client.
However I have a build of this daemon running on a Qnap storage device, which is super handy just for ad-hoc music searches, and people can also peruse my music library 24/7.
In my opinion, software engineering has about another 50 years before it matures to the point where it is a “proper” engineering profession.
“Just got to spin through a few trillion instructions to get things sorted before we go to standby! Won’t be a minute!”
You can set up fast roaming (802.11r) which is supported by most mobile clients and gives a much more organised handover between APs. Bit of config tangle but apparently works ok once you get it going.
Why the fuck isn’t there just a simple status LED that is on the same circuit as the camera?
Because cameras aren’t simple on-off devices powered by a single wire, that’s why. It’s always got power, and it’s turned “on” (send image data over the data bus) and “off” (do not send data) by software commands over the same data bus.
So the most convenient solution is then have the camera IC have an output that can drive an indicator light. And as camera ICs are basically full computers in their own right, they can be reprogrammed so that they don’t turn on that output.
End result is that you are much better off either having a physical cover over the camera lens, or having a USB camera that you can unplug.
The more we automate, the less people can do, so they don’t have jobs and no income, not able to survive…
Most solutions to this issue usually involve some variant of a universal basic income. However, that gets politically boiled down to “MOAR TAXES GOVERNMENT IS STIFLING THIS COUNTRY!1!1”, so in countries like the US that want to keep the freedom of being able to be homeless and starving, it’s not going to be possible.
Effective advertising has a clear and simple visual language, and this is what UIs should strive for.
Interfaces can be needlessly complex regardless of being flat or skeuomorphic.
But flat interfaces still require mental effort to parse. Especially when the interface is complex and/or crowded and you’re trying to pick out active UI elements amongst decorations like group boxes/panels.
Essentially, flat interfaces are currently popular because of touchscreen devices. Touchscreen devices have limited space and thus need simplistic UI elements that can be prodded by a fat finger on a small screen.
But I don’t need a flat touchscreen-friendly interface on my non-touch dual 24" monitors with acres of screen real estate. I need an interface that nicely separates usable UI elements from the rest of the application window. That means 3D hints on a 2D screen, which allows my monkey-brain with five million years of evolved 3D vision the opportunity to run my “click the button” mental command as a background process.
Is there any sort of way to get the best of both worlds? to have the PC be able to go from power button to jellyfin server started and still have some measure of security?
Windows with auto login? Not really. That is, anyone with a mouse + keyboard locally can get in there.
You can set up jellyfin to run as a windows service and then it should auto start and run as a particular user without you having to log in. Have a look in the “advanced” section in the jellyfin docs.
Train your LLM better.
You didn’t go to the library in the '80s and watch a DVD of a documentary to get the information you wanted.
So this is the concern I have with letting LLMs do all the heavy lifting. You’ve put in a nice summary of how we should be using LLMs and then here’s a glaring anachronism. So now that I’ve spotted that, should I take any credence in whatever else you’ve said?
I was thinking that I would have to switch to bsd.
Finally the year of Hurd on the desktop?
Well you see, engagement is down, and the whole “sponsored content” thing is in a death spiral due to AI slop. So Meta has decided to cut out the middleman and generate their own AI slop, because surely their version of personalised AI slop will solve the whole engagement problem and keep line always going up, because if it’s one thing users love, it’s an endless torrent of AI slop.