illright

In the comments there's a link to consumer rights wiki that includes Raspberry Pi's response to this:

> there may be a minor financial gain in buying a part with say a 2GB part, swapping it for an 8GB part of dubious origin, and reselling it as an 8GB device. However, this brings an element of risk to the user - the board hasn't been tested by us - and is a potential support burden for us, with customers complaining of unreliable devices. We therefore remove the commercial incentive by locking devices to their original RAM size.

Frankly, that sounds pretty reasonable to me

The software still remains free, it's the registry that would become a paid service under such a model, that's what was proposed in the article. And even then, the registry would remain free for individuals, so it would merely be a price of convenience

I think it's a great idea and I hope someone inside the teams of popular registries sees this and they actually give it a go. I imagine the shitstorm of reactions that it's gonna cause from people who aren't even included in the "required to pay" group and I wish the registry maintainers strength to get through this

I think this article presents a good rebuttal to this model — foundations will never pay the long tail, and sometimes the payout may not even be all that substantial for them to bother. The model in the article has the benefit of being automatic and applies to all packages that actually get used by for-profit companies

I wonder why they only support Apple platforms, citing CoreML. Doesn't Android have a similar framework, ML Kit?

That bit of text looks pretty unremarkable to me, what makes you think that no one talks or writes like this?

Thanks for the write-up, I've been interested in connecting RF-controlled devices to Home Assistant myself lately! Your experience will definitely come in handy :)

Are you following HA's latest push towards proxy devices? A few months back, they have been improving support for RF and infrared controlled devices through proxy hardware — as far as I understand, it lets you have just one RF transmitter in the home and have several integrations use it for sending their chosen commands on their chosen frequency, a sort of transport/logic decoupling. It's new, so there's not a lot of stuff that uses these proxies, but I want to play around with them and wanted to get them on your radar as well!

I'm glad that Lumo has image editing capabilities now, if it's half as good as ChatGPT, that means I no longer have to upload photos of my room to OpenAI to try a new paint color.

However, I trust Lumo only marginally more than ChatGPT, they really don't do themselves any favours by claiming it's all open-source when only the frontend is

I didn't read the title of the article in that way, and I believe it to be a bad-faith reading.

I understood the title to mean "the people in FOSS whom we treat with the same morbid curiosity and provokation as the mentally ill people and other deviants who used to be displayed in circuses for other people's entertainment".

It's also clear to me from the text of the article that Drew does not hold contempt for neither of the two individuals he talks about, and the postscript demonstrates that he approaches this difficult topic with due diligence and respect for the people

It's a good thing I found his personal statement about this report before I found the report itself — https://drewdevault.com/blog/Addressing-harassment/

Having read both now, I'm willing to take him at his word and accept his apology for the heinous things he has said in the past. I encourage people to read his statement to get a perspective for both sides and make up their mind

That said, I still find the original comment I responded to inappropriate, I believe it derails a genuine conversation about how we treat problematic people with mental illnesses

I don't find it appropriate to make these kinds of assumptions about the author. It's okay to disagree, but it would help the conversation if you elaborated why you didn't enjoy the article

Nice article, it does a good job of conveying just how simple it is to use the OS primitives for secure-ish* delivery of secrets.

I think it falls short in meeting people where they are when they use env files. I appreciate the section at the end that says "if you must use dotenv files", but surely there's gotta be a middle ground where you can still leverage the OS and not lose out on convenience.

I think it could even look like an npm package that helps you set up the service correctly for local development, accounting for the reality that there is likely several projects on a developer's machine and some of these projects are there temporarily.

* of course, there's still all that stuff about environment variables being visible to other processes, so we still need decent runtime infrastructure for envvars like `DATABASE_URL_FILE`, a temporary file where the application can read the secret at runtime. I'd love to see that in Node instead of the `env-file`

I disagree that this submission has no value at all or that it should be auto-banned — there is a difference between thoughtless vibe-coding and spec-driven coding, guarded by tests. It seems from the next thread that this project is hopelessly outmatched by the sheer complexity of taking on something as the Docx spec, but this project has value, and for someone, it may be all that they need.

However, I do agree with your point about live debugging. In light of that, I prefer to treat this submission as a curiosity about current model capabilities, and let the authors keep improving this project if they find it worth their time.

Let's be more respectful to the differing goals of people

You can see the dimensions in the full spec table on the preorder page

https://repebble.com/watch

Dang, I really liked that there were screws on the Time 2. Does that at least mean waterproofing will be better? :p For instance, would it be a good idea to swim with the Round 2 on?

Is the Round 2 also going to have screws to potentially replace the battery, like Time 2 does?

Also, while watching the announcement video, I noticed that the pixels are quite a bit more visible on the Round 2 on the same watchface, when compared to Time 2 — on the Time 2, the image is a lot more crisp. Is that a matter of adapting the watchfaces to the bigger screen?

I don't even go past their monthly allowance :) I don't have that much audio to process

A very worthwhile mention is also Stable-TS: https://github.com/jianfch/stable-ts

Out of the box it can transcribe with Whisper or Faster-Whisper, but it can also align audio with an existing human-written transcript, providing time information without losing accuracy. This last feature was something I really needed, and my attempt at building it myself ended up much worse, so I'm glad I found this

I self-host it using Modal.com, as do some other commenters