Hacker Newsnew | past | comments | ask | show | jobs | submit | marsavar's commentslogin

People are reacting quite strongly to this answer, but it is unfortunately correct. OP has essentially created an application for memorising vocabulary, which is... fine, and it's an achievement to be celebrated.

But no amount of flashcards will make you a competent language speaker. There is no substitute for immersion.

What made it really click for me for me was reading. Lots and lots of it. My suggestion is to start with short, easy stuff (stories for kids) and then move on to progressively harder material (short newspaper articles, essays).

I passed JLPT N1 back in 2013, and preparing for the test was just an exercise in memorising vocabulary and grammar patterns. What really made the language click for me was reading novels in Japanese. That alone helped me more than any amount of Anki-style JLPT prep material ever did.

Vocabulary is important, but it's much, much easier to absorb and retain if you learn it in context.


Do you know of a tool that can generate texts to read based on exactly your level?

I think that was Krashen’s input hypothesis. If I read a text in Vietnamese with more than one unknown word, it’s too much. Exactly one would do it.

Haven’t seen a tool doing that.


It's a numbers game. Sentence complexity within a given novel follows a distribution, and if you keep reading then you'll keep getting some input that is at exactly the right level for you to grow. It's normal to stumble on the exposition at the start of a chapter and then breeze through the dialogue.

I did find it helpful early on to go through web novels with a low 95% coverage vocabulary count, like the Narou stories indexed here: http://wiki.wareya.moe/Narou

Natively is a great resource too. It does Elo-style ranking of novel difficulty: https://learnnatively.com/browse/jpn/?language=jpn&lvl=

I highly recommend real stories over generated text and synthetic exercises, because the key to success is staying engaged long term. Stories are just more fun. Also get yourself a reading setup that minimises the pain of dictionary lookups, because there are going to be a lot of them. ttsu reader + yomitan is excellent.


That matches my experience, too. I passed JLPT N1—then called 1-kyū—back in 1985 (!).

I did spend a lot of time memorizing vocabulary with flashcards, but I spent even more time on extensive reading—novels, newspapers, magazines, anything I was interested in, even if at first I understood little. The repeated exposure to vocabulary in real-world contexts really made a difference.


My favourite, when I lived in Japan many years ago, was the Musashi-Koganei melody in Tokyo: https://www.youtube.com/watch?v=cT2xTUPveCw

It stood very much in contrast with all the other jingles, and I simply loved it.


Is that a fancy dancy version of Sakura Sakura (probably the most famous Japanese folk song)?

https://m.youtube.com/watch?v=jqpFjsMtCb0


It is.

There are possibly-recognizable tunes throughout the system. Vivaldi's Spring comes to mind. I think at Ooimachi.


OP's personal website deserves its own post! https://dulnan.net/ <3


"The best FREE scripts" got me!


This is absolutely amazing


When did the internet take a wrong turn?


When the nerds got a bunch of money and turned into uncaring business people.


arthur_fist.jpg



Gosh, that brings back memories! I remember having so much fun coding my own trivia bots in mIRC Scripting Language (https://en.wikipedia.org/wiki/MIRC_scripting_language) over 20 years ago.


Same. mIRC scripting was really what motivated me to learn to program. I had tinkered a little bit with BASIC before that, but mIRC was what kept me interested. It's one thing to write uninteresting "hello world" CLI apps, but mIRC scripting was something you and your friends could immediately interact with. I wonder if kids today have something similar...maybe Roblox scripting?


yeah I think the "this is immediately useful for me" part of mIRC scripting was a key driver in pushing kids to learn programming, and it seems lost in modern times.

But, perhaps, we're just not young anymore.


> maybe Roblox scripting

That, but also Discord is basically what people use on computers for communication today, and Discord also has bots.

Of course, you cannot just willy-nilly write your own client for a Discord bot without dealing with tons of hassle, compared to how easy it is to write something that uses the IRC protocol, so not sure it's the same "instant gratification".


This was a lot of fun to play. Congrats!


That's amazing! Could you elaborate more on your VACUUM planning?


Mostly, be careful with long-running transactions (hours long), and modify your autovacuum settings to be as aggressive as possible based on your workload. Be careful with the freezing threshold too.

I ran into

    multixact "members" limit exceeded
Quite a bit when starting out :)


Can you recommend a good rule of thumb for autovacuum settings given a large workload like yours?


Here's mine:

- autovacuum_max_workers to my number of tables (Only do so if you have enough IO capacity and CPU...).

- autovacuum_naptime 10s

- autovacuum_vacuum_cost_delay 1ms

- autovacuum_vacuum_cost_limit 2000

You probably should read https://www.postgresql.org/docs/current/routine-vacuuming.ht... it's pretty well written and easy to parse!


Do you ever need to VACUUM FULL?


It’s too slow at that scale, pg_squeeze works wonders though.

I only “need” that because one of my table requires batch deletion, and I want to reclaim the space. I need to refactor that part. Otherwise nothing like that would be required.


Yes, I was asking because you have mention deletes. Thanks for the answer, cool stuff!


Beautiful visualisation - I just wish I could hover over the grid and see which PIN my cursor is pointing at.


