Morphit

I expect (hope) it’s a small factor, but I wonder where pedestrian fatalities fit in. Several of the worst models seem to be large SUVs or sports cars - alongside these Teslas and some rather cheaper compact cars.

Go to the actual report. There is one table for the top fatalities by vehicle model and another for the top average fatalities by manufacturer.

As a note, it looks like the data they used is publicly available from the NHTSA. They mention that “models not in production as of the 2024 model year, and low-volume models were removed from further analysis.” I wonder where the Hummer and Rivian show up there since they are not mentioned in the report whatsoever.

Yeah the Rolling Stone article is written really weirdly. I don’t think it’s technically wrong anywhere but it reads really misleadingly when you compare it to the actual report.

Like it leads with “the group identified the Tesla Model S and Tesla Model Y as two of the most dangerous cars” - meaning they are in the list - at sixth and twenty first places respectively. The mix is really weird though. As you mention the top of the list is cars like the Chevy Corvette and Porsche 911, but also things like the Mitsubishi Mirage and a load of Kia models. So it seems like there’s a lot to interpret there.

Certainly it’s somewhat damning that despite the driver assistant technology, these models are not particularly safer. But I think other manufactures have a wide range of vehicles at different price points that also vary in safety, which brings their averages below Tesla’s in the final rankings.

For thin clients?

It’s Cannonical. They prefer implementing everything themselves fast, rather than developing a more sustainable project with the rest of the community over a longer timescale. When they do that, there will be very little buy-in from the wider community.

Others could technically implement another snap store for their own distro, but they’d have to build a lot of the backend that Cannonical didn’t release. It’s easier to use Flatpak or AppImage or whatever rather than hitch themselves onto Cannonicals’s homegrown solution that might get abandoned down the line like Mir or Ubuntu Touch.

It’s Cannonical. They prefer implementing everything themselves fast, rather than developing a more sustainable project with the rest of the community over a longer timescale. It makes sense that when they do that, there will be very little buy-in from the wider community. Much like Unity and Mir.

As you say - why would others put time into the less supported system? Better alternatives exist. If Canonical want their own software ecosystem, they’ll have to maintain it themselves. Which, based on Mir and Ubuntu Touch, they don’t have a good track record of.

They’re not converting it back into electricity, this is for industrial process heat. They have 100 units of electrical energy and 98 units go into whatever the industry needs to heat.

Lots of industries use ovens, kilns or furnaces. Mostly fueled by gas at the moment. Using electricity would be very expensive unless they can timeshift usage and get low spot prices. Since they need heat anyway, thermal storage is pretty cheap and efficient.

It’s heat though. They’re turning electricity into heat then moving that heat to where it’s needed, when it’s needed. Making heat from electricity is nearly 100% efficient, and pumping losses for moving fluids are going to be tiny compared to the the amount of heat they can move. They quote the heat loss in storage seperately as 1% per day. It seems reasonable.

I don’t know why you’re getting downvoted. It makes perfect sense that Cannonical made it’s own proprietary package ecosystem and while technically anyone can build their own snap store, ain’t nobody got time for that.

curl shit | sudo bash is just so convenient.

Plus it did land on the barge. Most of the debris should be there, though the remaining fuel would have mainly gone overboard. Probably the flight termination explosives also.

I think what you want is in Firefox nightly right now: https://blog.nightly.mozilla.org/2024/08/07/firefox-sidebar-and-vertical-tabs-try-them-out-in-nightly-firefox-labs-131/

That expands and compacts based on the sidebar state and can be flipped to the right side of the window in the ‘customise sidebar’ settings.

1xx: hold on
2xx: here you go
3xx: go away
4xx: you fucked up
5xx: I fucked up
6xx: Google fucked up

A balance has to be struck. The alternative isn’t not getting anything better, it’s being sure the benefits are worth the costs. The comment was “Why is [adding another decoder] a negative?” There is a cost to it, and while most people don’t think about this stuff, someone does.

The floppy code was destined to be removed from Linux because no one wanted to maintain it and it had such a small user base. Fortunately I think some people stepped up to look after it but that could have made preserving old software significantly harder.

If image formats get abandoned, browsers are going to face hard decisions as to whether to drop support. There has to be some push-back to over-proliferation of formats or we could be in a worse position than now, where there are only two or three viable browser alternatives that can keep up with the churn of web technologies.

I mean, the comic is even in the OP. The whole point is that AVIF is already out there, like it or not. I’m not happy about Google setting the standards but that has to be supported. Does JPEGXL cross the line where it’s really worth adding in addition to AVIF? It’s easy to yes when you’re not the one supporting it.

https://xkcd.com/927/

Adding more decoders means more overheads in code size, projects dependencies, maintanance, developer bandwidth and higher potential for security vulnerabilities.

Will you be able to handle all these panels as it becomes economically reasonable for people to replace them?

You wouldn’t print a mouse.