papercrane

The JDK team made a decision awhile ago to hold the Vector API until value types are final. That's why the API has been incubating so long.

> acquired a warrant IMO to compel the valid, non-destructive PIN be given over.

Compelling a PIN, even with a warrant, is legally questionable. Courts have held that it is a form of 'testimony' because it's compelling you to disclose something you know, while some state courts have ruled the opposite way.

In all likelihood the government wouldn't push it in this instance, to avoid creating any sort of precedent.

The problem will be aftermarket car alarms. There used to be a car that parked on the street near my apartment that would go off ever time a bus went by it.

> Settlement would mean it doesn't become legal precedence, right?

The summary judgment ruling from 2025 in this case will still be legal precedence, the settlement will mean that there is no trial for damages.

However, since this is a district court ruling it is non-binding on other courts.

Yes, the trick is having a "ground-truth" KB it can link to, and trying to avoid citogensis from happening. I'm not sure if it's a technique that's every going to scale because of that.

In this context, it would mean building a knowledge base the model can link to that is considered authoritative. So maybe the LLM is has some slop in it's training, and that causes it to try and output some junk, but if it can't match it to it's verified ground-truth KB it doesn't end up outputting it.

The problem is figuring out what is the authoritative sources.

It's when the models outputs are connected to sources.

So for example, you might have a company chatbot that responds with a summary of a Confluence page it was trained on and a link to the source.

Doesn't eliminate hallucinations, but it does reduce them and gives you a chance to verify.

It's possible they could AOT compile for iOS, but I think the idea they are unifying to be unlikely.

I have an Ergodox EZ for work, and I love it for that, but for my gaming I just use a typical keyboard, I couldn't get used to it for gaming.

Yes, because auto-boxing is just compiling to Integer.getValue under the hood, the bytecode for Integer.getValue(1) and ((Integer) 1) is the same. They'll both compile to something like:

   iconst_1
   invokestatic java/lang/Integer.valueOf:(I)Ljava/lang/Integer

    new Integer(10) == new Integer(10) // true

Before value classes this would always be false. The only time comparing Integer objects with == could be true is if Integer object was create by going through Integer.valueOf (or obviously if they were the same object reference.) By default the cached values where -127 to 127, but that is tuneable at runtime.

https://github.com/openjdk/jdk/blob/jdk-27%2B27/src/java.bas...

> I'd love to see pledge/unveil on (upstream) Linux - but I'm not holding my breath

There is Landlock now, I believe it would be possible to implement unveil and pledge on top of that.

There are a number of licenses that are named MIT that are all similar, but not identical.

The "Expat" here is the MIT license variant. It is referring to the Expat XML parsing library that first used this license.

Usually when projects these days use an MIT license this is the version they use.

The letter also said it was conditional on the combined entity maintaining investment-grade credit rating, which seems unlikely if the combined entity was saddled with $20B in debt.

> It works great as long as you squint just a bit, ignoring that it generally calls for long,lat and is designed with the assumption of a world CRS.

I thought the spec allowed you to specify the CRS, but I just checked the RFC and they removed that from the 2016 specification and WGS84 is specified. It does allow for alternative CRS with prior arrangement, but like you said that does require a lot of care.

The hot dogs are 1/4 lb, with the bun it's probably 500-600 calories.

Yes, they have a number of domain names, archive.is and archive.today are the most well known ones.

It's almost a certainty that you can't copyright code that was generated entirely by an AI.

Copyright requires some amount of human originality. You could copyright the prompt, and if you modify the generated code you can claim copyright on your modifications.

The closest applicable case would be the monkey selfie.

https://en.wikipedia.org/wiki/Monkey_selfie_copyright_disput...

You're using too much! Its commonly used to improve meat texture, especially in Chinese cuisine. It's called "velveting".

The thermal gradient in space is meaningless because there is hardly any matter to dump the energy into. This means you are entirely reliant on thermal radiation. If you look at the numbers given by Stefan-Boltzmann law you'd see that means to radiate a significant amount of energy you need a combination of a lot of surface area and high temperatures.

This means you need some sort of heat pump. For a practical example you can look at the ISS, which has what they call the "External Active Thermal Control System" (EATCS), it's a complicated system and it provides 70kW of heat rejection. A datacenter in space would need to massively scale up such a system in order to cool itself.

Is there a credible way to cool a space-based data center on that scale?

I love this book! I do wish there was a new edition that updated the version of Java used in the tree-walk interpreter. There's been some additions to the language, like sealed classes and exhaustive switches, that could really benefit the implementation.

The DMCA notice is available here: https://github.com/github/dmca/blob/master/2025/12/2025-12-1...

The notice has a list of files and says that they were copied from ffmpeg, removed the original copyright notice, added their own and licensed under the more permissive Apache license.

The memo at the time said the serifs can cause OCR issues.

https://x.com/John_Hudson/status/1615486871571935232

One of the reasons Calibri was selected over Times New Roman was it has a lower rate of OCR transcription errors, making documents using it easier for people using screen readers.

It could've been, but it was probably a turbocharged L-series engine.

When Doyle wrote most of the Holmes stories cocaine was a popular and novel new drug, it wasn't until later that it's risks became widely known. In one of his later stories, "The Adventure of the Missing Three-Quarter", Doyle portrays it as an addiction that Watson weaned him of, but is still concerned that his friend may fall back into.

"For years I had gradually weaned him from that drug-mania which had threatened once to check his remarkable career. Now I knew that under ordinary conditions he no longer craved for this artificial stimulus, but I was well aware that the fiend was not dead but sleeping, and I have known that the sleep was a light one and the waking near when in periods of idleness I have seen the drawn look upon Holmes’ ascetic face, and the brooding of his deep-set and inscrutable eyes. Therefore I blessed this Mr. Overton, whoever he might be, since he had come with his enigmatic message to break that dangerous calm which brought more peril to my friend than all the storms of his tempestuous life."

- https://en.wikisource.org/wiki/The_Return_of_Sherlock_Holmes...

In the US copyright just requires a level of originality. The bar isn't very high, but for example simple logos, like IBMs blue lines logo is not copyrightable.

There are examples of software code that is probably not copyrightable, but that's limited to very simple code that has only obvious implementations.

I believe the RTL in RTL-SDR is "Realtek Limited", the manufacturer of the chips used in the early days of SDR. I don't think the chips these days are exclusively Realtek, but the name has persisted.

For clarity, they're requiring apps to be signed by a verified developer on certified Android devices. You can still side load, but the verification is still required for the side loaded apps.