phh

Well, I have the first model of laptops using nVidia Arm chip, it's called Toshiba AC100, and rocked a nVidia Tegra 2 in 2010.

It launched with Android, but I helped put GNU/Linux on it (got my first mainline author-ship with it), and daily drove it for a year at university (back then not everyone had laptops)

It was a hell of a device, and to this day I'm still missing anything close to it. 10 inch, 900g, 10 hour battery life (which was unthinkable on PCs back then), removable battery so when traveling I just had two batteries, extremely sturdy (I used to launch it in the air, and failed some times without worrying) [1]. That being said, even back then it was seriously limited, and I would max out the RAM just by doing ssh + quasselclient. By killing quasselclient I could have enough RAM to launch firefox to a simple web page, but that's it.

[1] It wasn't made to be sturdy, it's just that there was basically no weight, the pcb was very small, and it had pretty huge bezels

Issue is not dev cost, it's existing ecosystem. To drive a smartwatch ("smart" as in running a smartphone OS giving it 2 days battery life tops)[0], you need a full already existing ecosystem, so either iOS or Google/Android. Only Google can get Google to support RISC-V in Google/Android.

[0] there might alteady be less smart watches that are already running RISC-V but we can't know since they are closed.

Yeah it's a bit weird to see the difference.

> Also, ExecuTorch does not rely on TorchScript, and instead leverages PyTorch 2 compiler and export functionality for on-device execution of PyTorch models.

I think this is the major change: AFAIU you could consider ExecuTorch as ""just a rewrite"" of PyTorch mobile using PyTorch 2 compiler.

BUT this is should still be pretty major. It was pretty hard to retarget models to be compatible with TorchScript. But most recent models will want to target PyTorch 2 compiler for the performance boost. Which means they'll automatically get ExecuTorch support.

As someone who got burned several times in trying to port models to TorchScript and run them on Androids, I'm happy they did that.

Those mini-suite definitely look pretty cool. I'm usually not a fan of sleeping in the same place of strangers (even with friends it's hard), but those mini-cabines look great.

> I just wish they were more financially competitive.

I'm doing Paris <=> Wien, and it was financially competitive to airplanes. 100€/person/way (ok cheapest seat rate, that's probably very uncomfortable... I'll see) I had 240€/person/2 way with airplanes. I just checked airplane prices again and that's still pretty much the same prices (a bit lower than train if you accept going to the FarAwayAirport, much higher if you take the Airport on a train/subway station)

It does take a lot more time because it's one train every 2 day and arrives a bit late to do train => home => workplace, so if you value being where you want to be when you want to be, you could probably say that airplanes is more competitive)

So this should be it for trying to regulate theft, right? If you can open a window without any tool other than your own body. It doesn't seem like it's possible to stop thefts without a totalitarian surveillance state (and barely event then!).

Or same can be said about media "piracy". Or ransomwares.

States have forever regulated things that are not possible to enforce purely technically.

Based on your comment, it seems you're assuming the people writing the standard don't intent to step on patents, but that's wrong.

Patents on standards is largely how standards are financed, by company who will then recoup their research spent on that via those patents.

I think that RAND patents and with the list of patents published alongside the standard are an acceptable compromise.

In France for the equivalent on ATSC 3.0, we standardized on Dolby audio (AC4), which is NOT available as RAND at all (and is stupid for many other reasons). So well, we've put ourselves in a worst situation.

Having dedicated code path for 911 is literally the problem. It makes it legally almost impossible to test in many countries (since in many countries you're not allowed to dial 911 if you don't have an emergency)

> * Isn't the US the great bastion of the free market, rugged capitalism, entrepreneurial spirit etc etc etc? I mean at least that's what I see all the people on here claiming vis-a-vis why SV is better than everyone else.

I'm fairly new to US point of view (as in, I witness it only since like 10 years ago), but I feel like this has evolved, and the actual theories of "free market" (which include not having roadblocks, promoting SMBs because of diminishing returns , the need to inform the consumers, the integration of externalities in the taxes) are getting more and more support in the US, especially in SV. I think this has been severely pushed by US' Healthcare sector, where there is no obvious blatant anti competitive behaviors, but have profound negative impact on US society. I think conglomerates killing small useful services by providing the same for free (but bad and without any further innovation) have also had impact there. I'm sure most people on HN has a service in their mind that used to be great, that got killed by Google doing the same for free, but that got enshittified and they are sad the original service is dead.

In India on TVs, (where people use 4g to watch content) Google added a data saver mode in Android TV. They just shape the trafic down (to 500kBps iirc?). I guess that was easier than trying to convince app developers to reduce their data usage on a system flag. That's stupid for a lot of reasons, but I guess in your case that would work.

I just don't understand the meaning of this, and thus I have a hard time not seeing malice in this.

I manage RMA for consumer electronics, and we don't do any shit like that.

We just have a factory test firmware that is signed by us, that can be booted from a USB storage, boots in RAM (so it doesn't need anything from local storage except first stage bootloader, and we disable secure enclave in first stage boot in that mode).

It's still a bit of a security tradeoff since that gives attacker an extra surface area to execute code on Application Processor, but still nowhere near giving your password to strangers?

Google pays gregkh to make LTS (not all LTS are made by him though), and he works mainline first, so it's already working that way

FWIW, I don't have numbers, but Horizon Europe is both a research funding scheme and an industry innovation funding scheme. https://research-and-innovation.ec.europa.eu/funding/funding... says "This is 70% of the budget earmarked for SMEs."

