millipede

It's because of the local model use case. Running AI locally on the phone is a feature (they brag about the processor), but it's constrained by memory and bandwidth.

The models are pretty useful, especially when it comes to asking things you may not want to go over a network.

The things we need to know most, RAM and Memory Bandwidth, are unfortunately missing from the release. Their absence doesn't bode well.

The joy being gone is the big painful one for me. Talking to colleagues and friends who like slopifying everything, they seem to love it. It's a divide between us and them; they never really liked programming in the first place and only considered it a job.

At my $work, people do this frequently, but instead of it just being insensitive, they have their personal Slackbot responding on their behalf. After I realized I wasn't even talking to a person, I've stopped opening their DMs.

Just need to learn how to exit it now.

Whenever I see Substack or Medium URLs, I avoid clicking them. They are so spammy with their call to subscribe, that it's not even worth me visiting them. People recognize quality, and they don't need to be pressured to subscribe. The websites I visit every day never asked me to come back.

> On top of that comes what is owed to oil companies and suppliers, claims stemming from expropriations under Chávez and the outstanding loans from China and Russia.

It seems like China or Russia would expect to be paid not in dollars, but in Oil Wells and Refineries. Why would they accept restructuring of the debt instead of repo'ing the assets?

Saltwater comes in the air. Just being near it corrodes everything. Both stainless steel and bronze are very expensive. Even if things were made of corrosion proof materials, not everything can be, for strength reasons.

Picking a random orientation depends on trigonometric functions. In order for this to calculate, it would be a lot cooler if it didn't depend on transcendental functions.

What other language does it better?

Why not unconditionally trust Bob?

Events like this show that the Internet is pretty heavily centralized. The original DARPA Internet was supposed to be resilient to stuff like this, but it's clear that the old Internet, and the new Internet, are not the same. We as Internet engineers really need to be better here, and design hardware and software to be ready to handle any errors, even unlikely ones like a state actor breaking things.

It's like installing smoke alarms; no one thinks they need them until they do.

Right, I think the lazy thing implies that it would happen post "commit" being returned to the client, but it doesn't need to be. The commit just needs to be wait for "an" fsync call, not its own.

I always wondered why the fsync has to be lazy. It seems like the fsync's can be bundled up together, and the notification messages held for a few millis while the write completes. Similar to TCP corking. There doesn't need to be one fsync per consensus.

Both ints and floats represent real, rational values, but every operation in no way matches math. Associative? No. Commutative? No. Partially Ordered? No. Weakly Ordered? No. Symmetric? No. Reflexive? No. Antisymmetric? No. Nothing.

The only reasonable way to compare rationals is the decimal expansion of the string.

> Ctrl-F "convertible"

> 0 of 0 matches

Some analysis.

Intuitively it doesn't sound like it would work. The program doesn't call listen() or accept() on the socket.

I've been throwing moderately large parties the past 2 years (12-40 people) and the lack of partying is definitely noticeable. Most people don't reciprocate, making it disheartening to keep doing it. I wanted to build friendships out of it, and hopefully get invited to more parties myself, but so far it hasn't happened. It's a decent amount of set up (cleaning, buying food, coordinating), and a lot of clean up after too. The ROI isn't where I want it.

I kind of wonder if people have just forgot what to do after the party is over. I had hoped it would be "that was so fun, we should host one", but instead it just kinda fades away in their minds.

Information is valuable; we just weren't charging for it. AI is just bringing the market for knowledge back into equilibrium.

"AI is Amazing" - People who barely write any code.

There's a great graph showing the wages stagnating compared to GDP growth. It looks like wage's haven't gone up. But, when adding back in employer provided health coverage and other benefits, the graphs align again. It just wasn't in dollars. TFA briefly mentions it but I think it should be front and center.

Type inspection is the flaw of Go's interface system. Try to make a type that delegates to another object, and the type inspection breaks. It's especially noticeable with the net/http types, which would be great to intercept, but then breaks things like Flusher or Hijacker.

Not with HTTP/2

Unbelievable, they actually missed one: 2^(log(22)/log(2)) has all even digits!

Neat, looks like it would work well for values clustered around 0.5, and which hang more closely around 0 and 1.

Hm, cool idea but ran into a glitch. Cutting 3 elongated ovals, end to end and forming a triangle, didn't remove the center piece.

This is exactly right. I've stopped publishing to my blog because I know my words will be slurped up and sausaged out into ChatGPT or Gemini. And, without people writing about technical subjects, or talking about current events, or other thoughtful discourse, Google has no future in Search nor ads.

In the future, maybe OpenAI or Google or Perplexity will pay for people to write about interesting subjects because doing it for free online is no longer rewarding.

Cursors are generally better, but they don't give any hint about how much more is in front of the cursor. I don't think there is a O(log n) algorithm to keep track of what position a given item is in a collection (k'th item out of n), which still keeps the other CRUD operations at O(log n). This would be useful to split a large, dynamically changing result set into smaller parts that can be scanned through using a cursor.

> EVCache

EVCache is a disaster. The code base has no concept of a threading model. The code is almost completely untested* too. I was on call at least 2 time when EVcache blew up on us. I tried root causing it and the code is a rats nest. Avoid!

* https://github.com/Netflix/EVCache