dev_snd

I've never worn a pebble, but I also have a banglejs 2 and I really love the watch for it's hackability.

I've written my own watchface and a couple of other apps and made changes to a number of existing apps, it's really simple because you can always test your code changes live on the watch while keeping it on your wrist. There's an IDE that connects to it using Bluetooth and the code can be modified during runtime

Lthere's also a great community of hackers and tinkerers that steadily improve the watch.

It might not have the same polish as the pebble had, but it makes it up in hackability. I can only recommend getting a banglejs2 (battery life is also pretty great, I get about 10 days with regular use)

It's an excellent point. In such a scenario, if you're bound to a rigid workflow system, you will probably have a hard time recreating all the intermediary steps required to get the system back into a consistent state with the external world.

Idempotency is key and the choice of the idempotency key as well ;)

Disclaimer: I don't mean to disrespect anybody by posting the title of this work. The content is not sexist or racist.

I posted this video in response to the thread about facial recognition evasion. [0]

I originally saw this at an exposition in France a few years back, and it stuck with me ever since. It's a very clever commentary about the state of AI and facial recognition software and that's long before the big AI boom. I hope you enjoy it as much as I did.

[0] https://news.ycombinator.com/item?id=42357372

It's true that e.g. medications and alcohol can increase the risks for the child, but many people also have apnoea (often without knowing) which can also lead to sudden movements in your sleep.

The article says there's no socio-economic factors; the data says the risk is the same for all babies.

If you don't want to change your habits because of an article, that is fine, but people should be aware of the risk.

We use a co-sleeping bed next to the parents bed and the infant is just as close, but safer from harm

Yeah, it's pretty much the same story for me; after having found out how to get all the same features in KDE that I had in xfce, there was no way back. I always like to use the lightweight and less resource intensive software, but arguably not using hardware acceleration is actually very resource hungry.

Great work and great write-up! I especially love the MAINFRAME BREACH PROTOCOL that comes with it.

It would be awesome if there was a mode in which the pronunciation of the names is also same in the two languages.

For example, in french there is the name "Arnaud", which exists in German as "Arno". For a bilingual child it's much more important for the name to sound the same that to be written the same.

I immediately have to think of the potential uses for VR: lightweight setups for Varifocal lenses are one of the missing corner pieces to create a life-like VR experince. I'm excited to see where they will go with this research!

Here's the link to the original full PDF: https://cme.h-its.org/exelixis/pubs/JanThesis.pdf

Redundancy also adds a lot of complexity, as fail-over mechanisms aren't simple either. That added complexity then turns into additional possible error sources.

A friend of mine builds a component for a satellite system and the FDIR mechanisms need to be chosen very carefully, as adding more fail-safes can actually make the system overall more error prone.

There's an interesting blog post on from AWS about that topic [0]. Turns out adding more fallbacks and fail-safes is actually discouraged there.

[0] https://aws.amazon.com/builders-library/avoiding-fallback-in...

It's not even a cipher, it's just an encoding. Nobody would be surprised if chatgpt would be able to also understand e.g. utf-16.

What's interesting here is that base64 does not have the exact same byte sequence (such as a single space) between tokens. So I still think for it to understand base64 is a bit surprising.

This is the reason it only works with directx 8 or 9: they have a pretty rigid render path with a clearly separated vertex and pixel shader pipeline.

For directx 8, you would for example specify light sources as part of special draw calls to the directx API, which makes it easier to translate them to a ray traced scene graph.

Shader tools like reshare also use heuristics to detect which of the framebuffers contains a depth map to add effects like SSAO to older games.

I imagine that rtx remix can to similar things to detect where in a shader the light coordinates are stored: after all most shaders look pretty similar when it comes to implementing a Phong shader etc.

The readme states that currently it transfers around 2 gigabytes per second which is 16 gigabits per second.

The rest I leave as an exercise to the reader.

I think that the bigger problem is the misuse of Tags as the sole method of marking a commit for release for example.

Tags could be utilised for naming commits which would make it much easier to reason about where in the process of your development you are instead of referencing commits by hash. Hashes are hard to remember after all.

Once steam OS is available to the general public for all kinds of computers this is definitely the last nail in the coffin.

I use Windows solely for the reason of gaming and otherwise have switched to linux and never looked back

That is true, the network needs to be stable in any case, but the keyframes have "bursty" traffic, therefore increasing the latency compared to b-frames (and p-frames)

Ever since h264 there is a feature called intra-refresh, which allows you spread key frames over time, reducing the burstiness of the transmission (and therefore improving latency in most scenarios)

Eversince that backblaze post I've been monitoring only these five relevant SMART values and replace a disk once I have 2 of these values being in the not-normal range. I haven't had a disk failing on me ever since.

Before that, I had lost all trust in SMART, as it would report drives as "healthy" that could fail next day.

Your joke is not as far fetched as one might think! There have been multiple instances [0][1] of a Tesla (and also other car makes, for that matter) where the image recognition system could be tricked.

I wonder at which point we'll deem fooling a autonomous car a misdemeanor instead of changing the underlying AI system.

[0] https://www.technologyreview.com/2020/02/19/868188/hackers-c...

[1] https://www.technologyreview.com/2019/04/01/65915/hackers-tr...

Looks great in theory, but does not work on Android Firefox unfortunately.

I never understood why incrementing a pointer while doing another operation, such as a comparison, is considered good style in C, TBQH.

I think the go example is much more readable.

I have authored a small library for reading audio metadata in Python and I can say with certainty that ID3v2 is really one of the weirdest of all the meta data formats.

For example between version 2.3 and 2.4 the size of the bytes to store information changed from 8 bit per byte to 7 bits per byte. In Version 2.4 they wanted to have the MSB always set to zero so it doesn't trip up hardware mp3 decoders in portable mp3 player that use the MSB so sync to audio dataframes. (Otherwise you may listen to the screaming loud noise made by the device turning your metadata into audio)

Many parts of the spec are just being able to parse different buggy implementations of audio tagging programs from the early 2000s.

It really was the wild west back then. Looking at it now is really charming as the post elaborated.

Interesting question. I guess what makes us call something an AI is that it's an interactive recombination of a corpus of knowledge as opposed to a static one (such is the library). GPT-3 is far from sentient, that's for sure, but it's more interactive than any library.

People used the clouds to tell the future in the past, so there certainly was a value to reading them. Maybe GPT-3 is the next best thing to do the same, maybe it's more, time will tell.

"Ceci n'est pas une pipe", as Magritte woyld say. It's always the observer, that makes a thing into what we believe it is. Without an observer nothing has any meaning.

The AI does not have to understand words and objects in the same way as you do to have real world use cases.