teilo

Or, just don't use JSON for config files. There are plenty of human-friendly config file options, so there is no reason to frankenstein JSON in this way.

I rebuild these things. I know the market.

Because the energy of a photon is a function of its frequency.

DX7s don't go for a high price. The market is flooded with them. If you pay more than $600, you paid too much. You can easily find one in the $400 range.

This is strangely written. NL is not a character. It is an operation. Every time he writes "NL character" he means an LF, i.e. 0x0A.

The NL operation is defined differently for different OSs. Yes, CRLF is a relic of the teletype, but in practice, it's really the Windows convention inherited from DOS. L/unix has always been LF. On old Mac operating systems it was CR alone.

So that's what I missed at the keynote.

I use Z-Wave for everything. I looked into Home Assistant, but went with a Hubitat C8 Pro hub bridged to HomeKit on my Apple TV. A Z-wave multi-relay and tilt sensor kit is under $100, and just works.

Home Assistant is more flexible, and has nicer dashboards, but it's also way more of a PITA to get everything working consistently. Hubitat, OTOH, has just worked, with very little tinkering. I had some issues early on with it running slow, but later firmwares resolved all such issues.

Why are enterprise SANs so good at dedupe, but filesystems so bad? We use HPE Nimble (yeah, they changed the name recently but I can't be bothered to remember it), and the space savings are insane for the large filesystems we work with. And there is no performance hit.

Some of this is straight up VM storage volumes for ESX virtual disks, some direct LUNs for our file servers. Our gains are upwards of 70%.

No. This is just wrong and fails to understand not just how transformers work but the conceptual mapping that results from their training.

All these arguments about whether or not LLMs think are missing the point. They do not “think” as humans think due to their intrinsically transactional nature. But calling them “fancy statistical autocomplete algorithms” is also wrong.

LLMs contain within their matrix a massively high dimensional concept map. In this coordinate space, high order vectors map the distance between abstract concepts. This is a natural result of consuming language, which by its very nature is a symbolic concept map.

The uncomfortable question becomes: Is the human brain similarly using a massively high-dimensional concept map? Can a significant part of human thought be described as a fancy autocomplete algorithm? Can a significant amount of human reasoning be mapped as a nested series of transactions?

Actual article title: Benchmarking Qualcomm's NPU on the Microsoft Surface Tablet

Because this isn't about NPUs. It's about a specific NPU, on a specific benchmark, with a specific set of libraries and frameworks. So basically, this proves nothing.

It’s an on-device RAG.

These are the same people that convinced 3rd-world mothers to feed their babies freely provided formula until their breast milk dried up, and then the formula ran out, and they could not afford more, and their babies starved to death.

That pretty much describes all the social sciences. A morass of non-reproducibility, lack of controls, cherry picked data, bias, and subjective conclusions.

Neal Stephenson predicted this in Anathem.

Sorry. That was a major brain fart. Yes. 8-bit quantization, and using 49G of RAM.

No quantization (8_0). The full 48GB model. As for token count, I haven't tested it on more than 200 or so.

I'm running it on an M2 Max with 96GB, and have plenty of room to spare. And it's fast. Faster than I can get responses from ChatGPT.

Audacity, despite its weaknesses compared to commercial tools, still excels at batch processing due to its Nyquist plugin suite. The macro tool is finicky, but you can still do things that nothing else can in a batch, like trimming leading and trailing silence and then adding an exact amount of silence to the front and end of a file. You would think functions so simple and obvious as this would already exist in Audition, RX, SpectraLayers, etc., but no.

In other words, just like VSTs on Bitwig. Even if the entire audio engine crashes, you just restart it with a click.

Why would they want to support AU? The only reason to use AU is Logic. Everything else (Protools excepted) supports VST, and all the plugin devs release in VST and AU, so it would be a waste of time. Bitwig is putting their time into CLAP, and for good reason. It's cross-platform, much easier to develop against, and much more advanced that all the alternatives. Even Avid has shown interest in CLAP. So has Image Line. I expect Studio One to support it in v7.

JetBrains Mono has been nerdified, in four varieties, for a long time, and is already there. I use it with oh-my-posh.

There are non-ligature and non-monospace (for icons only - everything else is monospace) versions.

Ignore the sour grapes. Yes, Docker is constantly evolving. But the title does say "in 2023."

I, for one, need tutorials like this, and I am grateful for it. Devs who have been using Docker for a while have forgotten how intimidating it is for those of us who have our entire dev tooling built around VMs (orchestrated with vagrant in my case). You need just enough hand holding to get you up and running efficiently, and this tutorial does exactly that. It gets you over the hump a lot faster than trying to follow years-old and incomplete advice from Googling.

Yes, a thing or two may be deprecated. So what? We are used to that. But 99% of it still applies.

Yes, we care about flavors of Linux in the enterprise. RHEL is an industry standard distribution that nearly all vendors support. We have a significant amount of commercial software that is only supported on RHEL or its derivates. Not running on a supported distro? You get no support. That’s a non-starter for mission critical systems in the enterprise.

We used to be able to run this software, fully supported, on CentOS. IBM pulled the rug out from under us. Rocky and Alma rose to the occasion, and the various vendors supported one or both, since it was still RHEL minus branding. Now IBM is cutting off those distros also in an attempt to force everyone into RHEL licenses and support contracts.

But because it IS Linux, after all, a lot of enterprises like mine have NEVER needed support for the OS itself. We just needed our vendors to support their software running on it, and we needed vulnarability patching on the OS, something which CentOS, Rocky, and Alma always provided.

If I had my way, 100% of what we run would be on Debian. I have Debian running everywhere possible. But not all 3rd party applications support Debian. I’m hoping that one consequence of IBM’s move here will be to drive them all to Debian.

Yes, it turned out to be wrong, but not for the reason you cite. Dawkins completeliy dismisses the critical role of mutational fixation of epigenetic gene expression. He admits it exists, but dismisses its relevance in speciation.

RISC-V may become the open source ARM, or it may be a complete bust. At this point, either could happen. As it is now, there are only a few hardware implementations and all of them are dog slow compared to ARM.

And you think it will stay that way? It won't. This is just the start. Next, they will ban private servers and raid those who do not comply.

I agree that this is clickbait, but then I come from a family that just doesn't think this way. I've been through this twice now, for both my parents. They died at home. Not in a rented apartment. At home. We, their children and grandchildren, were their caregivers. We did everything a nurse would have done. We took shifts. We spread the burden among us.

Yes it was a sacrifice. But it was a sacrifice that was worth it, and nothing less than what we owed to our parents.

Age also has a big effect on serum levels. When you are mid-40s+, you cannot absorb D3 as well, and it takes a significantly higher dose to get your serum levels up. This is why testing is important.

It is also important to take K2 when you are taking higher doses of D3. Even if you have a higher than normal serum level, K2 mediates the metabolism of D3, and prevents issues. I've seen numerous examples of this, verified by CAC tests which consistently score 0 in people taking 10,000 IU of D3 with 100mg K2 for extended periods of time.

I use Privacy.com, which basically turns every card I use with them into a canary. The first time you charge on one of their virtual cards, they become merchant-locked. No other merchant can charge to that number, and if someone tries, I get an alert.

I have uncovered flaws in online merchants this way, and notified them. They were usually grateful, especially so since the fraudulent charges failed.

Chastity != celibacy. Chastity is the broader category. Celibacy is specific.

chaste, adjective, 1: innocent of unlawful sexual intercourse 2: celibate, 3: pure in thought and act, modest.