Hacker News new | past | comments | ask | show | jobs | submit | rtodea's comments login

We love the things that we are good at.

If you start with easy math problems (but not labelling them as "math problems", but as logic, smart puzzles) you can build up from there.

Consistency and frugality of options is important.


We love the things that we feel good at. Even if we are objectively better than most peers playing an instrument, if the teacher (parent) gives the impression that we are not good all motivation might dwindle.

On the other hand if the approach of the author gives a slight competitive edge, this might snowball for the reason you've given.


> I checked his work, and its mostly wrong

Can you show some examples that make you draw this conclusion? Thanks!


Sure, but I will limit myself to one, which is fairly exemplary.

The question is to show that R2 is second countable. This means that it contains a countable base for the topology. The usual way to do this is to pick open balls with rational centers and rational radii, and use a little finesse to find one around any point in an arbitrary open ball.

The Author's answer was to take the set of all open balls, and pick only those with natural number radii. This is neither countable nor a base for the (usual) topology on R2. This answer has a check mark on it.


What makes you say that about Python?


it's become a complicated language with many different ways to do the same thing but none of them very good


To add to that, it's introduced as a 'simple' language. But, in 2023, the cumulative effort one puts into getting things "right" is bigger than effort you would put in learning something that's described as high learning curve.

As an anecdote: Someone described elixir to be a "high learning curve" language if I wasn't exposed to Functional Programming. When I inquired how long they mean by that, the answer was 2-4 months. Which really says something about our attention spans.


Popularity dooms the simplicity of almost every language.

Re: Elixir, not surprising. Makes for more opportunities for some though.


Simplicity is for elitists.

Maybe some of you remember Didier Verna's Lisp, Jazz, Aikido: https://www.didierverna.com/blog/index.php?post/2007/04/03/L...

It's not hard to sympathize with Verna's feelings. These are all based on a similar set of philosophical guideposts: maximal output for minimal input, maximal possibilities of self-expression from a minimal set of generative rules, and understanding the deep "essence" of the craft, so that when you add your own contributions, it is by finding the "essence" of the addition and harmonizing it with the essence of existing work. And these are appealing because they give the feeling of tremendous power and the sky being the limit.

But the real world runs on the philosophy of Visual Basic, punk rock, and mixed martial arts, which are all based on a different set of philosophical guideposts: a) focus on practical solutions to real world problems; b) make getting started as easy as possible for everyone; c) it doesn't matter if added components harmonize with the original; what matters is if they contribute significant value on points a) and b), i.e., it's okay to get messy.


I came back to Python after several years, and there's type-checking now, which we're required to use at work. Way to ruin the entire point of Python.

Also, Python's async stuff was always terrible until they introduced the new async/await feature, but I suppose that's part of the "many ways to do the same thing" you mention. They should've done it from the beginning IMO, but it was hard to predict maybe.


I don't have many complaints about the language itself, but I've found the packaging ecosystem to be the most complicated and frankly dysfunctional that I've used in the last decade.


I never understand this complaint when Javascript/Typescript is sitting there with a mess of .lock and .json files across multiple tools that sometimes interoperate and sometimes don't.

Package management isn't a solved problem; Python employs standard patterns for it; Significantly better than chained Makefiles from my C-development days.


I didn't say it was a solved problem - it isn't! - but rather that I find python's solution particularly bad. The situation in C (and C++) is indeed even worse, I just haven't been doing that for the last decade. Things I've used in the last decade that all work better, in my opinion: Bundler, NPM, Maven, Cargo, and even Go's packaging.

I always liked python, but I didn't use it at all really for a very long time until recently, and I've just honestly been surprised by the poor state its packaging seems to be in.


NodeJS has NPM, which is simple. Usually when you want to run someone else's NodeJS project, you `npm install && npm start`. Most Python projects have a Dockerfile, which shows you how bad the actual package management is.

TS is just JS but with extra things that can go wrong (Babel etc) and not being able to use simple `require` syntax anymore and some not-very-automatic type checking that you don't need; I don't use it. It's like taking a steak fresh off the grill and smothering it with mayonnaise.


unless you're Dutch


> Explicit is better than implicit

Says the language whose code-block-endings are invisible! I will never understand the desire of people to make the important punctuation in their language invisible. Python, CoffeeScript, YAML. These people just hate being able to see the important flow-determining punctuation in their language!


Does anyone the source of the movie sequences from the Filmbox presentation page? Thanks!


I would also be interested in this! A reverse image search didn't bring up anything.


I recently migrated to Windows from MacOS and have found the following paid apps really helpful:

1. [Pengwin](https://www.whitewaterfoundry.com/) - a customized Debian distribution with helpful scripts for setting up a dev environment

2. [X410](https://x410.dev/) - a customized X11 Server

They made all the setup really easy. I reuse some of the scripts from Penguin in the Ubuntu distribution.

One caveat for those who want to try WSL2, which is available if you subscribe to Windows Insider program, is the instability.

Some times, after an update, your WSL will not start. After some restarts everything is back to normal.


WSL2 became available outside of insider builds in version 1903 which came out May 2019.

I have not experienced WSL2 a not starting issue at all and I use it every day this year.

Also what benefit does Pengwin have apart from X server? I use the Ubuntu image and I've had no issues.


For me it really helped to copy and paste the commands related to the X server.

For instance I had a small issue with the scale of the applications. In Windows I have 250% (recommended) scaling applied.

``` export QT_SCALE_FACTOR=2.5

export GDK_SCALE=$(echo "(${QT_SCALE_FACTOR} + 0.49) / 1" | bc) #Round ```


Is the instability related to one of the above apps? I've been using WSL2 for at least six months without that issue.


I don't think so.

I blame it on being on the Dev release channel.

To be honest I thought I had to be on the latest and greatest to take benefit of WSL2.


I sometimes get wsl2 to be not responsive too, but you can force shutdown ask (WSL --shutdown) that solves most problems except networking.


> Choosing different wording and expressions for emotions will not change how you feel or how you act.

Cognitive behavioral therapy (CBT) tries to do that.

> CBT focuses on challenging and changing unhelpful cognitive distortions (e.g. thoughts, beliefs, and attitudes) and behaviors, improving emotional regulation, and the development of personal coping strategies that target solving current problems.


I fail to find "wording" in that quote. CBT deals with thoughts, not the exact words used to describe them.


First of all, thank you for the reference to Lysenkoism:

"as any deliberate distortion of scientific facts or theories for purposes that are deemed politically, religiously or socially desirable".

But I fail to see the connection with Dyson here.

Having an open discussion about taboo subjects is at the core of progress.

Conflating science with morality is adding another barrier to having those kinds of discussions.

I am aware of Brandolini's Law:

“The amount of energy necessary to refute bullshit is an order of magnitude bigger than to produce it.”

But if the criticism is done in good faith, with the tools of science, I think it's worth the debunking exercise.

Data is just data. It can be collected with a lot of bias in place and can be interpreted with the same biases. We are only humans after all.

When you say thinks like:

1. "it's just a dude not even bothering to learn research before dismissing it all as wrong" 2. "he stupidly used his social power to undermine the science"

I feel you are some priest that dismisses the sinner, by pointing out the vices of carelessness and stupidness, in the name of the sanctimony of science.


> as any deliberate distortion of scientific facts or theories for purposes that are deemed politically, religiously or socially desirable

This is precisely what Dyson was engaging in. There was no scientific backing for his critiques, he didn't even bother to learn about what he was critiquing. He just went on a political crusade and pretended that because he had scientific accomplishments in a different field, he was applying the same scientific rigor in climate science. In fact, he was not.

As far as being a "priest," excuse me for caring about basic scientific honesty. This sort of hero worship of Dyson, to the point of excusing extremely bad science and abuse of authority by calling criticism religious, is exactly the sort of stuff that happened in the Soviet Union! Lysenko and Dyson promote the right politics, so they are considered above reproach.


Morality isn't a binary thing.

If we adopt consequentialism, how do we quantify all the good things amongst the bad ones?

Please note that just because the bad ones are most vocal, doesn't mean they will outweigh the good ones.


>> Morality isn't a binary thing.

It is when you're young and know everything


Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: