quailfarmer

LWIP integrates quite well with the FreeRTOS scheduler, I’ve used that combo before. How does the “plus-TCP” stack compare?

It’s always baffled me that the default configuration of conda takes the liberty of contaminating the default shell environment. It feels incredibly bold to assume that because I install conda I want it to exist in every context in which I use my computer. It’s the terminal equivalent to an adware “toolbar” installing itself in your browser.

Yes, we used them for a rocket launch in the desert. The main distinguishing feature is that these (CF-31 and related) are fully dust proof, the fan of a normal laptop will pull in dust and die. These were certainly slower to use, but we were just hacking around small bugs as they caused issues.

Don’t be afraid to implement a basic buck converter! Something like the MPQ4572 isn’t too hard to get right (right enough for hobby projects at least).

> Now I'm an adult, watching this kind of thing play out on a global scale every single day with the collapse of public health, the collapse of our democracies, the collapse of global industrial civilization, and the collapse of our ecosystems.

I must have missed the memo on this situation.

But more seriously, there’s such a thing as constructive criticism, and there’s such a thing as fatalism. And there are people who burn their own reputation by complaining more than is useful.

I'm sure this will find use in Business-Class "Mobile workstations", but having integrated DDR4 in my own hardware, I have a hard time seeing this as the mainstream path forward for mobile computing.

There's lots of value in tight integration. Improved signal integrity (ie, faster), improved reliability, better thermal flow, smaller packaging, and lower cost. Do I really want to compromise all of those things just to make RAM upgrades easier?

And how many times do I need to upgrade the RAM in a laptop, really? Twice? Why make all those sacrifices to use a connector, instead of just reworking the DRAM parts? A robotic reflow machine is not so complex that a small repair shop couldn't afford one, which is what you see if you to to parts of the world where repair is taken seriously. Why do I need to be able to do it at home? I can't re-machine my engine at home. It's the most advanced nanotechnology humanity can produce, why is a $5k repair setup unreasonable?

This is not to mention the direction things are really going, DRAM on Package/Die. The signaling speed and bus widths possible with co-packaged memory and HBM are impossible to avoid, and I'm not going to complain about the fact that I can't upgrade the RAM separately from the CPU, any more than I complain about not being able to upgrade my L2 cache today. The memory is part of the compute, in the same way the GPU memory is part of the GPU.

I hope players like iFixit and Framework aren't too stubborn in opposing the tight integration of modern platforms. "Repairable" doesn't need to mean the same thing it did 10 years ago, and there are so many repairability battles that are actually worth fighting, that being stubborn about the SOTA isn't productive.

To get good efficiency from a solar panel you must continuously track the “Maximum power point” (MPPT). You would never want to run a motor directly off the solar, you want a power converter to maximize efficiency.

This is cool, but if you want to see the proper extension of this style of computer, check out a Panasonic CF-31 Toughbook. They can be had for pretty cheap on eBay, and are amazing. In every way the opposite of the cost-cutting and cheap material you tend to see in consumer electronics. It’s the level of quality of a MacBook, but directed towards an ideal of “rugged” instead of “sleek”. Really a testament to what modern manufacturing can do when unbridled from the constraints of cost and popular aesthetics.

Maybe one day, but I’ve asked the various models basic EE interview questions, and it’s painfully obvious that while it can make a paragraph that has the right words, it’s incapable of reasoning spatially, or describing complex connections. My job is secure (for this year).

Hey, I’m very curious if you’ve seen any improvement in latency over the past month.

Unfortunately, the routing to make this feasible doesn’t exist yet. Users need a single IP address from a range that’s homed at a single PoP. Starlink doesn’t support user-user connections through the mesh, you need to go all the way out to your PoP, then over to the other users PoP, then back through Starlink to that user.

The US State department provided specific authorization empowering telecom companies (incl. Starlink) to operate in Iran and disregard local restrictions. If you can smuggle a dish into the country, you can use it. But the signal wouldn’t be too hard to detect, so it’s risky.

That’s a pretty significant jump in complexity, the RK3588 is one of the most powerful SOCs available. A cool part, but I’m not sure it fits the OpenWRT use case

You can get a pretty high quality PCB built at Bay Area Circuits in Fremont, and assembled at Sierra in Sunnyvale. The speed is marginally slower than JLC, the cost is marginally higher and quality marginally lower than Compeq. But, you can drive there, and be on a first-name basis with your supplier.

