a-dub

18 ms·

331.
by a-dub 2y ago
"I think part 2 will be about the wiring and the custom ECU I designed to keep the actuators happy, and to implement things like the cruise control buttons, as well as reading out the speed, blinkers, … After that, I’ll do part 3 expla
332.
by a-dub 2y ago
opensnitch is pretty good on a linux desktop, especially with its cool ebpf kernel interface. mostly good for monitoring for malware though. doesn't replace good inbound firewall rules.
333.
by a-dub 2y ago
a good computer system would allow the application programmer to write their code in a synchronous and single threaded manner and all performance optimizations with respect to multiprocessing (and associated resource sharing) would be handl
334.
by a-dub 2y ago
google's playbook from the 2000s was to run a search engine based on a novel ranking algorithm that actually worked with a ui designed by engineers who like things fast, simple and minimal. the rest of the process described herein has
335.
by a-dub 2y ago
interesting, didn't know that the networking stack had ring buffer infrastructure as well. (i don't think this af_xdp stuff existed when i was in this world) the fib trie is the core of the ip stack - i was using it as proxy for
336.
by a-dub 2y ago
it is my understanding that io_uring is the generalized open source implementation of this, although i do not think it bypasses the kernel fib trie like openonload does...
337.
by a-dub 2y ago
i've actually made that exact mistake in the past. luckily it was for something that wasn't important. i reiterate my suggestion.
338.
by a-dub 2y ago
i'd consider using a signal generator (square waves), led and oscilloscope to characterize the response time of the sensor apparatus itself.
339.
by a-dub 2y ago
i wonder what the story was behind the 68000. did they design a cpu core for the future and then wrap it with a compatibility layer to meet the market or was it simply all they had to ship given time and/or engineering constraints.
340.
by a-dub 2y ago
wrote my first programs, accessed my first bbs and setup my first networks (midi and null modem) on the atari st. the best part was that an open box 520st that plugged into the tv with a single sided 320k floppy disk drive was just barely c
341.
by a-dub 2y ago
that's what i remember. i remember reading an academic paper about a cool hack where someone was getting the shaders in gpus to do massively parallel general purpose vector ops. it was this massive orders of magnitude scaling that enab
342.
by a-dub 2y ago
how does it compare with bastion?
343.
by a-dub 2y ago
after all of the headaches in the late 90s/early 2000s with truncating password hash functions, i'm just a little surprised that this sort of thing would still be an issue. i understand performance concerns and design trade offs,
344.
by a-dub 2y ago
why would someone in 2024 reach for bcrypt for building a secure hash key?
345.
by a-dub 2y ago
the term "open source" means that all of the materials that were used to create a distribution are available to inspect and modify. anything else is closed source. it's as simple as that.
346.
by a-dub 2y ago
is risc-v anywhere near the same efficiency ballpark?
347.
by a-dub 2y ago
a better analogy for frequency domain interference would be something like the spinning flashing lights on a fire engine or utility truck occasionally shining colored light on your detector.
348.
by a-dub 2y ago
i always thought it was "do the collection of things to mitigate the problems that have been learned over the years that cause problems with long running processes" changing the cwd so it doesn't cause problems with mounting&
349.
by a-dub 2y ago
thanks. you seem to think that a spiking multimodal variant of transformers on neuromorphic hardware would demarcate a goal of some sort, which one? for as far as i can see, the achievement would just be a spiking multimodel variant of tra
350.
by a-dub 2y ago
the goalposts for what?
351.
by a-dub 2y ago
neuromorphic hardware is just hardware that has biologically inspired designs. spiking neural networks are artificial neural networks that actually simulate the dynamics of spiking neurons. rather than sums, ramps and squashing, they simula
352.
by a-dub 2y ago
life is nothing but shaped noise
353.
by a-dub 2y ago
yeah but still not great for metadata operations, no? i remember it was really not great for large sets of search paths because it defeated the kernel's built-in metadata caches with excessive context switching?
354.
by a-dub 2y ago
does modern fuse still context switch too much or does it now use io_uring or similar?
355.
by a-dub 2y ago
education and socioeconomic sorting are at cross-purposes with each other.
356.
by a-dub 2y ago
it is, however, cross platform for a handful of embedded targets of which android is one of.
357.
by a-dub 2y ago
to be clear, this is only really useful for applications that present their ui through opengl and do not interact with the rest of the android system very much. the ndk is meant for writing little bits of c to speed things up in a classic j
358.
by a-dub 2y ago
it seemed to me that the bottleneck mostly revolved around the layers that were in system ram and that a lack of vram was really the gating factor in terms of reasonable inference performance. (although i would imagine that there's pro
359.
by a-dub 2y ago
idk, i have a pretty powerful laptop with 16GB VRAM and a 3080ti. when i've played with quantized llama2 and llama3 (with llamacpp), it was kinda underwhelming. inference was slow, the laptop would heat up and the results weren'
360.
by a-dub 2y ago
oh i have no problem at all installing stuff in my own environments, i'm all about having cool new tooling -- it just starts to get a little rude to ask others to do so in order to use something you're distributing (and therefore

More ›