HNHacker News
TopNewBestAskShowJobs

gas9S9zw3P9c

852 karma · joined April 27, 2020

submissionscomments
gas9S9zw3P9c··on Org-roam: Emacs org file based personal knowledge base system
I'm one of the people who started using emacs just for org-mode about a year ago. I've been using it daily to manage my schedule, TODOs and projects. It's nice, but IMO overhyped by a very vocal religious minority. I've been spending as much time configuring little things and working around rough edge cases than it has actually saves me. Elisp is still somewhat of a mystery to me and I'm mostly copying and pasting snippets without truly understanding what's going on.

I've encountered so many little issues that resulted in hours of debugging and searching for solutions that the net productivity gain has definitely been negative. True, if I stick with it for 10+ more years, the productivity gain will likely become net positive, but who knows if that will happen. IMO it's quite likely that better systems are going to be invented over the next 1-2 decades. It's clearly outdated.

Only do it if you are willing to invest a large amount of time into learning emacs, and if you are planning to stick with it for decades. The best for me way to describe it is: Emacs is the ultimate procrastination tool. It's great at making you feel productive by letting you do meta things with elisp and configs, when you really should just be getting stuff done.

gas9S9zw3P9c··on Ask HN: Startup CTO using big Twitter following to harass me, what should I do?
Wow. I actually agree with dhh on this one, but the way he responded to this, freaked out, and made childish accusations and Bitcoin jokes does not speak well for him. This really changed my image of dhh and his companies. This is beyond unprofessional for someone with such a large following.
gas9S9zw3P9c··on A Guide To Hacker News For People Who Aren’t Men (2018)
I'm sorry that this is off-topic, but we can we please ban Medium links? I cannot read the article because of the paywall and I have no interest in spending my time hacking around it.
gas9S9zw3P9c··on BLEU Score: Bilingual Evaluation Understudy
It's not so unusual in my experience. I used to do research in this area and we ran side-by-side human evals with our experiments. Yes, huge jumps in BLEU typically imply better human ratings, but the variance is big. Many times, significant jumps in BLEU had no significant effect on human ratings. ROUGE is even more useless.

You just don't see these things reported in the literature because 1. human evals are a pain to run and can be expensive 2. researchers need to publish or perish.

gas9S9zw3P9c··on Stop Taking Regular Notes; Use a Zettelkasten Instead
> How do you know? Maybe their set of habits are a complicated system which they are just not aware of being a system, thus not they are not talking about as they consider it as natural or just obvious.

You're absolutely right. But with that you are also just arguing about the definition of "system" - everything is a system in one way or another depending on where you draw the line. If I scribble stuff into my notebook and throw it away once it's full, you can call that a system.

By system here I meant a system or product that has a well-known name they can point to, like Zettelkasten or GTD or BASB, etc.

gas9S9zw3P9c··on Stop Taking Regular Notes; Use a Zettelkasten Instead
By using tools for this you will inevitably spend additional time organizing, tagging, and summarizing notes, most of which is probably a waste of time. The article has a pretty good example of this

> I find watching how this knowledge graph grows satisfying

Yep, it's satisfying. Because it's procrastination and gives you the illusion of "being in control" - just like engaging in social media is satisfying. Or cleaning your room while you should be doing homework. I don't think it's useful.

Taking notes is great. Research has shown that active note taking, e.g. while watching a lecture, is a crucial part of learning. Organizing them into some (in case of roam, PAID!) system that you will probably not even come back to? Procrastination.

gas9S9zw3P9c··on Stop Taking Regular Notes; Use a Zettelkasten Instead
This seems to be a new trend now. I'm skeptical. Many of the most successful people I know don't have complicated systems like this - instead of spending time on optimizing note taking, they actually get stuff done. In fact, many don't have systems at all. On the other side, most people raving about the systems are "productivity gurus" writing articles and books, not people who actually get stuff done or would be considered "successful" in the economic sense.

I see a similar trend with org-mode users. It's great and all on a technical level (I've used it myself in the past) but it's easy to fall into the elisp and customization rabbit hole. Over time, the few hours spent here and there to optimize the system add up to thousands of hours that could've been used to get stuff done.

These complex systems are the perfect excuse for procrastination because they make you feel like you are doing something productive while you actually aren't. Just get stuff done.

gas9S9zw3P9c··on GPT-3: Language Models Are Few-Shot Learners
Where do you draw the line between "generating" and "copy-pasting from its vast memory"? Why do you think what humans do is not copy & pasting different snippets of information they have come across in the past? Isn't that what grammar is? A bunch of rules you've come across a lot of times?