Given the extreme efficiency of international shipping logistics, I don’t think it’s reasonable to expect any given metro area to have competitive world class manufacturing of every type. The cost of real estate alone makes it hard to compete.

Active probes can draw several watts. Probably shouldn’t, but no one is optimizing them for efficiency

The analogy doesn’t hold because one of the hardest parts of this technology is building a device that can extract the optical energy as electrical energy, and dissipate whatever heat it generates. The power density of the fiber itself is easy by comparison.

Unfortunately because of spectrum rules it’s unlikely that cell plans will work outside of the regulatory coverage region.

> failed to achieve its aim with Falcon?

Yes, actually. But we aren't resting on our (very profitable) laurels. Development of Falcon slowed so attention could pivot to the new generation of even more reusable vehicles. Development on Merlin started around 2005, the first F9 was built in 2010, and first landing was in 2015. Raptor engine testing began around that time, and starship hopped in 2020. These things take time, but in truth we're right on schedule.

Hah, that video really ticked me off at the time. Technically true, but only in the least meaningful way possible.

I was just thinking back to my childhood homelab days. We would go to the regional recycling center and back our car up to the E-Waste trailer. One of us would pretend to move empty cases from the car to the dumpster, while another would crawl around and pull out treasures. I could field-strip a Dell Optiplex in 15 seconds, and pocket the RAM and CPU. I'm not sure we ever really made much profit, but we had fun doing it. I wonder if the waste contractor ever wondered where the high-value items were going?

Ah…

Maybe it’s because I just woke up, but why does a hexagonal pyramid not satisfy the definition? (Or any n-pyramid?)

Wow, perhaps I’m an exception to the norm, but this isn’t my experience at all. My family regularly sends interesting historical records of the lives of our ancestors. My great aunt composed a historical record of my great grandfather, who over his life built dozens of houses by his own hand. Even at university, I read a number of interesting historical letters and documents of the people who lived in the same dormitories in generations past.

I guess I may be ignoring all those documents that weren’t interesting enough to be remembered, but I imagine it’s hard to predict what will be interesting in the future. The fact that 99% of our lives are stored in computers vs paper would still vastly reduce the number of _interesting_ documents.

I think you’re in an uncanny valley where you want to have those powerful search features, but haven’t realized you probably wouldn’t actually use them. For ubiquitous passives, the DK search has all the relevant requirements, and then I’m just picking the manufacturer I have a relationship with. The supply chain on a single passive isn’t worth the headache to save a few percent.

Aside from passives though, you’re never going to get exactly what you want from any reasonably bounded database. For a FET you’re trading footprint, thermal impedance (both up and down), SOA, parasitic capacitances, turn on times.

When I’m picking a key part, I’ll start with a DK search, get a sense for which MFGs are leading the field, make a list of 10 or so parts that look promising, and go through the material on each, filling out my own spreadsheet as I go.

That’s even more true for anything with digital logic, where it wouldn’t be possible to sort by the enumerated features in a meaningful way.

A good way to handle the who-updates-the-updater issue is to use a triple partition updater. A updates B, and then B updates C, then C updates A. If anything about the new version prevents it from properly updating its neighbor, that neighbor won't be able to close the loop, and you'll fall back to A. This simplifies the FSBL, because it just boots the three partitions in a loop, no failure detection required. You don't need to triplicate the full application either, just the minimum system needed to perform an update, and then have the "application" in it's own partition to be called by the updater.

I recently got a BetaFPV Cetus X Pro. I was considering a DJI Mini, but FPV seemed like a lot more fun than the well-engineered computer controlled flight. And the vehicle costs much less to crash and destroy.

Those things are all possible, but you’re not going to find it in a COTS product.

The key terms in the link budget is directionality. If you use a phased array (like Starlink) or a parabolic dish, you can boost your range tremendously. This is how 4x4 MIMO WiFi can achieve gigabit speeds with the same power limit.

Plenty of Stanford Grads at Zipline! :)

Anyone care to recommend one of the two devices (Kobo Elipsa or RM2) mentioned? It appears the RM2 may have gotten more expensive since release? How is the relative build quality? Any key hardware differences to note?

Agreed with Hector, there’s so many bad-faith or misinformed opinions around this.

They should probably be obligated to provide calibration tools. However, the risk of providing those freely is that it increases the street value for stolen devices. The fact that a stolen iPhone can’t be erased and sold is a feature many people care about. Conveniently it also helps Apple’s bottom line. Someone needs to come up with a system to allow sustainable reuse without encouraging theft.