jtc331

A book, TV show, movie, video game, album, or comic book is not available on the internet served by the copyright holder’s own servers with no authentication or authorization checks. But the NYT is available in that way.

Maybe they don’t know all the intracies, but they certainly know the difference, that’s why there’s such a cultural big deal about blue versus green bubbles.

Everybody knows when one person in a text group isn’t using iMessage, and they usually hate it.

IME gas pumps are often slower than the 10 Gpm limit in the US.

That doesn’t follow. Meeting the emissions standards is possible on its own, but not necessarily along with other requirements like efficiency.

SELECT FOR UPDATE means two queries can’t _read_ the same row at the same time, right?

Those can be in a deadly spiral.

Risk aversion doesn’t actually mean you’re good at addressing risk; in fact it may keep you from implementing changes (change is risky) that address real risks.

That’s not what the parent said. The wired connection itself is better, but it has the tradeoff of physical wear and tear and ports that break.

You’re absolutely correct.

It’s just irrelevant since you don’t need anything even approaching hyperinflation to have real world impact on people, and the real wage loss over the past several years is clearly in that category.

The problem with this definition is that even CEOs at big tech giants don't own the company (in the vast majority of cases).

You might want to check your facts. The 2022 Honda Odyssey minivan (not a small vehicle) has 0-60 of 6.5s.

DC fast chargers are already 400V or 800V, so we’ve long since crossed the limits of AC charging at home.

My understanding (I used to say the same thing) is that this is now understood to be false, and that placebos don’t affect objective indicators, only self reported perception, and that perception reverts to the mean over time with the placebo.

You must have incredibly good insulation if 2kW is enough to heat that much space.

> It's weird to say, but we should crash the economy if it can slow down climate change.

To save humanity we should destroy humanity.

No thanks.

And even if in different data centers (say different AZs), query latency is probably 1ms; in practice that’s quite low. How many queries are you doing in a request?

If deploying on different racks is that big of problem for your latency then you’re severely limited in how far you can scale your application just by physical rack space. IME this just isn’t true.

It all depends on your workload. Those 500 instances may be make calls to other services, services, for example, or burning CPU in various ways.

Each of those reads may only be 1ms, which would translate into only 100 connections needed.

So the workload you describe — without more detail — may in fact be just fine with that connection count.

But plenty of people also run Postgres with well above 500 max connections.

PG doesn’t have a thread per connection; it has a process per connection.

It’s not true. And using the 70s as proof is either dishonest or lacking knowledge: gold rose in the 70s because an artificial pinning of its price was removed.

There’s a lot of emphasis here about just adding storage, but that’s missing the fact that unnecessary indexing affects writes and reads performance also (and potentially quite significantly).

Using a partial index where it’s obviously matches the use case (like when a majority of values are null) is just correct modeling and should not be considered a premature optimization or waste or developer time.

I think the bigger reason is simply that backwards compatibility is hard.

Some projects put in that work (e.g., PostgreSQL), but it comes at a significant tax on productivity that could otherwise be put into new features.

> The goal, in the name of the company bottom dollar, is to make it as inconvenient as possible for me to use an employer offered benefit. A truly mundane and recurrent example, unfortunately, as I try to navigate U.S. health insurance.

Never attribute to malice that which can be adequately explained by incompetence [and misaligned incentives].

Designing a good experience and flow for a piece of software is hard. You have to work for it not to be bad rather than work to make it bad.

Nuclear and renewables/batteries actually have the same biggest problem: the vast majority of the cost of future energy is up front capital costs. That’s what makes natural gas so attractive: the vast majority of costs are fuel amortized over the lifetime of power generation.

That’s the thing we need to fix regardless of the power source.

Liquid cooled charging cables is pretty standard for high speed DC fast chargers already. When that cooling is broken at an Electrify America charger, for example, charging is limited to something around 50kW.

I believe that is what pre-COVID masking in hospital settings has shown, yes. Possibly COVID is different from the flu.

But the difficulty for proving masking helps is: 1. If it only takes ~20 virions to get infected, then masking has to have a basically perfect filtering rate and be worn perfectly 100% of the time, neither of which is obvious. 2. With an endemic virus even if masks slow down spread they aren't going to reduce your likelihood of being infected overall.

4% in Italy isn't just "when hospitals were overwhelmed", it's with an aging population.

I think the “hack” language is correct because it’s a rethinking of GitHub at some level.

More importantly though I think you’re missing something very big about open source projects — the transition from “small project with single maintainer” to “group ownership” has always been a difficult one; that’s not unique to GitHub.

I don’t like the idea, but CPUs have long been an example (fused at the factory to disable cores, for example).