By counting grid points it looks like codes in the form 0[1-9][32-99] are the least common with a few exceptions (like 0990 or 0987).

I suspect this is leading zero bias: a leading zero is not meaningful mathematically and we tend to drop it. The exceptions are dates. The day first block doesn't extend vertically into an unused area but the month first one drops off a cliff around 32 because no month has 32 days.


It's nice but also flawed because you can't easily see if the first two digits are the same. Or the last two digits.


yeah, i too feel limited by our 3 spatial dimensions


It was so frustrating when I was studying mathematics, I feel like just one more dimension would make understanding lots of concepts much easier; for me there is simply not enough points to extrapolate from: 0D is the degenerate case, 1D is trivial, so only two, 2D and 3D, are left to play with. %(


color, fill, texture, shape - that gets you up to 7


doesn't work, those dimensions aren't bijections to the reals


It's up to the designer to deal with any physical limitations.

They could have added grid-lines to reveal this information, for example.


Kind of like GitHub's contribution graph? Which is an awesome little piece of design.


As someone who worked for the world's largest trade book publisher a decade ago, let me tell you that dealing with Amazon is the worst. They squeeze publishers' profit margins to the absolute minimum, and they aggressively force them to accept terrible deals because they have the upper hand.

Amazon has been horrible for the book industry. Please buy your books elsewhere!


Can I take this opportunity to recommend the lovely http://bookshop.org and http://uk.bookshop.org


https://www.hive.co.uk/

An online retailer that allows you to send a cut of the profits to a local independent bookshop of your choosing.

No connection. Just a satisfied customer.


I got my hopes up, but hive says

> Please note: eBooks can only be purchased with a UK issued credit card and all our eBooks (ePub and PDF) are DRM protected.

Bloomsbury OTOH sells DRM-free e-books.


I was about to suggest the same. They sell books, ebooks, music, films.


And while I’m at it:

https://libro.fm/

for audiobooks.


Book industry or just the publishers? Show me a single author who loves publishers


Don't worry, Amazon screws authors directly too when they self-publish, by using the cudgel of Kindle Unlimited to choke possible competition in ebook sales.

https://bsky.app/profile/glynnstewart.com/post/3leu7lzvy622g


KU is amazing.

There's entire genres like litrpg, progression fantasy and cozy fantasy that likely would either not exist or be a fraction of their current size without it.

And authors can make a living, there's plenty in those genres (not to mention romance) who via a combination of patreon + KU + Audible are doing just fine.

I too wish there was someone who could compete with amazon, but the thing is nobody seems to actually even try? I feel like the entire book industry would be quite happy if things had remained stuck in time circa 1990, on their own they would never have invented something like KU.


> I too wish there was someone who could compete with amazon, the thing is nobody seems to actually even try?

Apple was starting to compete, got sued by the US government for it, and decided they didn’t need to bother.


I mean, they did it illegally. Their fault for doing it that way. Kobo is competing with Amazon pretty well.

US market is heavily skewed in Amazon's favor though.


Sure doesn't help that unless you go out of your way to buy a third-party device, there's platform lock-in, which was never an issue with physical publishing.


It does rather feel like the shoe is on the other foot now. Go back a few decades and publishers were the ones rinsing bookshops for all they were worth. Two wrongs don't make a right of course...


This could be summarised as "companies leverage their power" which is unsurprising.

If you were to list occasions where entities acted intentionally against their best interests then that would be more noteworthy!


There's nuance to this. A company can achieve power by giving customers a better experience and in that way insert itself between customer and the industry. Thus wielding power in the interest of the customer. A company can also achieve power by giving producers a better experience and insert themselves between producer and industry.


I think my point is that in the majority of cases companies will do both. i.e. (when run "effectively") they will use all available levers.

If they fail to, it will usually be an oversight than a deliberate strategy.

Of course - some companies push harder, overstep more bounds and neglect the possible negative 2nd order effects more. But assuming there's an obvious lever that says "make more money legally" - the vast number of companies will reach for it.


Is it in Amazon's interest to destroy the publishing industry? Is it in publisher's interest for writing books to not be a viable career?


The real lesson is if you let a person or organisation get into a position where they can squeeze, they will squeeze. They won't even be doing it because they are "evil" because the hedonic treadmill makes everyone feel entitled to more. The problem is systemic. We know our failures but don't do anything about it.


Amazon has been more than just generically horrible - they use blatantly anti-competitive contracts (as if their near-monopsony position wasn't bad enough):

"Amazon fixed online retail prices through contract provisions and policies" that "prevent third-party sellers that offer products on Amazon.com from offering their products at lower prices or on better terms on any other online platform, including their own websites,"

https://arstechnica.com/tech-policy/2021/05/amazon-sued-over...


Frugal is the business term :-)


Except when that happens, a clarification is almost always added at the bottom of the article ("This article was amended on [date]. An earlier version said xxx" or some variation thereof). You're not gonna get a second push notification from an AI summary saying "Oopsies, the previous notification was wrong". Once it's out, it's out, and that sort of damage is difficult to repair.


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

Search: