Klaus23

Gaussian splats can have colour components that depend on the viewing direction. As far as I know, they are implemented as spherical harmonics. The angular resolution is determined by the number of spherical harmonic components. If this is too low, all reflection changes will be slow and smooth, and any reflection will be blurred.

A .step or .stp file encodes the model as mathematical shapes, rather than approximating it with polygons, but it doesn't save the entire parametric workflow or history, only the final result. As far as I know, there is no widely adopted file format that also saves this information.

Nice. That wasn't on my radar. I was already happy when Blender made significant progress in parametric modelling with geometry nodes. Together with CAD Sketcher, it looks pretty usable for modelling. I hope the whole thing improves quickly, but I suspect they have a long way to go before Blender can be considered a proper CAD program.

Blender is great, but it still can't replace a CAD program. I tried using it for simple CAD tasks (before geometry nodes were released), but the experience was so poor that I quickly switched to FreeCAD. It was worth it, even though it took some time to learn how to use the new program.

FreeCAD is pretty buggy, confusing, and sometimes limited, but its workflow can't really be replicated with Blender. Once you have worked with a CAD program for a while, you realise that certain things that are almost impossible or annoyingly difficult in Blender can actually be pretty easy.

It would be great if the two programs could be merged. Blender could benefit from better CAD functions, and FreeCAD could benefit from everything else Blender provides.

Nice site! I have a suggestion for a prompt that I could never get to work properly. It's been a while since I tried it, and the models have probably improved enough that it should be possible now.

  A knight with a sword in hand stands with his back to us, facing down an army. He holds his shield above his head to protect himself from the rain of arrows shot by archers visible in the rear.

I was surprised at how badly the models performed. It's a fairly iconic scene, and there's more than enough training data.

Regardless of whether she's right or not, isn't she the same person who recently said that Europe needs to invest tens of billions in foundation models as soon as possible?

I have tried it and it is very usable, if a little bare-bones. Unfortunately, Bluetooth doesn't work for me, but everything else works fine.

Development seems to have slowed down a little, which is a shame. There is still no good widespread application for peer-to-peer communication.

In terms of functionality, I miss a "one to many" feature. For activism, disaster management, or even just organising a slightly larger group, it would be immensely useful to have a channel that people can subscribe to (but not write to) and that syncs via trusted contacts, without having to manually reblog everything. Forums are also a bit problematic without moderation and when anyone can invite new users.

I think remote work will continue to grow because it is so attractive to employees. It will just take time to be fully normalised and it is not available for every job.

As for mass transit, well, some countries just like 26-lane highways right through their cities. It is an ideological issue. Here in Europe I've seen a shift towards better public transport and cycling infrastructure, but that whole change will take time and won't be cheap.

Yes, there are a variety of methods that could work. It should be seriously studied to see what works best.

I hadn't thought about remote work and you're right. It's an excellent solution, at least for the jobs that don't require presence. I don't think politicians necessarily need to do much for remote work to be successful, as long as they don't block it. It is very attractive to employees, because commuting is terrible, and I have seen a lot of pressure in companies to allow remote working. Hopefully it will be more widely adopted in the future.

No, that's not what I'm saying. At the moment, cities are too attractive, and this is largely due to the availability of jobs. What I am suggesting is incentives, in whatever form, to encourage companies to move to less populated areas.

It is not necessary to obliterate the local business environment, only to limit it to the extent that the available jobs roughly match the available workers. Yes, it is true that house prices would fall to be more in line with construction costs, but that is the point, if you want cheaper housing, the housing must actually be cheap.

The businesses wouldn't go away. They would just move to less densely populated areas.

Businesses would perhaps take a small efficiency hit because of the reduced availability of labour in the region, but even this would probably be compensated for by better worker mobility, because businesses would be less crowded in the cities and there would be more labour available in the low-density areas.

Businesses would probably also be able to pay a little less as less income is tied up in housing costs. A restructuring of business and population distribution would probably not have that much impact on overall productivity, but it could shift the available money between income levels significantly. If less money goes to rich property owners, then more money stays with the average person.

It seems paradoxical to me that the only "solution" to housing shortages, which exist because the area is too attractive in large part because of the availability of jobs, is to build more houses and thus make the area more attractive to businesses because of the increased availability of workers. It looks like a battle against windmills that is bound to get out of hand. Efforts to alleviate the problem only exacerbate it.

It would be interesting to see if the shortage could be reduced by taking a different approach and making the area less attractive. For example, you could tax businesses much more if they are located in very dense areas, or even just limit the total revenue of all businesses in a certain area. Such things would have their own problems and challenges, of course, but there are few economic problems as bad as the housing crisis, and there is more than enough land to go around.

If account settings are not possible because you could fingerprint users, then client-side filtering or reordering might be a solution.

Safe-search or not, just transfer both result lists and make the client only show the one you want. The same could be done with languages, where you at least get the results for the bigger ones. Blacklists would hide your blocked crap sites. It may even be possible to implement the ranking adjustments to some extend.