It has both direct research grants (ERC grants), and funds for big consortium projects, where most of the money goes to big companies, with some money also going to SMEs and researchers (It probably has intermediate funding like direct SME fund, I don't really know the full extent).

I wouldn't be surprised that UK joined Horizon Europe because of industry lobbying rather than research.

Possibly it's the author removing them at the first one star rating to keep their author score high?

> It's clear at this point that acquisition didn't work and the companies were incompatible with each other on basically every level.

I agree that the companies look largely incompatible, but it looks like it does work? An acquisition that keeps releasing products after the merge is pretty rare and is usually considered a success

> Oh, so then you are ok with those companies offering those placebos for immense prices and the insurance (aka the people) paying for that?

The price isn't immense though... We're speaking of like 3€ for a month of "treatment". I would definitely rather the state to pocket that money but for the placebo to work you need a bit of marketing, so I don't really know what's the best way.

> You could just give some natural products as placebos that are not hurting trust in medicine by legitimizing quackery.

homeopathics are completely natural products, since it's literally sugar...?

(As a french) I argue that it's better for physicians to have a way out to not give anything active to the patient who is adament they need a drug for their flu to be over faster. Also, placebo effect do work.

> I would rather force companies to release schematics, specifications and documentations after, say 3 years

On top of that I would had that this documentation must have been released to an escrow before being releasing the product (many hardware companies come and go within few years. I wouldn't mind an exception to the escrow for "big enough" companies). Also, the secure boot keys must also be released if a major security (~ local root privilege escalation without hardware access) issue hasn't been fixed for one year.

As the original comment mention, this issues doesn't exist on Linux because the FS is inode based. The file content will be removed only when the last user has been removed.

Thanks for the complementary infos. YCbCr420 (which you can't just mux on the channels directly?) dates back to HDMI 1.4 though

> - Initialising the DRAM controller and sizing memory

Hum I don't really understand that one. Based on the article I understood that the code before SBI was directly bootrom, and usually bootrom doesn't init DRAM (at least that's not how rockchip/allwinner/amlogic SoCs behave). So what's the expectation on how SoC vendors implement DDR init code? It's expected that SoC vendor push this code to bootrom (thus costing more in maskrom)?

Sadly the issue with this is that you can offset both the cost and the carbon impact in just few years. (I say sadly because of course I'd much rather people would reuse than put working hardware to trash)

Based on what you said, you didn't actually look at HDMI protocol, only the protocol exposed to your HDMI encoder chip. You could have such a protocol on Thunderbolt 3 encoder chip.

FWIW, yes HDMI is still pretty simple, but not as simple as you describe it. Even though there are 3 pairs of data it's not one pair R, one pair G, one pair B (highest-bandwidth HDMI uses 4 pairs), it's just one data bus. The data pairs aren't only used for data colors, but also conveys audio, and info frames (which will include various stuff like HDR or VRR metadata). Of course there is the matter of DRM: the content will often be encrypted (but negotiation of that encryption happen separately, over I2C)

Assuming you're not just following the water flow, you should also be able to have an under-water turbine to generate even more water than you could with a wind turbine. (technically that power would still come from the wind, but from the sail you already have)

It's carbon neutral on the order of magnitude of a century. If we were to plant a forest in the Sahara, it'll largely be net positive capture for at least 3 decades, and then we'll be able to replace coal with that wood, making a global net zero (but we'd already have reduced co2 in the atmosphere). Replacing oil with wood would probably require quite a lot of work though (as making the desalination plants to make a forest)

Yup you're right. Thanks for correcting me.

My biggest gripe against the title of the poll "Climate or economic growth?" is the underlying assumption that economic growth comes with comfort growth. Going from oil heating to heat pumps will decrease GDP while improving comfort. Making durable fridge will decrease GDP while improving comfort, etc. Not fixing the climate will increase the GDP because of all the mitigations we'll have to implement (iirc French government estimated 15% of GDP will go to climate mitigations in 2050)

Now about the thesis of OP, even if we assume a full solar + battery world, and manage to keep that +2C, electricity pricing will swing a lot (it'll probably often go negative or at least to 0) which will lead to discomforts (people not able to pay for electricity in the winter) without state protections, even if economic growth is here. And then electricity is not enough. Simply keeping the production of autos has a considerable CO2 cost.

> Sony Xperia X Compact

You can run custom ROM on it (the suupport is provided by Sony themselves), though maybe you need to actually build it yourself (which yeah not great, but you don't have to write the drivers, the build scripts and everything related yourself)

> Samsung A40.

You can run custom ROM on it (using Generic System Images)

> I buy the smallest phone I can find. Apparently no ROM developer is interested in those phones.

Well, I am interested in those phones :P. I currently daily driver Asus Zenfone 9 (compact in nowadays standards, but not really compact), and Qin 3 Ultra (actually compact, usable one-handed). And I definitely run custom ROMs on it.

Uh, that doesn't make sense to me. Could you provide some details? The point of SIM card is that there is no user-controlled software in the path between SIM card and antenna. So if the pipe from SIM to antenna was secure before custom ROM, it's still secure after it.

I don't get it. Why are they using such a complicated infrastructure for something that can just be a static file polled by devices? Let's get nifty and maybe have a small php intermediary for slow roll. Assuming 10M devices, 24h polling sounds reasonable that's 10M requests/day. A VPS should be able to handle that.

(I'm saying this as someone who managed OTA for 1M smartphones that way 10yrs ago)