HNHacker News
TopNewBestAskShowJobs

hardwaresofton

12,857 karma · joined January 16, 2013

writing at vadosware.io
submissionscomments
hardwaresofton··on Self-hosting your own media considered harmful according to YouTube
Yes, sometimes -- I think TikTok's content/goals are a bit different than YouTube.

TikTok is a direct competitor to YouTube Shorts, but not YouTube as a whole -- YouTube also competes with Netflix and surprisingly paid course sites (did you know YouTube has courses?)

I don't think it's as easy as thinking TikTok will unseat YouTube. Also, I personally think TikTok's... approach is a bit hard to sustain. Just like Facebook's approach of initially showing you a feed of friends activities, but morphed into something else over time (some of that is not FB's fault, humans have certain behaviors that can be toxic all on their own).

hardwaresofton··on Self-hosting your own media considered harmful according to YouTube
> In that case, I was happy to see my appeal granted within an hour of the strike being placed on the channel. (Nevermind the fact the video had been live for over two years at that point, with nary a problem!)

Looks like some L-(5|6|whateverthefuck) just got the task to go through YT's backlog and cut down on the mention/promotion of alternative video platforms/self-hosted video serving software.

Quick appeal grant of course, because it was more about sending a message and making people who want to talk about that kind of software think twice before the next video.

> But until that time, YouTube's AdSense revenue and vast reach is a kind of 'golden handcuff.' > > The handcuff has been a bit tarnished of late, however, with Google recently adding AI summaries to videos—which seems to indicate maybe Gemini is slurping up my content and using it in their AI models?

Balanced take towards the end (after the above quote), but yep, the writing is on the wall.

I really wonder where the internet goes in this age. The contract between third party content hosters and creators is getting squeezed, and the whole "you're the product" thing is being laid bare more and more.

Is it a given that at some point creators will stop posting their contents to platforms like YouTube? Is it even possible at this point given that YouTube garners so many eyeballs and is just so easy? Does a challenger somehow unseat YouTube because programming and underlying libraries (ffmpeg et al) becomes so easy to use that spinning up a YouTube competitor goes down to basically zero?

Seems like there needs to be a new paradigm for anyone to have a choice other than youtube. Maybe AI will enable this -- maybe "does jeff have any new videos" -> a video gets played on a screen in your house and it's NOT hosted on YouTube, but no one knows and no one cares?

hardwaresofton··on Test Postgres in Python Like SQLite
Ding ding ding. Testcontainers is a fanatstic way to write the most important tests (arguably) for your app. Don't test E2E with a mock, just use a real database.

If that feels hard to you (to set up your app pointing to another DB, run a single E2E-testable part of your app with it's own DB connection, etc), fix that.

hardwaresofton··on Test Postgres in Python Like SQLite
TIL this exists, thank you! It's seems obvious in retrospect.

Would love to see a benchmark between this, pglite and testcontainers.

hardwaresofton··on AGI is not multimodal
Not to be overly snarky, but I'd argue that your comment here is a failure of creativity.

If the AGI mechanism can learn from a real world, it can learn from a simulated one (that it can similarly operate within and act upon) -- and in fact that can cut down the time it would take to train the AGI from years/decades (humans) by many orders of magnitude.

We already see things like this in robotics environments, it's a matter of fidelity/simulation quality. Even without perfect quality, if the mechanism of learning is correct, you'd get an intelligence with incomplete ideas/intelligence, not a completely different thing.

hardwaresofton··on AGI is not multimodal
This is a really interesting paper.

The discussion about the importance of decoders strikes me as a parallel to the human eyes, ears and other sensory organs. We actually dont have a good grasp of what our eyes see, they just see (produce data and relay it) and children figure out what is what.

I guess AGI will be achieved when we can sit a program in a simulated world with completely fabricated input and get a general intelligent program out. Maybe we’re in that simulation right now.

hardwaresofton··on Swift at Apple: Migrating the Password Monitoring Service from Java
It absolutely was. All it takes is a quick look around other language ecosystems and JS itself.

Rust is an excellent language for embedding in other languages And underpinning developer tools.

That said, someday the new typescript binary will compile to WebAssembly, and it won’t matter much anyway.

hardwaresofton··on Cap: Lightweight, modern open-source CAPTCHA alternative using proof-of-work
No need for the SIM, just being logged in to something will probably be enough to stop most crawlers.

Then, if someone is logged in, you can throw TOS their way, and make it a legal problem.

hardwaresofton··on How tech workers feel about work
I think this is a(n accurate/expected) reflection of Lenny's readers/listeners, for better or worse
hardwaresofton··on A thought on JavaScript "proof of work" anti-scraper systems
anonymity and the open web are different things, and neither of them were promised/guaranteed to anyone on the internet.

For people that value anonymity, they'll create their own spaces. People that value openness will continue to be open.

What we're about to find out is what happens when the tide goes out and people show you what they really believe/want -- anything other than that is a form of social control, whether via browbeating or other means.

hardwaresofton··on A thought on JavaScript "proof of work" anti-scraper systems
You were right on the second one! Facebook wasn't even a thought in my mind per say (they're not unique in that every social network wants to build a walled garden).

My focus was more on the areas outside the large walled gardens -- they might be come a bunch of smaller... fenced backyards, to put it nicely.

hardwaresofton··on A thought on JavaScript "proof of work" anti-scraper systems
> Yeah. People over-estimate the flashy threats from AI, but to me the more significant threat is killing the open exchange of knowledge and more generally the open, trusting society by flooding it with agents which are happy to press "defect" on the prisoner's dilemma.

I don't think societies are open/trusting by default -- it takes work and a lot of anti-intuitive thinking, sustained over long periods of time.

> "High trust society". Something that took the West a very long time to construct through social practices, was hugely beneficial for economic growth, but is vulnerable to defectors. Think of it like a rainforest: a resource which can be burned down to increase quarterly profit.

I think the trust is downstream of the safety (and importantly "economic safety", if we can call it that). Everyone trusts more when they're not feeling threatened. People "defect" from cultures that don't work for them -- people leave the culture they like and go to another one usually because of some manifestation of danger.

hardwaresofton··on A thought on JavaScript "proof of work" anti-scraper systems
Fantastic work by Xe here -- not the first but this seems like the most traction I've seen on a PoW anti-scraper project (with an MIT license to boot!).

PoW anti-scraper tools are a good first step, but why don't we just jump straight to the endgame? We're drawing closer to a point where information's value is actually fully realized -- people will stop sharing knowledge for free. It doesn't have to be that way, but it does in a world where people are pressed for economic means, knowledge becomes an obvious thing to convert to capital and attempt to extract rent on.

The simple way this happens is just a login wall -- for every website. It doesn't have to be a paid login wall of course (at first), but it's a super simple way to both legally and practically protect from scrapers.

I think high quality knowledge, source code (which is basically executable knowledge), being open in general is a miracle/luxury of functioning, economically balanced societies where people feel driven (for many possible reasons) to give back, or have time to think of more than surviving.

Don't get me wrong -- the doomer angle is almost always wrong -- every year humanity is almost always better off than we were the previous year on many important metrics, but it's getting harder to see a world where we cartwheel through another technological transformation that this time could possibly impact large percentages of the working population.

hardwaresofton··on Advice for a Sick Hacker with Nothing
Hey sounds like youre having an incredibly difficult time, I hope things turn sooner rather than later.

> language, I know the linux shell very well, I use git, I consider myself a Python expert

It sounds like applying to every python job you can find is probably your best bet.

No job too low, no company too boring. Hope you find something

hardwaresofton··on Meta's Pattern of Failed Big Bets: From Metaverse Meltdown to AI Brain Drain
You can’t have it both ways. You can’t claim it’s only luck and acknowledge this truth:

> with his only real successes coming from spending his company’s fortune to acquire other people’s successful ventures—Instagram, WhatsApp, Oculus.

There is clearly skill involved on top of the initial luck (and likely before it).

Meta has clearly made mistakes but it didn’t get to where it was by accident.

The decision to open source models was something others weren’t planning to do and absolutely changed the atmosphere. It’s either visionary or an incredible competitive move, or both.

Also, seems like this article was written by AI, the repetition gave it away.

AI is actually getting quite good

hardwaresofton··on Look Ma, No Bubbles: Designing a Low-Latency Megakernel for Llama-1B
You’re right — I thought it was one of the papers with better UX that has been coming through recently — it’s just a blog post but wow I wish all the papers read like this.
hardwaresofton··on Look Ma, No Bubbles: Designing a Low-Latency Megakernel for Llama-1B
Meta note but this paper is wonderfully written and incredibly approachable — excellent work by the authors.
hardwaresofton··on Official MCP Metaregistry
MongoDB as the first DB paradigm you write an adapter for in 2025? Really?

Someone is getting paid an AI salary and picking Mongo as a backend in a public project that’s supposed to be run by many people and orgs? At least there’s an interface in place[0] so it can be changed at some point.

Is this isn’t payola I’m incredibly disappointed. Payola I would at least understand.

[0]: https://github.com/modelcontextprotocol/registry/blob/main/i...

hardwaresofton··on Coding without a laptop: Two weeks with AR glasses and Linux on Android
Has anyone tried the Viture Pro? There are myopia adjusters built in.
hardwaresofton··on A kernel developer plays with Home Assistant
Oh yeah I’m aware, thanks for pointing it out. I also have zigbee2mqtt available for me for API-first interaction.

Setting up HA at least at first seems like a good idea so I can figure out the best parts and worst parts and have a baseline

hardwaresofton··on A kernel developer plays with Home Assistant
Yeah these are all good points, thanks.

I really need to figure out how deep I want to go -- HomeAssistant is clearly the best off the shelf option. Maybe I'll set up HA first and then see if it really is worth trying to build something better.

hardwaresofton··on A kernel developer plays with Home Assistant
This is a good point I haven't considered until now. Inability for the ad-supported model to ever sneak in.

Now I'm really wondering why over wifi didn't take off.

hardwaresofton··on A kernel developer plays with Home Assistant
Yup, you're right -- looks like zigbee2mqtt is a huge unlock, hard to build without it since it supports so many things!

Not excited about having to essentially now also bring along a MQTT broker but... It's probably pretty painless to run most brokers and it's likely a single-machine-is-fine affair.

hardwaresofton··on A kernel developer plays with Home Assistant
Thanks for sharing your experience here -- agreed that it's better to avoid the chatter and also rotating your WiFi passwords is indeed an issue to consider, that would be quite a pain.
hardwaresofton··on A kernel developer plays with Home Assistant
Thanks! This is what I eventually got to. That said, I'm leaning towards putting the USB stick in a Pi or something like that, which is attached to and powered by the router!

Just want to have the one device and I think that's maybe the simplest way to get it without trying to run stuff on the router.

Basically, it is a hub, but it's more of an attachment to the router than anything else.

hardwaresofton··on A kernel developer plays with Home Assistant
> If you have a $70 nicely designed, documented IoT sensor but the DIY home automation people think they can put ESPHome on a $10 Amazon device and accomplish nearly the same thing, which one do are they going to buy?

I agree with you that targeting devs/people with an under-developed sense of buy/build doesn't work!

I'm hoping that the devs with money/slightly more accomplished who don't have time to mess with stuff would pick it up.

A bit of a jump but I think this was one of the big strategic mistakes of products like FirefoxOS -- aiming for the dev with money who can afford a $600 I-support-foss-and-mozilla signaling device that also happens to be a phone they can hack on may have worked better than targeting $60 feature phones.

> If you go through the forums you can already find some semi-premium devices that are a little better constructed and might have better feature sets. They’re always followed by comments from people recommending a cheaper option.

I think that's OK! It's similar to HN dropbox effect. IMO you actually want those people to find the cheaper option, those are bad customers for a premium brand.

The product has to actually be noticably better/more consistent/etc than the cheaper option though. And IMO there is a large subset of nerds that don't actually want to write 5 files and flash an OS to do smart home stuff -- they want to play at the application layer.

A good counter example might be the NAS industry.

This is all theoretical of course, so... We're just armchairing at this point.

hardwaresofton··on A Linux kernel developer plays with Home Assistant: case studies
> Aka buy lightbulbs and switches from ikea and you can right out start using them, I believe you only need a hub to create groups of devices which then can get controled with one switch. You then could unplug the hub and still use them, only needing a hub for ethernet bridging and automations.

Yeah thanks for pointing this out -- just need a single Zigbee coordinator (if my light research has been correct so far) and I'm ready to go.

I think IKEA bulbs will also be in my future.

hardwaresofton··on Rustls Server-Side Performance
> Too bad there are those who think they should use Rust to write GUIs and end-user applications, which is where Rust ergonomics breaks down fast.

Disagree here, happy that those people are experimenting, and Rust is being used all the way up and down the stack. I may not prefer using Rust for web pages, but I'm super glad that some people want to -- projects like Dioxus and Tauri are really fantastic to use, and reflect well on the ecosystem as a whole.

I think Rust gained from this enthusiasm, because it's one of the languages that I think goes almost everywhere, if you pay the high upfront cost of learning. There really aren't that many domains where you absolutely couldn't write Rust. The fact that Swift is also chasing this quality suggests it's valuable.

hardwaresofton··on A Linux kernel developer plays with Home Assistant: case studies
Welp that's depressing.

Thanks again for laying this out -- I've been seeing zigbee2mqtt everywhere and this explains why someone would add mqtt to the mix. Sounds like this is another thing that needs to be run/managed on the software side to be robust.

This is an insane goal (and who knows when I'll actually get to work on this project), but what I want to build is an all in one something that "just works". So roughly:

1. Pick a good enough physical comms stack to hit most things

2. Write software to fill in the rest

It's going to be difficult but it feels like the setup for all these tools is just hard, when it doesn't have to be if you could pin down the hardware/install instructions, then write a really decent software layer to pull it all together without making people go homelab.

That said, that's probably what home assistant devs thought before they reached the current level of complexity, I'm probably preparing to attack a windmill here.

I think my secret sauce here will be WebAssembly -- if I can nail down the hardware below, build/convert a ton of adapters via WebAssembly, and then build a compelling/easy to add/install/manage/configure UI on top of that, I might have myself something worth posting to HN someday.

hardwaresofton··on A Linux kernel developer plays with Home Assistant: case studies
Yeah I think so -- I like to think I can control my router at least, so I don't have to worry about it. That said, probably not protected from the botnet case.

Also, unfortunately up until now I've been saying the wrong thing -- I mean Matter over Thread versus Matter over Wifi. Matter over Wifi seemed like a winner to me because I could just use it.

It looks like going forward I'll be plugging a small SBC into my Router's USB (+ ethernet) and connecting a Zigbee + Thread dongle. That should cover me for most communication options, then from there it's "just" a software problem :)

← PreviousPage 7 of 34Next →