So awesome. I've been watching the Data Haskell initiatives with a lot of interest, ever since it started. Haskell is such a great language that it deserves to have a good data science ecosystem.
If you want to make a keyboard more efficient, QWERTY is a very poor choice. QWERTY was designed in the nineteenth century, for mechanical typewriters. It was designed to be inefficient, so that the typewriters wouldn't jam up. As a result, we have such features like: placing the E key such that you have to move your left hand to reach it. E is the most frequent letter of the alphabet for the English language, and putting it anywhere but the home row is nonsensical. What is on the home row instead? A semicolon.
What you've done here, amazingly, is made QWERTY twice as inefficient, by requiring two touches to register certain keys.
To summarize:
- QWERTY: Bad
- QWERTY Mini: Worse
- QWERTY Mini Pro: Still worse, but now costs you money
As with the "Zen of Python" or countless other "Zen of X" treatises, the connection to Zen or to Buddhism is tenuous if it exists at all. As far as I can tell, this author believes "Zen" to mean a collection of short, syllogistic aphorisms. Tim Peters is not a "zen master," like this author says, and has even said about the "Zen of Python": "If I were to change anything, I'd drop the reference to 'Zen.' That wasn't part of the original, and was added by someone else."
Using the word Zen to mean something else—simplicity, calm, a collection of aphorisms, or whatever—only seems reasonable if you've never studied Zen and have no more than a surface-level understanding of what it's about. Could you imagine any other religion being used to describe some aspect of programming? "The Islam of Python." "The Baptist Christianity of AI." These sound ridiculous. It only seems like a reasonable thing to do with Zen because Zen is being caricatured here.
I'm upset that there are 700+ courses across 70+ languages, but none is my language, Esperanto, the most successful international auxiliary language of all time, with an estimated 2M-5M speakers worldwide. I'm unable to select Esperanto as the language I want to learn, nor as the language I already speak. This is the kind of thing Esperantists have come to expect of language learning resources which support a low number of languages, but when there's an app which supports 70+ languages, and Esperanto isn't one of them, it feels like a deliberate omission.
There are tons of great reasons to learn Esperanto, but one is its propaedeutic value: learning Esperanto for a year, and then French for a year, leaves you knowing more French than if you had studied only French for two years. Some cite the ease of learning it (I learned it to a conversational level in about two weeks), and some the intuitive grammar (only 16 grammar rules, no exceptions). Whatever the case, the value of learning Esperanto is huge, even if you have no plans to become active in the Esperanto-speaking community.
It's great that Lairner allows one to learn endangered languages and minority languages—I'm learning Irish, for instance, which only has ~200K native speakers, most of whom are in Ireland. Learning Irish (and other Celtic languages) is immensely mind-expanding, and it unlocks a rich literature and culture. But the original literature and culture of Esperanto is even more vast, rich, and international, and is not restricted to any geographic area.
Another issue I want to bring up is the unnecessary conflation of language with geographic region. In Lairner, languages are given flags of nations, as if there's a 1:1 map of language to country. That is a really bad assumption. Languages go beyond the borders of nations. Just think of Basque, for example. Or Esperanto. (And Esperanto has a flag!)
A Turkish speaker can't learn Basque, by the way, using Lairner. Basque is not a language which is supported at all, as far as I can tell, and Turkish doesn't seem to come up as a possible L1 language for most languages. The site says "700+ courses. Every combination." But there is not every combination. To learn Irish, for instance, you have to already know English or German. There's no Chinese -> Irish, or Japanese -> Irish.
If it were me, I'd feature Esperanto as the language everyone should learn first. Then I'd ensure that, using Esperanto, you can learn any other language. That way you wouldn't have to write lessons for every language pair, only X -> Esperanto and Esperanto -> X.
It's incredible what knowledge we'd have, if it weren't for Christianity and the Dark Ages it engendered. There are tons of palimpsests like this, like the Archimedes Palimpsest, in which the beginnings of calculus was invented, almost two millenia before Newton, but were scraped off to make yet another Bible. Imagine what the West could have accomplished if monks weren't so busy erasing science and math.
Without deep OS support for this kind of thing, I can't imagine it will feel anything like a native experience. There are already much better touch screen experiences with years of development behind them, in the Linux, Windows, Android, and iOS platforms. There's no need to have a janky hacked together version for macOS to try to make it to the same thing.
It strikes me as a little unimaginative to want to improve on smartphone software keyboards but (1) stick with QWERTY, a layout designed to be inefficient, and (2) require multiple taps to enter some letters. It seems like you've invented a way to make smartphone typing even more of a pain than it already is.
I've used dozens of notetaking tools over the years. Some cloud-based, some markdown-based, some flashy apps, some plain-text, some open-source, and some closed-source. My takeaway from years of jumping between them is this: don't use closed-source notetaking software. Just don't do it. Even if your data is in markdown files, on your own computer, you're still probably stuck with proprietary markdown extensions, and at the very least, you're stuck with muscle memory for the app's UI that you'd have to translate to some new system eventually. Startup companies come and go, on a monthly basis. Developers move on to shiny new projects. You can't take that risk, or any other security risks, with your personal notes.
> ... many current and former professional educators have taken it upon themselves to do their own research. Some then wrote books collecting their findings. These books tend to be insta-dismissed as the ravings of crackpot loonies given their alarmist titles and appearances. But they are typically very well written, well researched, and full of primary sources for their claims. So many sources, in fact, that they make for good jumping-off points for more in-depth research.
This sounds like the frustrated research log of someone lacking research skills. An undergrad-level annotated bibliography from someone interested enough to gather sources, but undiscerning enough to tell the good ones from the bad.
Those of us that use Nix don't have the problems you complain about. Just write (or vibe-code) a flake.nix for each project, and you can have arbitrary versions of whatever language (Node, PHP, whatever), arbitrary environment variables, port setups, and whatever else. Add in Direnv and you don't have to do anything but `cd` into your project directory. Then simply add an extra output to create a Docker image for your project.
The idea of paying double digits annually for a service that's basically a weaker version of Nix, and which lacks Linux support, is absurd.
It might not be a new area of inquiry, but the insight seems reasonably novel. If you read the article, it's about how Symbolic Show of Strength (SSS) beliefs are predictors for belief in misinformation. That speaks to something more granular about the psychological mechanisms behind the spread of misinformation. "Repeating known falsehoods is about declaring tribal allegiance" instead speaks to the existence of the phenomenon, more generally.
While folks were building improved ls, cat, and so on, and jq for manipulating JSON data, Nushell has been happily doing all this in a consistent way and making it easier, to boot. I'm surprised to see Nushell missing from this list.
I saw that this was on GitHub and I totally thought this was going to lead to a repo of some sort which solves this problem. But no. It was just complaining.
> it's absolutely undeniably that the Buddhist cannon is full of batshit insane stuff, just like any other religion
Buddhism is not like Christianity, where the source of truth is a book or a canon, and that the book must be believed in order to subscribe to the belief system. Speaking of Zen, at least, it's one of the foundational tenets that it's "a separate transmission, outside the scriptures." In fact, there's a lot written about how Zen isn't a religion at all, at least not in the Western sense, with beliefs, faith, and doctrines. You don't need to believe anything to be a Zen Buddhist. So even if the "Buddhist cannon" has "batshit insane stuff," who cares? Shakyamuni was a great teacher, but that doesn't mean that he can't be wrong.
> I think it's more honest to say that you practice meditation with Buddhist characteristics than to say that you're a real Buddhist if you don't have the time of day for spirits and deities
You might be under the impression that Buddhism is somehow theistic or dualistic. But the Buddha, for one, outright rejects mind/body dualism, which therefore rejects the possibility of spirits and deities. Some traditions, like Tibetan Buddhism, have tantric practices like deity yoga, which involve visualizing deity-like figures, but even then, there's no presumption that these deities actually exist, in some kind of spirit realm. But even if there were Buddhists who believe in "spirits and dieties," again, who cares? It's not like you have to believe anything to study and practice Buddhism.
My main point is that, if you're writing about meditation, or meditative practices, that either originated with Buddhism or were developed through Buddhism, it's fairly disingenuous to completely divorce it from its context.
So let me get this straight: HTML was originally built to have markup and style both in the same containers, and it was clunky and hard to maintain, so CSS was invented, to abstract away the styles so that content and style could be separate. And that worked pretty well. And then someone came along and said: you know what would be great? Removing all the useful abstractions and going back to the HTML 1.0 way of putting it all in one container. And rewriting HTML in JavaScript to make it less maintainable. And so on. Am I getting this right? Correct me if I'm wrong.