Client-side filtering would put more load on the server and search sources, but I hope the cost increase is tolerable. Blacklisting and reordering could be virtually free. This could make Privacy Pass available to many more users who don't have overly complex account rules.

The downside of this is that if you are not on a larger network, the IP address will probably deanonymise you. Kagi knows you are logged in, and if you open a private browsing window to do a spicy search, they could link the searches. Fast switching between modes is undesirable.

We're going round in circles because you're just responding to points I didn't actually make. Please read my comments carefully before replying.

Your point is valid, but that's not really what we were talking about.

You said that using data centre heat is an alternative to reversible computing. This is not the case. Neglecting efficiency improvements and instead moving the data centre to a place where the heat can be used is worse than making it more efficient and using heat pumps with the saved energy at a much higher efficiency. It is even worse if the heat is dumped into the air, as istjohn said.

Also using the heat to heat homes would be even better. The two concepts aren't mutually exclusive.

istjohn is right. Using a heat pump instead of resistive heating (which is basically what a data centre is) is many times more efficient.

That doesn't mean we shouldn't use the heat a data centre provides. It just means that it is not a good idea to neglect the development of energy-saving technology because the heat produced can be used somewhere else.

I am no expert, but I like the aesthetics, especially the colours, much more than the superhero comics of the time.

The chances of the rocket reaching orbit on the first attempt are not too bad. Ariane 6 did it and only failed on the third engine relight.

Nailing the landing is a lot trickier, but if they have any early problems they would ditch it in the sea. They would have to have a problem just before landing for the rocket to take out the barge. If you want to count tipping over as a crash, that would make it more likely, but still less than SpaceX because of the soft landing and more legs.

New Glenn has the throttle range to hover and they test more than SpaceX. You will probably be disappointed.

Because it is a book download site, which is illegal in every country that has copyright, and revealing one's identity with a bank transfer would be a stupid way to go to jail.

Oh, it's 100% trustworthy until it suddenly isn't.

I have driven a number of level 2 cars on the motorway and almost all of them can do extended zero-intervention driving, but that does not make them safe. The failure rate compared to humans is still sky high.

Multiple independent FSD tests have shown that you need to take over several times an hour to avoid dangerous or illegal situations <https://electrek.co/2024/09/26/tesla-full-self-driving-third...>. The number will be lower on a motorway and you will sometimes have time to correct even if you are not looking, but the number of failures is still significant. If you take your eyes off the road, it is only a matter of time before you end up in a ditch.

I stand by my statement. The system is _never_ trustworthy enough to take your eyes off the road.

After a bit of research, I found out that they apparently did. Obviously every manufacturer would like to be able to use only proven technologies such as cameras and radar because they are cheap. One of the early Mercedes prototypes seemingly didn't have lidar <https://www.youtube.com/watch?v=DlgGTi4Gs50&t=79>.

Since then, the consensus has been that without lidar, the systems would not meet safety standards. For example, the cars need to be able to detect fairly flat objects, such as pallets that have fallen onto the road, which are very difficult to see optically, especially in difficult lighting conditions. For this reason, and because the technology has come down in price, virtually everyone except Tesla, which is developing advanced driving systems, is using lidar.

This development is nearly a decade old. It is for this reason, combined with the overwhelming amount of Musk-related nonsense, that I objected so strongly.

I doubt it. Yes, FSD is more flexible and can also drive reasonably well on city streets, but there is a reason why it is not certified for level 3 on motorways. It would most likely fail certification. With a level 3 system, I can take my eyes off the road and watch a movie. Doing that with FSD, even in the best conditions, is suicidal. Level 3 vehicles must have an extremely low failure rate. Any crash would quickly be picked up by the media.

FSD is a versatile level 2 system, but at best a prototype for level 3. If we are talking about prototypes, it has to be compared to prototypes from other manufacturers like this <https://www.youtube.com/watch?v=0uSph0asNsk> fully autonomous system from ... 11 years ago. The reason FSD is available to the average consumer is mostly a matter of philosophy, not technology.

I ignore the Chinese because it is difficult to get reliable English information. Apart from those, these are the only level 3 systems available, and level 3 is the most advanced system that private individuals can currently get their hands on. Have I missed any?

Blue Cruise is level 2+, not 3, and does not rely on lidar.

This is simply not true. Let's look at the best autonomous driving features available today, i.e. level 3:

Mercedes Drive Pilot: Uses a lidar (and a dummy unit) up front.

BMW Personal Pilot: Uses a lidar (and a dummy unit) up front

Honda SENSING Elite: Uses 5! lidars

They all use lidar, and some of the placement locations are downright hideous (Mercedes EQS). I think further development will require even more/better sensors, and manufacturers tend to agree on this point.

As Starz0r said, one of the main reasons is that the market is just very small. I think it was CSGO that had basically no protection on Linux for years, and the developers just ignored it because the small number of players didn't make much of an impact.

If it essentially calculates the area of the polygon inside the pixel box and then assigns a colour to the pixel based on the area portion, how would any spatial aliasing artifacts appear? Shouldn't it be equivalent to super-sampling with infinite sample points?