They are working tirelessly to form a cartel :)

30% of posts on X are already done by bots. Grok will make it even worse :)

I'm like you. I have 2 printers and never printed 3D Benchy as well.

It works like this in more European countries: Ireland, Belgium, Netherlands, Finland, Slovakia, Italy.
This rounding to 5 cents is done only for cash payments.

You can get rid of some limitations of Fourier transform by moving to Laplace transform.

You need a lot of neurons to approximate some more complicated functions. E.g. try to approximate sine function over its full domain with a 3 layer neural network. As running out of available neurons, you will very quickly see that Fourier transform is better in this case. NNs are universal appropriators but only over a limited subset of R^n.

The nice thing of Fourier transform is that the computed parameters have a very well defined meaning. Computing them will give you clear insights into the behavior of the transformed functions. The point is that GP claim that some other function approximators are "arguably" better than NNs is correct. It depends on your scoring system.

Do you even drive it or is only stored in a garage at a low temperature and about 30% charge?

Because itsme1234 discussed power delivery over usb-c and that it is negotiated over D+/D- lines. I only pointed out that it is negotiated over CC line. But maybe itsme1234 meant maximum current request over usb device descriptor. Or some other options I mentioned which use D+/D- lines. But yes, I was completely focused only on USB PD specification. I did not even mention standard data communication anywhere in this thread. Good you pointed out that D+/D- are required by standard.

I guess it depends whether you consider USB Battery Charging (USB BC) specification (which can run over D+/D- lines) to be part of USB PD specification (which uses CC lines). I downloaded USB PD specification and it did only reference USB BC. It did not contain it. Strangely I could not find USB BC specification on usb.org. Only some agreements and testing protocols. Maybe USB BC is not that open, maybe it derives something from the proprietary standards (like Qualcomm QuickCharge etc. there is a bunch of them).

The physical layer description of USB PD says CC wire is used (i.e. USB-C is needed). Based on a web search some early devices supported a data wire for this but "these devices were very rare". I have not seen such a device personally. Every USB PD capable device I have ever seen always used USB-C.

Every usb-c cable should be able to handle at least 5V and 3A. USB power delivery (PD) is negotiated over CC data lines (configuration channel). That is the reason USB PD works only on cables with usb-c connectors. There are some charging standards (like e.g. Qualcomm Quick Charge) which use D+ and D- data lines of USB 2.0. But this is unrelated to USB PD.

They should also cut the bundling of Copilot from their Microsoft 365 subscriptions. Especially since it looks like an illegal thing to do for a monopoly.

How do you define momentum of a moving shadow?

There is a big difference between Haskell and Lisp. Haskell types are checked at compile time. That by itself is a huge difference. It has other goodies like algebraic types with pattern matching, type classes, monads ... and really nice libraries. And one does not need to deal with parentheses so much :)

They should be taught Haskell. Fundamentals is important.

The weight of emission component is likely higher than the weight of mass component. This way the higher mass of EVs will not dominate. Anyway I do not know the details but weighted average is an easy and simple way to achieve it. The actual algorithm may be more complicated though.

This will shift to "different price for everybody" over time. It will get worse when AI will do shopping for us. Actually this might be the best way to monetize AI. They give us an AI assistant which will propose be highest price we will be willing to pay to merchants. Something like our own personalized marketer working against us. We should be careful if our future AI assistant will be closed source and cloud based!