I think there are different aspects to it.

Amazon’s delivery service is better than ever. You get products in half the time, with less packaging, and fewer miles traveled to deliver it to you, without any significant increase in delivery fees.

Price is still competitive when you take into account delivery cost and speed. If you don’t care about those, Amazon isn’t the cheapest.

Search and reviews are down the tubes. It’s like Amazon no longer cares if their site is overrun with crap products as long as people are buying them.

Amazon still works great if you only buy name-brand products that are fulfilled by Amazon.

So wouldn’t the fees be proportional to the price? The added taxes on a tiny cheap holiday home would be cheap too.

It explains the answer is 4 before the 5 minute mark.

Part of the reason is because it goes into the story of the SAT being wrong and a student being the one to catch it, which I found interesting.

After that it mathematically proves it several different ways and then shows how it relates to some real problems in astronomy.

Can you elaborate on what happened when you tried to search? I’ve never had trouble.

Those are all protocols for accessing an entire calendar or sharing your whole calendar, not for general-purpose inviting one user to one event.

I’m talking about using the ChatGPT API to make a chat bot. Even when the user’s input is just one sentence, it can cause ChatGPT to forget its prompt.

Is it possible to be a productive programmer with slow typing speed? Yes. I have met some.

But…can fast typing speed be an advantage for most people? Yes!

Like you said, once you come up with an idea it can be a huge advantage to be able to type out that idea quickly to try it out before your mind wanders.

But also, I use typing for so many others things: writing Slack messages and emails. Writing responses to bug tickets. Writing new tickets. Documentation. Search queries.

The faster I type, the faster I can do those things. Also, the more I’m incentivized to do it. It’s no big deal to file a big report for something I discovered along the way because I can type it up in 30 seconds. Someone else who’s slow at typing might not bother because it’d take too long.

GPT-3.5 seems to have a problem of recency bias. With long enough input it can forget its prompt or be convinced by new arguments.

GPT-4 is not immune though better.

I’ve had some luck with a post-prompt. Put the user’s input, then follow up with a final sentence reminding the model of the prompt and desired output format.

Also, did you fully cream the butter and sugar before adding any other ingredients?

If you just dump everything into the bowl and then mix, this is what happens

Did you scrape the bowl while mixing?

KitchenAid mixers are great, but depending on what you’re mixing you need to scrape the sides of the bowl with a spatula and then mix some more.

I don’t think it’s over mixed, I think the cookies made from the batter that was stuck to the sides are under mixed.

Some people say there’s no malware for macOS and that’s obviously not true.

But others say macOS has malware so it’s no better than Windows in that regard, but I don’t think that’s true either.

Look at this example. It only works if it tricks users into downloading and running an unsigned executable, bypassing sometimes multiple warnings.

I agree with you, but politics is complicated. If she felt like continuing to fight for nuclear at that time would be unpopular, it might not have been worth it. It probably would have made it impossible to achieve other goals.

I’m a fan of randomizing the test order. That helps catch ordering issues early.

Also, it’s usually valuable to have E2E tests all be as completely independent as possible so it’s impossible for one to affect another. Have each one spin up the whole system, even though it takes longer. Use more parallelism, use dozens of VMs each running a fraction of the tests rather than trying to get the sequential time down.

I think the reality is that there are lots of different levels of tests, we just don’t have names for all of them.

Even unit tests have levels. You have unit tests for a single function or method in isolation, then you have unit tests for a whole class that might set up quite a bit more mocks and test the class’s contract with the rest of the system.

Then there are tests for a whole module, that might test multiple classes working together, while mocking out the rest of the system.

A step up from that might be unit tests that use fakes instead of mocks. You might have a fake in-memory database, for example. That enables you to test a class or module at a higher level and ensure it can solve more complex problems and leave the database in the state you expect it in the end.

A step up from that might be integration tests between modules, but all things you control.

Up from that might be integration tests or end-to-end tests that include third-party components like databases, libraries, etc. or tests that bring up a real GUI on the desktop - but where you still try to eliminate variables that are out of your control like sending requests to the external network, testing top-level window focus, etc.

Then at the opposite extreme you have end-to-end tests that really do interact with components you don’t have 100% control over. That might mean calling a third-party API, so the test fails if the third-party has downtime. It might mean opening a GUI on the desktop and automating it with the mouse, which might fail if the desktop OS pops up a dialog over top of your app. Those last types of tests can still be very important and useful, but they’re never going to be 100% reliable.

I think the solution is to have a smaller number of those tests with external dependencies, don’t block the build on them, and look at statistics. Sound an alarm when a test fails multiple times in a row, but not for every failure.

Most of the other types of tests can be written in a way to drive flakiness down to almost zero. It’s not easy, but it can be doable. It requires a heavy investment in test infrastructure.

Bulk mayo makes sense if you’re a restaurant or cafeteria or running a summer camp or something like that. Probably not for many other people.

Check out Linear. The startup I was at nearly switched to Jira and then thankfully when a bunch of us protested, we tried Linear and ended up really loving it.

I think there’s a huge range of stuff that’s legal but that many people might want to filter out because it’s gross or disgusting.

The tag “NSFL” may have been created for the most egregious pictures (that might be illegal in some cases), but it was generally applied to a much wider range of stuff in practice.

I’m not sure what point you’re trying to make.

Chromium and Android really are open-source. There are hundreds of products like Electron and Fire OS built on top of them without any involvement or consent from Google.

Just because Google Chrome and Pixel phones have some proprietary code doesn’t mean that Android and Chromium aren’t open.

Actually, just changing the file name doesn’t change the format. If it works, it’s because whatever place you uploaded to already supported webp.

If you download a webp file and you really want jpg, you need to actually convert it, not just slap .jpg on the end.

It’s also just an open file format. Anyone could implement it, and in fact I found dozens of completely independent implementations of webp decoders on GitHub in various languages.

There really is no secret ulterior motive in this case.