j4k3

joined 2 years ago
MODERATOR OF
[–] j4k3@lemmy.world 2 points 2 hours ago

I think that one is a Wayland issue. I had it at one point, but I think there may be a fix in the settings or something like that. I forget the details but I'm not at a comp at the moment. Wayland doesn't follow the same ordered operations as X and the frames come up in the wrong order or something like that.

[–] j4k3@lemmy.world 5 points 6 hours ago (3 children)

We need some kind of subletting system. Package a stupid simple Raspberry Pi image with the minimal amount of configuration required for me to self host, and set up a fediverse based DNS and certificate authority. Make it stupid simple for a user like me to self host my traffic without needing to pay for a domain and sort out all of the requirements. Like I technically have money but don't control my finances and am heavily subsidized by family due to my situation and disability. I could probably set it up if I was super motivated but networking complexity is a rabbit hole that feels intentionally convoluted and a pain in the ass every time I mess with it. Locking down a Linux image with an immutable base and well configured PAM, SELinux, and automatic updating is equally daunting when a person has no experience or professional contacts familiar with the setup to ground one's understanding.

I don't see why we need to rely on the traditional internet infrastructure. We should be able to make something like fediverse.lemmy.user.j4k3 where we collectivise the address fediverse.lemmy, use it as a DNS server, and then distribute the traffic to individual servers.

[–] j4k3@lemmy.world 2 points 7 hours ago

No, not a conspiracy but, it is not actually "emotional issues" either. It is dogma, and an emotional solution will not work directly either.

Emotional solutions may seem to help in some cases but not all. The core underlying issue is the human tribal scope. Dogma exists specifically at the level of tribalism. Tribalism is why logic and reasoning are ineffective. Any information from humans outside of the tribe is invalidated based upon membership alone. No information from outsiders is considered valid.

So if you can reach someone with an emotional connection, you are really convincing them that you are part of their tribe. However, if they accept you while the rest of the tribe has not, or they are of low social hierarchical rank in the tribe, they risk becoming an outcast other too, and preventing propagation of reasoning logic.

This dynamic of tribalism is why religious leaders are so powerful and why secular authorities engage with them. If a leader or high ranking members of a tribe endorse such a secular authority, all of the tribe must follow blindly because there is no logic in dogma, only tribalism. It can sound rather absurd, but the reinforcement mechanism is social network isolation of the tribe. Most people don't actually believe the tribal mythos, but if they leave, they lose access to their social network they were born within because this group is mutually exclusive. This social network isolated reinforcement mechanism is why there are regional faiths, and these do not compete on merit of logic or reasoning. If you were born into a region of Jewish, Christian, Buddhist, Zoroastrian, or Muslim faith tribes, you will be a part of one of these. The issue is not the validity of the tribe. Battling the mythos or ethos of the tribe is only capable of provoking the most combative and least broad-scope self aware members of the group. The only way to effectively change people is to offer them a better tribe and social network than what they have and all that they have ever known. For most humans, leaving their tribe is a death sentence in the same primitive instincts that are at play with dogma.

So, if you follow this logic, first off, welcome to the tribe, because we are doing it too! Tribalism is inescapable at this point in human evolution. Second, it should be rather obvious that any logic or argument with dogma results in the exact opposite effect of what you are trying to accomplish. If you truly want to change a dogmatic person, you must welcome them into your tribe openly. If this action seems difficult, it is because your own tribal scope is not what you believe it to be, and in a sense you were coming to a battle. Third and finally, a High Machiavellian type person with broad stroke abstractive skills can see this type of dynamic like playing with Lego bricks to make a small box; it is trivial. The skill doesn't have anything to do with sensing or emotions. It is like watching a cutaway of an engine assembly turning and intuitively seeing how all the pieces work together well enough to understand the mechanism. And this is why a person that regularly visited Epstein Island, and solicited a porn star for sex because she looked like his daughter, or an open polygamist oligarch is able to buy the heads of dogmatic tribes and get blindly accepted. These people are high Machs too. We are rare relatively speaking in terms of functional thought. High Machiavellian is not necessarily bad. It can be used for good, or like myself – to be one of the few people dumb enough to abandon their isolated socially exclusive network of dogmatism. It also enables raising awareness of those that are dangerously wielding the skill, though only at smaller scales of within a tribe, unless one makes a goal of broad scope influence but that is a challenge as well. Like, I can understand collective motivations like a machine, but I do not understand emotions of influence and popularity well. The same applies to others like Trump and Musk. If they could understand these elements, they would displace the leaders of the dogmas. Instead they must still work to appease these intermediaries.

[–] j4k3@lemmy.world 6 points 16 hours ago (1 children)

I'm not very emotional like that. Things are not traumatic in ways I process with immediacy. I still think about him from time to time over 30 years later, so I'm still processing it in a way.

With my almost dying from the crash that disabled me, I have an ongoing hypothesis that humans do not actually experience death in a way that can be processed and understood. Death happens in the limbic system which is like the human kernel space while the conscious mind exists more in a operating system like user space. We can experience the pain and all, but that isn't related to actual death.

We were inside a gym with a large basketball court running laps. I thought he had just tripped over his shoelaces or something. He was only a few feet ahead of me. Then I saw his face was dark reddish purple and his eyes had the most blank stare. He convulsed a few times and that was it. He was likely gone before he hit the ground and never processed or knew it.

In my crash, I lost 3 hours where my memory failed. I was lucid through parts of that, but it is like the blackest black in my mind. I had extensive fractured damage but of particular concern was the base of my skull and C1 around my brain stem. That whole experience processed different than other serious injuries.

I watched a 36yo woman die after crashing on an e-bike in 2022 and she appeared much the same like she did not see death coming at all.

I abstract about everything as a curiosity. At the time he died, it was an anomaly I didn't really understand. It might have made headaches a little more concerning. Now, I think he died super quickly and in a way he did not likely experience in the conscious mind. It was not bad at all for him. His family must have been devastated and traumatized, but I didn't know them.

[–] j4k3@lemmy.world 8 points 18 hours ago (3 children)

Gym teacher ran a friend of mine to death right in front of me. He had a brain aneurism. That was when I was in fourth grade.

I had a highschool chemistry teacher that failed two whole semesters of students with botched and terrible curriculum management with substitute teachers that had no instructions. The teacher had his wife diagnosed with late stage terminal cancer but was a year away from retirement. The lack of any reasonable safety nets for such a terrible situation wrecked many students lives mine included. How the whole situation was handled was so bad I hated anything academic related.

[–] j4k3@lemmy.world 2 points 18 hours ago

Very cool project

[–] j4k3@lemmy.world 4 points 1 day ago

*I killed Heime at 30 as a true test of faith because it would have been too easy to see my divine nature if I had remained at the age of perfection for centuries before giving myself as sacrifice.

[–] j4k3@lemmy.world 5 points 1 day ago

In most cases of a dead battery it is just a lack of the circuit topology needed.

The thing is that the cells inside the pack have differences in resistance and this can drain the pack over time. That said, a cell can go bad and short circuit with a low resistance that will heat up massively and cause a fire if you are not careful.

If the ideal circuit topology were used, each cell would be individually controlled by the battery management chip. A less ideal situation is for each cell to have a thermistor, aka a low resolution temperature sensor to monitor each cell. Even less ideal, the batman needs at least one thermistor to monitor charging and ensure things do not get out of hand. The worst kinds of batman rely solely on the charge circuit topology to „safely” charge the pack.

The actual circuit topology for most lithium cells requires three different modes of operations. There is 1) initial trickle charging, 2) constant current mode, and 3) constant voltage mode.

If your battery does not charge, there are likely one of two reasons why. It could be well monitored and the batman has detected an over temperature anomaly in a cell or a short circuit in part of the pack. The second potential scenario is that the batman circuit is missing the trickle charge topology and there is a threshold pack voltage that must be met before the constant current mode can activate. If the cell has sat for a long period of time without charging, the second scenario is more likely, (but not guaranteed to the extent you should ever leave such a charging cell unattended for any amount of time).

[–] j4k3@lemmy.world 6 points 1 day ago

Super lucky you. I only have my parents. So once they are gone, I guess so am I.

[–] j4k3@lemmy.world 2 points 1 day ago

I haven't looked into the issue of PCIe lanes and the GPU.

I don't think it should matter with a smaller PCIe bus, in theory, if I understand correctly (unlikely). The only time a lot of data is transferred is when the model layers are initially loaded. Like with Oobabooga when I load a model, most of the time my desktop RAM monitor widget does not even have the time to refresh and tell me how much memory was used on the CPU side. What is loaded in the GPU is around 90% static. I have a script that monitors this so that I can tune the maximum number of layers. I leave overhead room for the context to build up over time but there are no major changes happening aside from initial loading. One just sets the number of layers to offload on the GPU and loads the model. However many seconds that takes is irrelevant startup delay that only happens once when initiating the server.

So assuming the kernel modules and hardware support the more narrow bandwidth, it should work... I think. There are laptops that have options for an external FireWire GPU too, so I don't think the PCIe bus is too baked in.

5
accouterment (lemmy.world)
submitted 1 day ago* (last edited 1 day ago) by j4k3@lemmy.world to c/wotd@lemmy.world
 

accouterment (plural accouterments)

  1. (military, chiefly in the plural) A soldier's equipment, other than weapons and uniform.

