Absolutely, it is essential to always run the numbers. I was once offered a sizable rebate if I accepted a non-0% car loan, but no rebate if I paid cash or had my own financing. Since their loan had no early-repayment penalty – and I demanded this in writing – I accepted their loan and paid it off upon the first statement.

My suspicion is that that sort of offer was to boost the commissions earned by the loan brokers, rather than to move cars. Or maybe both. Who knows.

0% interest offers show up fairly frequently in the USA, often as general-purpose credit cards, or for car or furniture payments, in addition to the many buy-now-pay-later services that allow financing almost anything. However, the motives for offering 0% are slightly different for each of these products.

But answering the question directly, a 0% offer is beneficial if you were already going to make the purchase and would finance it. Cheap credit makes it easy to overspend, since the payments will be “tomorrow’s problem”. For people who can afford to pay for something in full, it might still be beneficial to finance with 0% just to conserve cash on hand. But the tradeoff is having to service the debt with regular payments; missing one payment can cause the debt to resume at an exorbitant rate. It takes a decent amount of financial discipline to make a 0% offer work in your favor.

Going back to why 0% offers even exist, I’ll use furniture and cars as they’re the historic examples. Furniture is expensive, whether it’s a sectional sofa or a queen-size bed with frame and storage. There’s also a sizable markup for furniture, and competition between furniture stores is strong. Thus, to help entice people to buy furniture, sellers will offer 0%, outsourced to a loan company, with the loan subsidized by some of the profit margins.

For cars, the equation is slightly different. Sure, cars are an order of magnitude more expensive, but that also means the opportunity cost for dealers to offer 0% is correspondingly larger. Instead, 0% financing for cars is almost always subsidized by the manufacturer, not the dealers. This is a financial and business strategy that allows a car company to create more sales in a given quarter, if perhaps they need to meet certain year-end targets but are reluctant to reduce their list prices.

0% car loans induce more sales fairly quickly, but will draw on the company coffers in the years to come, because the loan company still wants their cut to be paid by someone. Consumers will usually benefit from these offers, as it’s rare for people to buy a new car outright.

It’s my opinion that if a car company has to subsidize loans to move their product, that’s a tacit admission that their product is wrongly priced or the competition is better. I would take this into consideration, although it wouldn’t necessarily carry the day when considering a purchase. After all, car payment interest is not insignificant.

A quick “rule” is to see how old the word/concept is. “plaintiff” would have existed almost as long as the English legal system came into being, or probably even older to the court of Assizes pre-12th century.

Whereas firefighter as a profession might have only become a word after the establishment of fire departments by insurance companies, which I think might have been a 19th century development.

This entire series by Cathode Ray Dude is a wonderful dive into the world of PC boot sequence, for the folks interested in a touch of embedded architecture. His delivery is also on-point, given the complexity and obscurity of the topics.

From this video alone (41:15):

The way this worked was: they installed Xen hypervisor on your PC, put Hyperspace in a VM and Windows in another. Now, you either know what a VM is – and I don’t need to explain why this is terrifying – or you don’t and I need to make you understand so you never independently invent this.

And (43:59):

This is just a bad idea, ok? Virtualization belongs in data centers. Putting some poor bastard’s whole OS in a VM is a prank. It’s some Truman Show shit. It’s disassembling the coach’s car and putting it back together inside the gym. It’s not remotely worth the trouble and it probably didn’t work.

My prior comment on hydrogen mobility:

Hydrogen for mass- or space-constrained mobility (eg bikes, automobile, aircraft) faces all the known problems with storing it inside inconvenient shapes and contending with the losses from liquification. Real Engineering has a video on this aspect (Nebula and YouTube) when compared to simply using battery-electric storage.

With that out of the way, I’m skeptical as to the benefits touted on the HydroRide website. Specifically, the one about storage:

Hydrogen storage offers extended longevity, surpassing 10 years, ensuring reliability and sustainability over time.

