jdyer9

My thoughts as well, I love this!

Easy to do, easy to implement but hard to bypass. Also it tells me something about the network that is not vying for a slice of the attention economy and isn't going to do everything it can to keep me on the site.

This comment got me thinking that it might be worth using their second-to-last location to try and derive some vector. Obviously that's super informative as you already know the edge of the map they left, but maybe it's useful?

It's a good question, but in a closed system (like you have in space) the heat from the turbine loop has to go somewhere in order to make it useful. Let's say you have a coolant loop for the gpus (maybe glycol). You take the hot glycol, run it through your heat exchanger and heat up your cool, pressurized ammonia. The ammonia gets hot (and now the glycol is cool, send it back). You then take the ammonia and send it through the turbine and it evaporates as it expands and loses pressure to spin the turbine. But now what? You have warm, vaporized, low pressure ammonia, and now you need to cool it down to start over. Once it's cool you can pressurize it again so you can heat it up to use again, but you have to cool it, and that's the crux of the issue.

The problem is essentially that everything you do releases waste heat, so you either reject it, or everything continues to heat up until something breaks. Developing useful work from that heat only helps if it helps reject it, but it's more efficient to reject it immediately.

A better, more direct way to think about this might be to look at the Seebeck effect. If you have a giant radiator, you could put a Peltier module between it and you GPU cooling loop and generate a little electricity, but that would necessarily also create some waste heat, so you're better off cooling the GPU directly.

So, in doing a bit of research from a link in one of the other comments, this is lcos, levelized cost of storage. I understand that to be roughly equivalent to the marginal cost of using it, including the capex divided over the unit volume. That same article uses $125/kwh as the capex, which is in line with your (and my) expectations of the cost to install.

$65/mwh works out to $0.065/kwh, so that makes sense. Effectively you can read this as "it costs $65/mwh to store and then consume electricity using these batteries"

Oh man, I've been playing with GCP's vertex AI endpoints, and this is so representative of my experience. It's actually bananas how difficult it is, even compared to other GCP endpoints

If it matters to you, Reolink is a Chinese owned company. Not passing judgement one way or another, but if avoiding Unifi over the remote incident matters, I could see this factoring in as well.

Kharon | Individual Contributor and Data Engineering Leadership Opportunities | REMOTE or HYBRID in London, Madrid and Denver | Full Time - https://kharon.com/careers

Kharon is on a mission to revolutionize the current global security landscape. When you look at any major global crisis event, we’re providing intelligence that’s at the heart of those circumstances. We connect the dots in a way that’s meaningful, and we're currently growing our Data Engineering vertical. Operating at the intersection of global security + international commerce Engineering team tackling massive open-source data at scale Stack: Python (Pandas, NumPy, FastAPI), SQL, Spark, Databricks, Neo4j, Elasticsearch, AWS, Docker, K8s Feel free to check out open roles at the link above.

Kharon | Individual Contributor and Data Engineering Leadership Opportunities | HYBRID in London or Madrid or On-site in Denver, CO | Full Time

Kharon is on a mission to revolutionize the current global security landscape. When you look at any major global crisis event, we’re providing intelligence that’s at the heart of those circumstances. We connect the dots in a way that’s meaningful, and we're currently growing our Data Engineering vertical. Operating at the intersection of global security + international commerce Engineering team tackling massive open-source data at scale Stack: Python (Pandas, NumPy, FastAPI), SQL, Spark, Databricks, Neo4j, Elasticsearch, AWS, Docker, K8s

Apply here: https://kharon.com/careers

This is exactly the system I would have designed!

Nitpicking on the mechanic point, but this is pretty common, just not at the same level of detail as medicine. Certain brands, models, and parts are more likely to fail in certain ways, so if a model comes in with symptoms of a known, high frequency problem, many times that work will be done first rather than taking more of the car apart to inspect individual parts.

Certainly I didn't think there's huge bodies of work on those statistics the same way there is for medicine, but any car repair forum online will give you some sense of this

Imagine that your reputation didn't matter, getting sued was out of the question, and there was no criminal liability: your job, backed by the government, is to be employed by as many tech firms for as long as you can, you'd probably work pretty hard on coming up with a reasonable but very good resume and work hard on how to interview well. Now, you're a professional interviewer and might conduct 10x-100x more interviews than your average dev, and have a network of people helping you optimize your cheating.

Given that background, I personally find it unsurprising that they're having success and AI tools are just making it that much easier

For what it's worth, the article says the exact opposite

  Protip: Look for all-caps PYREX graphics which can either indicate that is vintage or that it’s from Europe, where a company called Arc International owns the Pyrex brand and still makes its cookware out of borosilicate.

It's likely often repeated because if you try driving 55 in a 55mph zone where people are driving between 62-70, it'sterrifying, it feels like you're stopped. Whether the stat is true or not remains to be seen, but intuitively, it makes a lot of sense. Sure, your risk of rear ending someone at that point is probably negligible, but the odds of being rear ended? Hard to say

I always struggled with loving dark mode, but I usually keep my brightness as low as tolerable without straining, and your comment made me wonder if it's related I do like dark mode if I'm working in a super dark room though

I'd argue it's about customisation, not as much about the process. A couple can't be made anywhere other than a factory, but we're seeing this exact thing happen more and more with custom chip design. Sure not hand made, but skewing towards custom. Previously, there were a handful of CPU's and that was it, but then GPUs came along, and now Tensor, Graviton, etc all with more specialized purpose. Maybe it won't be too long before people start building fabs in their basement, but I doubt it.

Also, custom PCB's are getting easier and easier, it seems like there's a ton of YouTube videos sponsored by and about doing custom PCB's for projects in 2024. Heck, I have an Air gradient in my house, and while it's not a custom CPU, it is a custom, relatively low volume PCB designed by (as far as I can tell) a single person, almost like an artist...

I'm not trying to add a lot to the conversation, just suggest that even chips are moving towards customisation and artisanship.

As for how to do it, based on my reading, imagine a mattress on the floor. And then, to follow your comparison, imagine a prayer, particularly an Islamic prayer pose (knees on the floor, hips in the air, head and hands touching the floor). Except as the author describes it, your knees would be up on the mattress, and your head and hands would touch the floor, rendering your torso largely inverted.

Also, I realize that there's a lot more to the Adhan than just this pose, but it's the only easily searchable thing I can think of at the moment.

As another commenter started to point out, the risk is essentially cloning the key. So, if you were out to dinner and it was cloned while you were out, you might not realize you'd been compromised whereas if it was stolen, it might too late, but you'd know. It seems that for many/most people the risk is low, but anyone at risk of a state sponsored attack should be aware.

For the threat model of keeping out random online attackers with no physical access, it seems this vulnerability doesn't matter.

Same, I don't even like washing my hands with cold tap water, but cold bidet water is no issue.

Except Walgreens. They say unsubscribe and then they just don't do it.

Presumably the difference with a condo is that the condo owner has a fair vote in the buildings management (usually an HOA in the US) and enough votes can be used to destroy the building if it achieves high enough liability-to-value ratio. That does not exist for a timeshare, the timeshare owners can never band together to change the governance of the building

That's the thing about Mongo, it's great until it's not. And Not comes quickly for many projects.

There's probably tons of small Mongo installations all over the internet doing great in their little niche. But the moment a project gets too big for Mongo, it becomes a problem very quickly. I suppose it's like comparing a bike to a car. If your problem is getting around a small town with nice weather, the bike wins. But the moment you need to move 5 people on the interstate to the next town in bad weather, there's no real choice besides the car.

That's been my experience with mongo every time

I have no affiliation to either locality, but I thought this whole saga was interesting from a legal perspective. The sheer audacity of the developer to sue for unjust enrichment. I think it also would have been a pretty important case if it hadn't gone this way, as others pointed out, it would have opened a lot of potential avenues for developers to abuse neighbors or steal land.

Similar, with a slightly different goal: https://starlink.sx/ It was designed more with the intent of evaluating coverage of Starlink, especially in the early days when coverage was more sparse, but it also shows the satellites in orbit. The total numbers do have a slight mismatch between the sites.

My guess is that it's not that significant of an increase in weight. They're removing everything including the axles, and putting in their own. Guestimating based on the video, I'd say each motor+axle weights similarly to a transmission, the engine is a lot smaller but add the generator, that's probably a wash. Batteries, they're planning to be small and cheap (because they have a generator), so lets guess 20kwh, so maybe 300lbs in batteries. Fuel tank stays, drive shaft is out. So, basically a total increase of one transmission (150lbs) and 300lbs worth of batteries, so 450lbs in all. Definitely heavier, but I think it's a far cry from 10x. 2x would be my guess, which seems well within the realm of reasonability

It's a long watch for sure, and it's just laying out the plan. To answer your question, basically they want to build locomotive style kits (diesel power, electric drive train), except with the addition of batteries. So, 3.6L Caterpillar diesel (next year) to generator, with smallish batteries and two electric drive axles, and probably update dashboards depending on the kit. Cool plan, but if you don't have 40 minutes now, I'd wait and watch later in the year as they release more and talk about the build process.