  2. (chiefly in the plural) An article of clothing or equipment, in particular when used as an accessory.
    Synonyms: equipment, gear, trappings, accessory

  3. (by extension) An identifying yet superficial characteristic.

  4. (archaic) The act of accoutering; furnishing.

 

Is it super standardised, like where all 30 or 40 pin LVDS connections are the same, as in pin and voltage compatible?

Are there hardware peripherals in a microcontroller that just drive LVDS like how UART, SPI, CAN, etc. work? Or is it a messy complicated thing with display specific power supply voltages, and unique power management requirements, baud rates and such?

I can find lots of old style monitor to HDMI or VGA conversions for an old laptop screen based on display model number. But what I am looking for is a USB-C/USB-3 to LVDS converter board small enough to fit into an old apple laptop top shell and act as a second monitor with all power and functionality controlled through the USB interface. I have the fab skills. If there is a simple chip that does USB-C PD/display to LVDS, I'll toss it in KiCAD and etch it myself if I can get the chip. In my past experience with small displays for hobby microcontrollers, they were anything but standard in most cases. I have never messed with the larger stuff though. It appears like most of the old style VGA/HDMI converter boards are mostly sold with the same hardware/board with the proper LVDS connector installed.

I can take care of the backlight driver part. I'm mostly concerned with what is going on with LVDS in practice. Anyone familiar with the subject on Lemmy?

 
 

It is not in the list of combinations. Most of the LLM packages seem to lack an easy way to run a llama.cpp server with the load split between CPU and GPU. Like Ollama appears to only load a model simply with the whole thing in the GPU. The simplification pushes users into smaller models that are far less capable. If the model is split between the CPU and GPU one can run a much larger quantized model in GGUF format that runs nearly as fast as the smaller less capable model loaded into the GPU only. Then you do not need to resort to using cloud hosted or proprietary models.

The Oobabooga front end also gives a nice interface for model loading and softmax settings.

gptel is at: https://github.com/karthink/gptel or MELPA

Oobabooga is at:
https://github.com/oobabooga/text-generation-webui

With a model loaded and the --api flag set the model will be available for gptel.

In packages.el:
(package! gptel)

In config.el:

(setq
gptel-mode 'test
gptel-backend (gptel-make-openai "llama-cpp"
                            :stream t
                            :protocol "http"
                            :host "localhost:5000"
                            :models '(test)))

This splits the load to easily run an 8×7b model. Most probably already know this or have other methods. I just thought I would mention it after getting it working just now. Share if you have a better way.

 

 

Everything seems to have shifted to pipes instead of caching in almost all posts here on LW through the Alexandrite interface. Is this a permanent change?

 

I've never messed with this layer before. I usually play with Arduino or FORTH, the first with the IDE, and the latter with a simple UART connection after using the Microchip toolchain to load the FORTH interpreter.

I was looking at putting a new (to me) version of FORTH on a MSP430F149 that I have had lying around for years. I have a homemade Goodfet 42, so I can technically use it to program through JTAG. However, it would be more fun to see how far I can get into the hardware from scratch. Perhaps I might connect another microcontroller to do the I/O through the terminal within emacs. What is the simplest path to sending byte data and manipulating a couple of I/O like the additional pins of a CH340 or RS232?

I just got Doom emacs running. I would like to get as far as developing a filesystem and tree to write assembly. I also dabble in AVRs, Espressif, and Micropython on STM32s in addition to FORTH on AVRs and PICs. If anyone has any advice please share. Call me a noob in all of them though. I'm doing good to make a cat excited by a servo and LED.

Any advice or references are welcome.

 
 
 

I assume they don't make them like they used to. It was otherwise good. I overloaded it in the first place.

I got it for free from the junk store a decade ago and put new MOVs in it. Three US Quarter size MOVs, each with a thermal switch attached, an overall thermal/current breaker, and all the bypass class X and Y capacitors is something I haven't seen in most power strips, so it seem worthwhile to save for the cost of 1/10th of $4 in a replacement thermal switch. Miser miser miser... That is all.

 

cross-posted from: https://programming.dev/post/26664400

Tarlogic developed a new C-based USB Bluetooth driver that is hardware-independent and cross-platform, allowing direct access to the hardware without relying on OS-specific APIs.

Armed with this new tool, which enables raw access to Bluetooth traffic, Tarlogic discovered hidden vendor-specific commands (Opcode 0x3F) in the ESP32 Bluetooth firmware that allow low-level control over Bluetooth functions.

In total, they found 29 undocumented commands, collectively characterized as a "backdoor," that could be used for memory manipulation (read/write RAM and Flash), MAC address spoofing (device impersonation), and LMP/LLCP packet injection.

Espressif has not publicly documented these commands, so either they weren't meant to be accessible, or they were left in by mistake. The issue is now tracked under CVE-2025-27840.

view more: next ›