pkage

I think the problem comes down to fingerprinting. For example, resources used to be cached in the browser regardless of site exactly as you described, but then ad companies figured out you can track a user's browsing history by timing the cache accesses to see if a user had loaded a bundle from another site. To remove that vector, browsers split up caches per domain.

What does this have to do with high speed rail?

> And then imagine that you can buy — hopefully you can buy — an apartment in the $400,000 range, or a nice house for $600,000. Those figures are a little bit up in the air, but it will definitely be more affordable.

These... are still relatively unaffordable housing prices.

What is the relationship between Rye and uv? Aren't they now both under the Astral.sh umbrella?

From the video description:

> A passively actuated gripping mechanism grasps the powerline cable during landing after which a control circuit regulates the magnetic field inside a split-core current transformer to provide sufficient holding force as well as battery recharging.

any tips on making the jump to Kagi? I've been considering using it for a while now and it's just been inertia keeping me on Google

For an identical price point you can get a brand-new ASUS vivobook[0] which features:

- a 15" 1080p (!) screen

- 6 hour battery life (vs 18h)

- plastic frame (vs aluminum)

- inferior cpu performance (Intel Core i7 1255U vs M1)

Though it does have some pros:

- 16gb of ram (vs 8gb)

- 1tb of disk (vs 256gb)

Seems like the M1 MBP is a great option and I'm glad it's more accessible.

[0]: https://www.walmart.com/ip/ASUS-VivoBook-15-15-6-Touchscreen...

Looks great! This is super cool. Where do you source your information for object locations?

Other way around, 10^9 bytes is a gigabyte (GB) and 1024^3 is a gibibyte (GiB).

As always with studies like these, there's the question of applicability of the measures to the hypothesis.

Is the measure of cognitive ability relevant to the claim made in the paper? That is to say, is the measure of cognitive ability ("g" in the paper[0]) actually a good measure of abilities and relevant to long-term financial optimism?

Are both these variables conditional on a third hidden variable? This isn't thoroughly addressed in the paper, the authors choose a series of variables "which are thought to be in principle unaffected by cognitive ability" but even there show that optimism is dependent on e.g. age (Measures -> Control Variables), and additionally mention that "g" is also dependent on age (Measures -> Cognitive Ability).

A high p-value doesn't necessarily indicate that the two variables are causally related, just correlated—and there are plenty of highly correlated datasets that are definitely not causal[1].

It's important to try to disentangle the claims from the data supporting them. The claim that people with lower cognitive ability are more (recklessly) optimistic about their futures seems like it would make sense at least from a social-Darwinist perspective, but it's exactly those claims that seem reasonable that we should examine more closely. Confirmation bias is a powerful set of blinders.

[0] https://journals.sagepub.com/doi/10.1177/01461672231209400 [1] https://www.tylervigen.com/spurious-correlations

I'd also be interested—is there anywhere we can take a look?

Beepys have a pretty big issue with randomly shorting out due to a design flaw, according to the Discord forum anecdata. It happened to mine, unfortunately.

A very capable little device before it cooked itself, though—having a raspberry pi in your pocket is a lot of fun.

I just wish you could turn the multilingual keyboard off—I find that I usually only type in one language at a time and having the autocomplete recommend the wrong languages is quite frustrating

I actually went to middle school in a former converted Nike missile radar site, we had the radar up and a lot of old hardware lying around. Definitely a cool experience as a kid.

I'm going to guess it's ISO.

You don't pedal on an scooter, so the battery needs to be big enough to carry around both you and the larger battery (adding weight). Conversely, e-bikes can assist your pedaling (rather than having to replace it in all cases) so they can be lighter.

Agreed, top comments smacks of gatekeeping. I work in both Python(ic) languages and in C/C++, both have their strengths and weaknesses. No need to start flame wars.

Combining embeddings is the backbone of multimodal LLMs, such as InstructBLIP[0] or LLaVA[1]. Those architectures take the output tokens from a (frozen) vision transformer and train a very small projection layer between the output token space of the ViT and the input space of the LLM.

[0] https://arxiv.org/abs/2305.06500 [1] https://llava-vl.github.io/

That just moves the trust requirement up the chain, and assumes that the image wasn’t compromised pre-signing (either intentionally or unintentionally)

This is also the case in software, improvements in hardware performance simply cause software to consume more resources, and not meaningfully increase the overall productive capability of a system.

How does this compare to a traditional ORM? I looked through the slides and the paper and this seems to largely provide the same kind of functionality as a regular ORM but with the caveat that you can only do key/value transactions.

I’ve had good results with SD1.4/2 with MPS acceleration on similar hardware (M1 Max, though with 64gb). No stability issues with MPS, either. I’d say don’t rule it out just yet.

The Linux distribution of Prism actually runs in a flatpak by default, which provides pretty OK isolation.

HTTPS seems to be broken and the HTTP link brings me through a series of crapware redirects before landing somewhere random.

Is ... that intentional?

I strongly recommend TablePlus for sqlite/postgres/other databases, it's made my life significantly easier.

https://tableplus.com/

It's a CSS animation from the Jekyll theme. Pulled from the minified CSS:

    @keyframes intro {
        0%   { opacity: 0 }
        100% { opacity: 1 }
    }

Applied via:

    animation: intro 0.3s both;
    animation-delay: 0.15s;

It seemed to be briefly listed on Steam, though the link is now dead. Here's the Wayback Machine archive: https://web.archive.org/web/20230322031642/https://store.ste...

Wouldn't that be a gigagram? Even more cool.