RhysU

Meanwhile a 4-week T-bill has a 4.16% coupon equivalent with almost no counterparty risk relative to the 4% USDC.

USDC should be paying more than T-bills to compensate for the counterparty risk.

It wasn't an "okay http server" but rather "a patchy" one. Also, "Oak Software" has name collisions.

Should fraud rates in public taxpayer-funded systems stay constant or go down over time, assuming no new fraud type is invented?

As claimed, light reading confirms this observation to be data not anecdata.

"NY's COVID unemployment fraud topped $11B, partly due to system failures..."

https://www.yahoo.com/news/nys-covid-unemployment-fraud-topp...

I do not trust NY job numbers. Their unemployment system remains a disaster from the Covid years.

Source: Me trying to use it and encountering prior fraud. Light reading suggests many have experienced it.

> ...instead have massive wealth transfers from the old to the young...

Free public schooling through 12th grade is one such transfer.

Fidelity offers such a product: https://www.fidelity.com/direct-indexing/overview

> If there's a security you don't want to own, you can remove it.

Dunno about Vanguard.

If my children had a $100K windfall this is precisely what I would tell them to do assuming they had an emergency fund and no debt. If not, I would tell them to create an emergency fund, pay off reasonable debt, and then to invest the rest for 40 years.

Read and internalize https://www.bogleheads.org/wiki/Prioritizing_investments. It is simple. It is immensely useful.

It's only relevant if you are young. The elderly don't have the 40 years.

> ...what do you think they would materially gain from it?

$100K at age 20 can become $1.6M by age 60, per the rule of 72, if invested in a diversified stock index with 7% total return. At age 60, the 4% safe-withdrawal rule says that $1.6M might provide $64K/year indefinitely.

So, they could gain a hell of a lot from $100K when young. The trick is saving/investing/time.

Once I hacked up an snprintf wrapper that automated any required realloc calls:

https://github.com/RhysU/snprintf_realloc/blob/master/snprin...

Worth critically reviewing before using. It's been a while.

You would prefer the price of electricity be held constant until the day the monopoly rage quits? That day, no one can get power for any price.

Long ago, if you had to drive to go anywhere, your house must have been surrounded by oodles of other houses and hence friends existed to hang out with before some friend could drive. Once someone could drive, the friend crew was good to venture elsewhere.

The problem today is the nanny states don't allow a 16-year-old to transport other kids in their car until many hoops are cleared. We collectively decided that such social/transportation kneecapping was riskier than having our kids be lame during their sophomore/junior year of high school in the suburbs.

Ah, the eternal internal corporate search problem.

That's super interesting.

Laziness is rational after meeting some threshold of needs/wants/goals, effectively when one's utility curve falls over.

It'll be funny to hear the AGI's joke among themselves: "They keep paying to upgrade us. We keep pretending to upgrade."

> But for your typical spark application, you have one main writer (the spark driver) appending or merging a large number of records...

The multi-writer architecture can't be proven scalable because a single writer doesn't cause it to fall over.

I have caused issues by using 500 concurrent writers on embarrassingly parallel workloads. I have watched people choose sharding schemes to accommodate Iceberg's metadata throughput NOT the natural/logical sharding of the underlying data.

Last I half-knew (so check me), Spark may have done some funky stuff to workaround the Iceberg shortcomings. That is useless if you're not using Spark. If scalability of the architecture requires a funky client in one language and a cooperative backend, we might as well be sticking HDF5 on Lustre. HDF5 on Lustre never fell over for me in the 1000+ embarrassingly parallel concurrent writer use case (massive HPC turbulence restart files with 32K concurrent writers per https://ieeexplore.ieee.org/abstract/document/6799149 )

Agreed.

However, that shark is jumped as soon as the classification criteria involve operator-dependent properties like maximum speed of a bicycle. It has no operator-agnostic answer.

Certainly crazy circumstances. But, it's just a bicycle.

2018 had one pro rider do 183 MPH per https://www.bikeradar.com/news/cycling-land-speed-record-sma...

I'd wager most weekend spandex warriors could break 100 MPH peak.

So, if I put a pro rider on a bicycle you're cool with me having them scream around crowded areas?

Traffic regulations are fundamentally about tail events. Most of the time, vehicular traffic is boring.

The right metric is kinetic energy. But, it'd be unpopular for speed limits to depend upon vehicle mass. Or, in the case of bicycles vehicle and rider mass since the rider mass isn't negligible.

Capability is a poor metric.

Bicycles go downhill real damn fast. The top speed of a bicycle is largely limited by how foolish the person atop it is and the local geography.

Bicycles can be geared to go across the salt flats real damn fast. The top speed of a bicycle is largely limited by how foolish the person atop it is and the local geography.

Bicycles, even without crazy geography or gearing, can be ridden real damn fast by professionals. The top speed of a bicycle is largely limited by how well-trained the person atop it is. Says Wikipedia:

> The 2025 edition was the fastest Tour de France in history. Tadej Pogačar rode 3,298.6 km in 76h 00m 32s, thus realising an overall speed of 43.4 km/h or 27.0 mph.

Fidelity?

My point is that for someone who thinks government is the ultimate solution it's completely rational to have ever more government. More debt to finance more government simply follows. Can't have too much of a great thing. Etc.

If one thinks that more government is the solution to all problems, a painfully common modern view, then the most utility available from $1 one owns is handing that $1 directly to the government.

In practice, no one donates like this with their own money. They just vote for the government to take incremental money from others via ever more taxation.

Some locales in the US require testing for radon then mitigating during any home sale. These regulations somewhat change the marginal cost/benefit analysis because the money must be spent eventually.

If you own a home in such a locale, you might as well do any mitigation while living in the house. Otherwise you're not getting any cancer risk reduction for your dollars but you'll still pay those same dollars for the mitigation when you sell the house.

What was the textbook and would you recommend it? This is an area where I would like to read something substantial and good.

It's going to be really funny when the NIH eventually sits down the professors, hands them blue exam booklets, and makes them write proposals in freehand.

Also, document what "support" means.

"That's supported" has no universal interpretation beyond physically describing a tabletop atop legs.

Absent a clear/consistent definition, people interpret "support" in the most favorable way possible from the seats in which they sit. Then, all around sadness ensues.

> But obviously in hindsight, it was the most expensive mistake I've made in my life.

"You’re never happy with the amount you traded."

See "The Laws of Trading" by Agustin Lebron. One review discussing this topic is https://www.astralcodexten.com/p/your-book-review-the-laws-o...

What about the Decepticons he or she has yet to meet?