Hacker News: sinkasapa

New comment by sinkasapa in "California is chasing wealth that has feet"

sinkasapa — Thu, 24 Sep 2026 23:21:43 +0000

Has there been a good study that firmly demonstrates that tax rates are the primary motivation that rich people have to live in one place or another?

New comment by sinkasapa in "Companies promote incompetent employees to management tolimit damage they can do"

sinkasapa — Tue, 18 Aug 2026 22:26:00 +0000

A department chair at a university is usually not the most productive researcher, as far as I can tell. And if they are, they tend to leave the position quickly to refocus on research. As far as I've seen it is not unusual for people who are engaged with what they do to avoid admin tasks. A dynamic like that would seem to push less engaged or capable people into admin.

New comment by sinkasapa in "Wired headphone sales are exploding"

sinkasapa — Sat, 14 Mar 2026 12:11:30 +0000

You can get cheap wired earphone splitters and share what you're listening to on long bus rides or flights. I miss that. Maybe one wants something between needing to be on and conversing and the isolation of separate bluetooth listening.

New comment by sinkasapa in "$3T flows through U.S. nonprofits every year"

sinkasapa — Sun, 08 Mar 2026 01:08:32 +0000

Having a detailed and auditable report of how money is being used is really helpful for creating the understanding you are talking about. That is what accounting is for and why it is so essential to modern life.

The site is obviously just an advertisement for a weird camera surveillance system, but the concern about incomplete accounting is very real. In many places one might want to contribute to non-profit efforts, IRS information isn't even available. In my work in Ecuador, I have seen a lot of fraud, and half-baked charities. Some rich NGOs sometimes walk in on some field trip that donors have paid for, make some statements about all they are going to do and disappear without follow-up. Basically they are just tourists on a free vacation taking publicity photos. There is a specific organization that comes down to build environmentally safe toilets. Not only are these built by young middle class volunteers that know nothing about building anything but their CVs, the communities they are helping don't even need new toilets. The building supplies tend to be repurposed after the volunteers are gone, every single year. I'd like to know if I paid for that. There are seeds of merit in the program, but also unnecessary waste.

Despite negative examples, there are many worthy things that are done, and could be done in the region. Northern money can go very far in the areas I work. It can do a lot to not just improve but transform people's lives. So you suggest that an answer to money misuse is to have personal experience with any organization you donate to. How many people who have the money are going to spend any real time in Amazonian Ecuador? They aren't there now. What is going to change? Since there are few people with money who can be personally involved, does that mean that no effort should be made to better people's lives there? Obviously, that is what accounting is for. I think the article is absolutely right about that. I find their solution to be creepy and invasive. Maybe just having better auditing and reporting standards makes more sense than pointing cameras at hospital patients, but what do I know?

New comment by sinkasapa in "Lean 4: How the theorem prover works and why it's the new competitive edge in AI"

sinkasapa — Sat, 21 Feb 2026 19:32:53 +0000

Lean 4 is uses constructive logic. If a closed world assumption requires that a statement that is true is also known to be true, and that any statement that is not known to be true is therefore false, that is not true of constructive systems. I only use Rocq, but I believe the type theories in Rocq and Lean 4 are basically similar variations on the Calculus of Constructions in both cases, though there are important differences. In a constructive theory something is true if a proof can be constructed, but the lack of a proof does not entail that something is false. One needs to prove that something is false. In constructive type theory, one can say, that something is true or false.

New comment by sinkasapa in "Go away Python"

sinkasapa — Tue, 30 Dec 2025 17:39:13 +0000

Those are both valid reasons to use both languages. The "only" (whether true or not) is what the argument hinges on. It is roughly the same as saying that the only advantage of X is that it is popular, but Y is also popular and has additional advantages, therefore, Y is better than X. That is a valid argument, whether the premises are true or not.

New comment by sinkasapa in "Why are 38 percent of Stanford students saying they're disabled?"

sinkasapa — Fri, 05 Dec 2025 04:24:33 +0000

It is interesting how accommodations can reveal dysfunctions in educational practice. In my courses, requests for accommodations generally change the course design for all students. This is because the accommodation does not alter expected learning outcomes, but it is clearly something that aids students with difficulty learning the material succeed. My goal is that all students succeed in the expected learning outcomes. I don't want the course to be challenging for the wrong reasons. So often the request reveals something I was doing that is unnecessary and makes the class more difficult for little reason. That isn't to say that there are not learning environment that should add additional stress. Sometimes such conditioning is needed so that one can succeed in the challenges they are being trained for. That isn't the case for my students.

New comment by sinkasapa in "Mount Proton Drive on Linux using rclone and systemd"

sinkasapa — Mon, 24 Nov 2025 04:08:02 +0000

I want to be happy with proton but their poor linux support across all their products makes it difficult.

New comment by sinkasapa in "MIT study explains why laws are written in an incomprehensible style"

sinkasapa — Tue, 17 Dec 2024 21:13:19 +0000

It seems like the central claim of the study is that there is an increased use of center embeddings in legal texts in comparison to non-legal texts and casuals peach. I'm seeing a lot of debating and presentations of opinions about various aspects of legalese in the comments, but the number of center embeddings in a document is measurable, and is in no way associated with clarity, or precision. It is known that beyond one or two center embeddings, most people find a sentence incomprehensible. There is no way that a sentence like "The dog the cat the rat chased ate saw the man wink." is more clear than some semi-paraphrase, which does not use center embeddings, such as "The rat chased the cat who ate the dog who saw the man wink." It seems reasonable that the only explanation the researchers could find for such behavior is that it seemed to have some sociological value in invoking authority. There is no sense in which center embeddings help to avoid lexical ambiguity. They are simply difficult to parse.

New comment by sinkasapa in "8 months of OCaml after 8 years of Haskell in production (2023)"

sinkasapa — Tue, 03 Dec 2024 06:14:13 +0000

I got turned off of F# because it seemed like knowing C# libraries and tooling was assumed if one wanted to do anything non-trivial, kind of like some of the functional JVM languages always assume some amount of Java knowledge and use of Java tooling. F# seemed nice, but it didn't seem like a real stand-alone language. Unlike Elm or Purescript, where one should also know JavaScript and its tooling, I don't find learning all the C# and Java stuff independently compelling enough to use F#, Scala, Frege, etc.

New comment by sinkasapa in "Turkish language has a gossip tense"

sinkasapa — Sat, 12 Oct 2024 19:03:33 +0000

It is called evidentiality. It doesn't have any direct relationship to time, as one would expect with a tense. Southern Quechua has a reported event evidential affix as well. A friend who is a speaker of Quechua was once horrified, hearing that a man was going to be given a life sentence for murder in Bolivia. They played a portion of his confession over the radio, and the accused man used the reported event evidential through the entirety. Literally, saying that all his words were second hand, dubious information. To my friend, the implication was that he was saying what the police had told him to say. Apparently, those judging the case were not aware of the subtlety, and it did not come through in the Spanish translation of the confession, resulting in a conviction. Whatever the facts of the case were in the end, what is interesting is that for Quechua speakers like my friend, due to the use of the reported event evidential, there was no confession, even though all of the events of a murder were stated in the first person.

New comment by sinkasapa in "On 17th century "cocaine""

sinkasapa — Thu, 10 Oct 2024 07:01:55 +0000

It definitely seems misleading to talk about coca leaf use as cocaine use, given the common expectation for what that means. People reading the newspaper probably don't know all these details, and it isn't spelled out. I can't imagine some guy in a club not feeling like he was robbed after paying for a powdery drug and then receiving a handful of leaves. I can't imagine an Andean woman sitting down for morning tea being pleased to have a bunch of powder dumped in her cup.

New comment by sinkasapa in "On 17th century "cocaine""

sinkasapa — Thu, 10 Oct 2024 06:46:49 +0000

I suppose the better term would be synecdoche, since the similarity in pronunciation isn't coincidental, like "The sun rose in the morning." and "Look at the pretty rose petals." but is a form of motivated polysemy.

New comment by sinkasapa in "A Tour of Hell – Shell scripting Haskell dialect"

sinkasapa — Wed, 02 Oct 2024 18:48:51 +0000

There is no reason why one has to enjoy a haskell-like syntax over Bourne shell syntax but I think you're wrong about the tool. The author states that "It should have no abstraction capabilities (classes, data types, polymorphic functions, etc.)." Granted, there may be types but without those abstractions, I don't think you need to know much of anything about type theory.

New comment by sinkasapa in "WordPress.org bans WP Engine"

sinkasapa — Thu, 26 Sep 2024 20:01:09 +0000

As far as I could tell, they weren't denied use of the code, just a bunch of other services that are not covered by the GPLv2.

New comment by sinkasapa in "Why wordfreq will not be updated"

sinkasapa — Wed, 18 Sep 2024 17:57:19 +0000

Maybe it is "including linguistics" but most of the world's languages don't have the data available to train on. So I think one major question for NLP is exactly the question you posed: "Is there any other paradigm that we think can enable a computer understand language other than training a large deep learning model on a lot of data?"

New comment by sinkasapa in "SAMA – open-source Chat server"

sinkasapa — Fri, 06 Sep 2024 14:57:33 +0000

Toasters have been around for a long time. So has Javascript. I suppose that I can buy a modern toaster or an antique toaster. Still, they both seem unremarkably like toasters to me. Likewise, I can run an old Javascript program or a new one. It feels like being modern isn't really all that special. I wonder if no matter how old the tradition of writing software becomes, we'll always want to associate it with modernness. Will certain software writing ever become artisanal?

New comment by sinkasapa in "The long road to ropeless fishing"

sinkasapa — Sun, 01 Sep 2024 17:34:21 +0000

I was kind of curious about what the equipment is supposed to look like. There aren't a ton of details at this site but one can see some diagrams.

https://www.fisheries.noaa.gov/new-england-mid-atlantic/mari...

New comment by sinkasapa in "Do Art Scenes Lead to Gentrification? (2018)"

sinkasapa — Sat, 17 Aug 2024 02:07:30 +0000

I feel like the art scene is where the artists go to get a drink, meet one another, and then go back to their places to collaborate. It seems like a lot of the most accessible venues are places that are not explicitly arts venues. I've seen a number of cases where artistically inclined bar tenders at dives, who have their friends come in to hang out, thus creating a scene, work to convince the artistically indifferent owners that allowing them to have shows, open mics or something more interesting than a photo of some used up sports star on the walls will draw customers. What begins as a quirky place to sing karaoke and pay bargain prices for beers are where people who build careers often begin communicating and collaborating.

New comment by sinkasapa in "Ask HN: How did you learn Regex?"

sinkasapa — Thu, 08 Aug 2024 20:50:48 +0000

That is a hard question because there are so many ways that one can understand regex. I learned how to read and use them using Unix tools like sed, but I think that my path to starting to understand them probably began with papers like "Regular Expression Matching Can Be Simple And Fast" by Russ Cox (https://swtch.com/~rsc/regexp/regexp1.html), well after feeling like I was pretty good at using them.

Then, as an expert in linguistic morphology, I started learning about things like subregular languages, as talked about in works such as Aural Pattern Recognition Experiments and the Subregular Hierarchy, by Rogers and Pullum (https://www.cs.earlham.edu/~jrogers/JoLLI.pdf). And I continue to wonder what the relationship is between these classes of languages and word formation.