This might be true in static conditions, but hydrogen automobiles have to vent some of the hydrogen while parked, simply to deal with the buildup of hydrogen gas, since even with excellent insulation, the liquid hydrogen will eventually get warm and evaporate into gaseous hydrogen, building up pressure. The fact is that automobiles must withstand broad environmental factors, especially temperature. And we expect bicycles to do the same: how the hydrogen tank would behave in warm climates is unclear.

There’s also not that much hydrogen in the tank. The website appears to indicate 20 grams. At 33.6 kWh/kg, the total energy in the tank would be 672 Wh, putting it at par with electric bikes of similar range and speed. Any hydrogen losses would be balanced against battery capacity loss over time.

Overall, as the article states, the target audience of rental operators might still be inclined to go with battery electric bikes rather than hydrogen. Requiring a supply of pure water in addition to electricity at charging locations – compared to just electricity for battery charging – is an extra logistical consideration. The “charge” time of 5 hours for 20 grams of hydrogen is also a potential issue.

There was a ton of hairbrained theories floating around, but nobody had any definitive explanation.

Well I was new to the company and fresh out of college, so I was tasked with figuring this one out.

This checks out lol

Knowing very little about USB audio processing, but having cut my teeth in college on 8-bit 8051 processors, I knew what kind of functions tended to be slow.

I often wonder if this deep level understanding of embedded software/firmware design is still the norm in university instruction. My suspicion has been that focus moved to making use of ever-increasing SoC performance and capabilities, in the pursuit of making it Just Work™ but also proving Wirth’s Law in the process via badly optimized code.

This was an excellent read, btw.

To say this is a complex story is an understatement, and the article does a fairly reasonable job at unpacking the details. That said, I’m disappointed they did not include a PDF to the filed claim, since even a cursory analysis of the claims relies heavily on the details therein.

For reference, in California and other states, lawsuits against the state itself, or the counties, cities, agencies, special districts, or anything else that came into existence by the law, must have first filed a claim with the entity being sued. The California Tort Claims Act requires this procedure in order to relieve the state’s inherent sovereign immunity and also permit the courts – which are also an arm of the state – to have jurisdiction.

Focusing on the relevant facts to the claim, the claimant (the former Pedal Ahead manager) states:

• Pedal Ahead is a non-profit, distinct from Pedal Ahead Plus, a for-profit company
• Pedal Ahead was contracted by San Diego Assoc of Govts (SANDAG) and CA Air Resources Board (CARB) to operate an ebike program
• The claimant was hired in 2022 to work on this ebike program
• The claimant did not receive paychecks after April 2023

Accordingly, the claimant brings back-pay claims against SANDAG (for $40k) and CARB (for $58k), based on the premise that he was working on behalf of those two state entities. Presumably, the claim argues that he was unlawfully terminated for whistleblowing, but the article doesn’t specifically say this and we don’t have the PDF of the claim itself.

This is… an uphill battle, IMO. Not the whistleblowing part, but the part where he wants money from those two entities. The state action doctrine distinguishes between things done by – or instructed by – state employees and representatives, versus things done by private entities and their employees. In a lot of ways, this mirrors the distinction between hiring W-2 employees versus 1099 contractors.

If he were directly on the CARB or SANDAG payroll, this would be a slam dunk. But he ostensibly only had employment with the non profit. This means the scope any state action claim will depend heavily on the text of the contracts with CARB/SANDAG. Generally, the narrower and more specific the contract, the less it would be the state’s responsibility to pay up. That said, the contract can’t be hiring the nonprofit to do something which CARB or SANDAG normally couldn’t do on its own.

As with most things in civil law, it’s a careful balance to apportion liability onto the entity which is best positioned to avoid the problem. Would making CARB/SANDAG pay $100k – and thus letting Pedal Ahead off the hook for $100k – address the problem of constructively firing someone who noticed malfeasance? That’s left as an exercise to the reader.

What often happens in civil litigation is the “shotgun” approach: aim the lawsuit at as many people and entities as possible, especially ones with deep pocket, hoping some will settle. So it’s possible the claimant will later file suit against all of Pedal Ahead, its owner personally, every individual mentioned personally, plus SANDAG and CARB.

California allows for joint and several liability of money claims, so if a court/jury finds that CARB or SANDAG were even 1% partially responsible, then 100% of the winnings can be extracted from their deep pockets. This is especially relevant if the entity with the most responsibility – possibly Pedal Ahead – is broke or is about to go bankrupt. Joint and several liability prioritizes claimants getting compensated, even though the most responsible party escapes via bankruptcy. Again, civil law is a balance.

What’s truly staggering is the degree of corruption alleged, which is only ancillary to the tort claim, but it suggests that there are villains everywhere. No one comes out looking good here, whether its SANDAG and its other scandal, the owner of the nonprofit and his other ventures, the ebike program itself (why have a hard goal of 100 miles per week?), the former San Diego County Supervisor who helped award the contracts and then resigned after a sexual assault scandal, and the state entities for not properly auditing Pedal Ahead and still paying out the contracts.

All the meanwhile, $30 million of the $31 million dedicated for ebikes remains unused, likely to the disappointment of lots of San Diegans who are eager to take up ebikes or participate in a pilot program. That’s the real travesty amongst all this.

At this point, the bill has already passed the Senate and the Assembly, so it’ll soon head to the governor’s desk, as is the norm around June of even-numbered years.

If you are a California constituent who cares about bicycle and ebike legislation, I would encourage you to sign up for Calbike’s email updates on impending legislation of great significance to bicycling in this state. This bill is fairly minor, but who’s to say something more momentous won’t come next year?

Likewise, if your region has its own bicycle advocacy group (eg San Francisco Bicycle Coalition, Sacramento Area Bicycle Advocates, BikeSD), get updates from them too.

Today’s ebikes have only been legalized since 2016, so what is advocated for now will set the tone for decades to come.

I’ve been tracking this bill since it was first proposed, and the author never got back to me on exactly what is within the scope of the bill. The relevant part of the bill reads:

(e) A person shall not sell a product or device that can modify the speed capability of an electric bicycle such that it no longer meets the definition of an electric bicycle under subdivision (a) of Section 312.5.

For example, would a generic serial programmer device fall within the definition of such a “product or device”? How about software that is used with that device to modify the top speed of a specific ebike model? Common sense would say “no” to the first, and “maybe?” to the second.

But we can look to the author’s own comments before the Assembly Transportation Committee:

According to the author, “AB 1774 will ban the sale of electric bicycle (e-bike) speed modification devices (i.e. tuning kits) in California. These devices modify e-bikes far beyond the legal definition of an e-bike in State Vehicle Code.

But the committee staff already see the obvious flaw:

While this bill would prohibit the sale of devices that allow an e-bike to go beyond the speed permitted by law, many e-bikes can be modified to bypass the speed limiter by reprogramming the bicycle. The limiter can be tricked by changing the tire size on the panel on the bicycle. Other e-bikes, like the Super 73, allow you to change the speed of the bicycle using an app that comes with the e-bike.

I would hazard a guess that ebikes manufacturers might start to do the same trick that e-motorbikes do, which is a wire that – if severed by the owner – would disable the speed governor.

As is somewhat typical now in the Legislature, these bills written against ebikes tend to come from SoCal, ostensibly because ebikes are an exceedingly popular alternative to driving on congested roads. What bothers me is that these representatives are not taking in the whole reality, which the committee staff report kindly summarized:

• Ebikes are outselling electric automobiles in the USA, by double digit percentages
• Ebike injuries are up, although the LA Times reported it as the low three digits in a 2023 article
• The best way to increase safety for all cyclists is through sheer numbers: more cyclists

Given that reality, legislative effort apparently isn’t following evidence but rather towards clumsy distractions that emulate a “tough on ebike” stance, as bizarre as that sounds.

This sort-of happened in the USA, in a small way, during the fallout of the 2016 Wells Fargo scandal. Public sentiment of the big-name, national retail banks was awful and credit unions capitalized on the moment with advertisements contrasting profit-centric national banks with local, cooperatively-owned credit unions.

In this article where consultants to credit unions were queried a year later, there’s still some questions as to the long-term effects that may have benefited the credit unions.

I once came across a comment somewhere online that suggested – sadly without hard evidence – that the scandal may have been a win-win, since the sort of customers willing to uproot themselves from Wells Fargo tended to have smaller balances while still incurring the bookkeeping costs. And that credit unions were able to scale up to take in new customers while saving on advertising dollars.

It’s a plausible idea, that a new equilibrium would be found in the banking market. Logically extending the idea further, though, would lay bare how much additional integration credit unions would have to do with each other to achieve a truly seamless customer experience. Of course, with more young people mostly sticking to online and mobile banking, this might come in the form of backroom operational improvements, rather than a revamped brick-and-mortar experience.

Most commercial publications in the USA and UK – UPC/ISBN or not, regular or not – will often send copies of their work to the national library (ie Library of Congress). That said, those copies might not be prioritized for digital viewing. So seeing them in-person might be the only way to access them.

As for whether the publishing houses keep them, it’s probably very individualized, so who can say.

Couldn’t it still fall forward/backwards though when stationary? That doesn’t seem like it’s very useful even when compared to a conventional unicycle that can fall over in any direction when stopped.

Looking at this photo as an engineer, I decided to have some fun and see if there’s a way this design could somehow be a unicycle after all, based on no prior knowledge.

Supposing that maybe the two wheels turn independently, this just ends up being topologically identical to the so-called hoverboards. Except that hoverboards are configured with small diameter wheels and set to a wide gauge, allowing them to make zero-radius turns without scrubbing (aka tank steering). But this contraption, with its large contact patch and narrow gauge, would surely scrub quite hard for zero-radius turns.

Ok, so that doesn’t really work. Maybe it’s a solid axle. But… that’s somehow even worse since its performance is nearly identical to a single, wide tire. The Onewheel skateboard is proof positive that a single, wide tire is valid, but the crucial aspect is that the Onewheel still banks like a bicycle or airplane, rolling onto the edges of its tire. So too must this Inmotion unicycle, but then the outside tire is less involved in stabilizing the turn.

The only example I can think of where it’s beneficial to split one large contact point into two smaller ones is at sea, where a catamaran uses two parallel hulls for its hydrodynamic benefits. And for sharp turns, one hull lifts entirely out of the water, which is desirable. Whereas lifting a wheel off of land is not.

I can only see downsides with the dual wheel, solid axle unicycle design, reminiscent of this video by Practical Engineering on rail wheels (Nebula or YouTube). In short, optimizing for stationary stability means giving up dynamic stability while in motion, either inhibiting a turn or turning uncommanded. Trains use solid axles with conic wheels for dynamic stability and rely on truck/bogie geometry for stationary stability.

TL;DR: I can’t come up with a good reason for this design

My prior comment on hydrogen mobility:

Hydrogen for mass- or space-constrained mobility (eg bikes, automobile, aircraft) faces all the known problems with storing it inside inconvenient shapes and contending with the losses from liquification. Real Engineering has a video on this aspect (Nebula and YouTube) when compared to simply using battery-electric storage.

With that out of the way, my takeaway from this article is that manufacturers are working on broadening the appeal of bikes and micromobility with new models, be it harkening to the classic “Dutch bike” or designing for less smooth surfaces. Some are clearly gimmicks and fads, or advance the surveillance state. But by and large, innovation is taking place, proof that this market will not stand still for even a moment.

If you have a particular job focus for after you’ve graduated with your CS degree, would an internship with a related company be an option? Experience with web will be of limited use for an embedded job, and embedded experience is of limited use at a quantitative analysis company.

That’s not to say the experience is entirely pointless, since many skills across the various disciplines of CS are transferable.

I would say the idea is right, but the focus is wrong. In order of what will have the most beneficial improvement, and not just for cycling, people can:

• wear correctly-sized shoes
• adjust the bike to fit them properly
• address all other problems as they arise

This 2018 report concluded that:

A large proportion of the population wear incorrectly sized footwear, which is associated with foot pain and foot disorders.

Good shoes are something that will (literally) support you in all your endeavors, whether it’s biking to your destination or help you get around once you’ve dismounted and parked the bike. The author’s laser focus on cycling shoes that are rigid and don’t store energy is admirable, but the same shoe would be deeply uncomfortable for any amount of walking. A holistic approach to footwear would mean optimizing first-and-foremost for good fit, which is always applicable and is what most people get wrong. And only then consider features or extra pairs of shoes for specific uses, such as cycling.

I would guess that many casual riders are uninterested in hauling bike-specific shoes with them, so the shoes will have to be well-rounded, rather than task-specific. But that brings me to the next thing to optimize.

The venerable Sheldon Brown wrote many articles on how to adjust bike fitment to address pain in various parts of the body. Even if nothing hurts, I would encourage people to have a cursory read, as an ounce of prevention.

To drive my point, the author is correct that a proper cycling shoe improves power delivery and comfort at the feet. But a properly fitted bike improves the entire body’s experience on two-wheels overall, without compromising walking performance. And adjusting a bike is basically free.

As usual, I also need to point out that the editor of that publication has chosen to reframe and dramatize the headline.

When it comes to what insurance does or doesn’t cover, the best answer will come from the text of the policy itself. This is, unfortunately, very dry reading and most people – although instructed to keep a copy handy – don’t have the full text nearby. That said, because of the regulated nature of insurance in the USA, standardized forms of policies exist, and homeowner policies are no exception.

The common homeowner policies are numbered HO-1 to HO-8. HO-1 only pays out only for the ten listened “perils”, and is thus the most barren policy available. Not all HO-1 policies are verbatim identical, but the gist usually matches.

We can look at this sample text from a random HO-1 (issued by American Family Insurance). Page 5 shows that “fire or lightning” is covered, so that’s a good start.

On page 6, we find the exceptions to the coverage, so if any of these apply, the policy will not pay out. Nothing in Part A would seem to apply to a DIY LED project, unless you tell me your LEDs are radioactive. Part B also doesn’t apply, unless you’re somehow perpetuating a fraud using LEDs.

Part C reads like it could apply, because it mentions “construction”, “design, workmanship or specification”, and “maintenance”, but this section only applies to the dwelling and so refers to those things which are permanently affixed to the house. That would include things like ceiling fans and light fixtures, but wouldn’t include stuff that is attached to the walls using thumbtacks or 3M Command strips. It even says that:

However, we do cover any resulting loss to property described in Coverage A - Dwelling and Dwelling Extension not excluded or excepted in this policy.

This clause basically means the exceptions on Page 6 should be interpreted narrowly, not broadly.

The point is, in the entire policy, there isn’t a clause that requires listed equipment, and remember that this is the most bare bones policy commonly available. If such a requirement did exist, then building your own PC wouldn’t be possible, since the standards bodies do not test individual computer parts – except the PSU, because that plugs into the mains.

If a fire that damages the house does occur, the most probable causes would be due to: 1) an unlisted power supply or power brick feeding the ESP32 or the LEDs, or 2) no current limiting (eg a fuse) to cut out the power supply. Other failures like a shorted LED are unlikely to actually cause a house fire, and the insurance companies and UL know this; they’re more focused on preventing arc-faults that contribute to an estimated 50% of electrical house fires every year.

Good design and clean installation on your part, and using properly listed low-voltage power supplies, will mitigate the major fire risks, leaving just software bugs and lighting snafus for you to deal with.

As a matter of completeness, if there is an unlikely fire, be it from an LED project or from a candle falling over, the insurance company will still pay. But big or small, the claim will be recorded in the CLUE database along with the payout amount. This often reflects negatively on homeowners, so future rate increases may occur. But that varies by state. In any case, though, the insurance policy has still done its job: cover a non-intentional loss.

I would nevertheless advise you to have a look at what sort of homeowner policy your dwelling is covered under. Everything beyond HO-1 is nicer, and some even include limited claim forgiveness of some kind (for a price). Also consider talking to your insurance agent, who should be able to help interpret how the policy applies.

What range do you get, and under what sort of driving conditions? I occasionally want to do the quick jaunt to the light rail station that’s 20 km one-way.

An electric Grom would be adorable and highly approachable. I want one.

Oh man, I posted earlier about looking for a mid-drive Class 3 ebike. My wallet is going to hate me for seeing this article lol