Other than the given prompt, the models don't have a goal. So what other than copying and adjusting would they do?

gas9S9zw3P9c··on Ask HN: If you are fully/partially remote, would you hire devs in India?
No. At least not through the normal application process. I have worked with many brilliant Indian people, and some of my favorite colleagues were Indian. However, there is so much "noise" in India's talent pool [1] - bad communication, fake degrees, fake profiles, sloppy work - that it just takes too much effort filtering through it. I understand that it's biased, but simply not hiring from India seems to be the better business decision when there is good talent elsewhere.

Of course, the exception are people that come through the network, not through a "normal" application process.

[1] https://restofworld.org/2020/india-engineering-degree/

gas9S9zw3P9c··on Ask HN: Do you still use MongoDB?
I use postgres for pretty much everything these days, sometimes with Hasura if I need GraphQL.

There was a short period of time (1-2 years or so?) where I did my projects with Mongo, but right now I don't see what it can offer over something extremely stable like postgres, which also handles JSON amazingly well.

gas9S9zw3P9c··on Ask HN: Do we have a metrics store?
There is Prometheus [1]. It has bindings for pretty much all languages and integrates nicely with Grafana [2], which is used to plot Prometheus queries onto nice-looking dashboards. I have been using Prometheus for many projects and couldn't be happier with it. It's incredibly small, fast, reliable, and memory-efficient. On one of my clusters, it has been 2 years without any prometheus downtime.

Most commonly used infra services, e.g. databases like postgres, Minio, Kafka, Docker, or k8s, ship with Prometheus metrics out of the box, so you don't need to setup anything to monitor these. Just enable the metrics endpoint. It also integrates with long-term storage (like timescaledb) for more advanced queries on historical data.

Alternatively, there is InfluxDB. I haven't used it, so I can't speak to the differences. These seem to be the two big ones, I'm sure there is a whole range of smaller "startup-ish" competitors/projects as well. Splunk or other log analysis systems are another option I see commonly used. They are not a metrics store, but many companies seem to "abuse" their logs extract metrics. IMO not a good idea.

[1] https://prometheus.io/

[2] https://grafana.com/

gas9S9zw3P9c··on Ask HN: What programming skills are required by HFT or Trading firms
That's a good point and it makes sense, but couldn't you hedge the inventory risk?
gas9S9zw3P9c··on Ask HN: What programming skills are required by HFT or Trading firms
To me, MM is the activity of providing liquidity with passive buy and sell orders - profiting from the spread while taking on inventory risk. That's also how exchanges define it, paying (or giving discounts to) traders whose orders are filled passively - the whole maker/taker incentive that was invented to attract customers a few decades ago. And we had human market makers well before electronic markets, and we still have them for illiquid markets today. They are (still now) called Market Makers even though they trade on minute scales!

So why is it not considered Market Making if I make decisions on millisecond or second scales? I don't understand how that is related to time horizon. So that's what I meant with original question, why can't I be a profitable market maker without HFT speeds? I still don't see the reason... Or are we just arguing about definitions here? Maybe when you say MM, what you really mean is HFT, but in my head the taxonomy looks completely different.

DMMs are a different story since that's basically just an SLA you have to adhere to, but not talking about DMMs in this case.

gas9S9zw3P9c··on Ask HN: What programming skills are required by HFT or Trading firms
> But if you are trading wide enough where you think latency isn't a factor, aren't you really just predicting where you think the book will go "far" ahead in the future

Yep, or rather, predicting where the market will not go to avoid the price moving against my quote. My impression was that HFT is all about being fast, as opposed to smart, since you can't make complex predictions on nanosecond scales. Complex models don't fit on an FPGA. So couldn't you get an edge by being just a little bit smarter with predictions but slower and quoting wider spreads? And just to be clear, I'm not talking about minutes here, but maybe milliseconds to seconds, which I think wouldn't be considered HFT today?

gas9S9zw3P9c··on Ask HN: What programming skills are required by HFT or Trading firms
Oh yeah, when I said "lower profits" I didn't mean per roundtrip, but overall lower profits over a longer period of time due to fewer trades happening at wider quotes. HFT almost always quotes minimum spreads, so my question was why you couldn't be successful quoting larger spreads (not $1M, but let's say 2-10x the normal spread) while trying to be "smarter" as opposed to faster, since with wider spreads you'd already be in front of the order queue when the price moves, in front of the HFT trader who has to react to the movement to quote his tiny spread.

I'm trying to understand where the market maker = necessarily HFT comes from. After all, we had markets makers well before HFT too. I understand that MM can be particularly profitable when doing HFT, but why is HFT a necessity?

gas9S9zw3P9c··on Ask HN: What programming skills are required by HFT or Trading firms
Hm, maybe I'm fundamentally misunderstanding something. Let's just say the price is a random walk. It's not directly relevant to the argument, but for simplicity. An HFT MM will make money by continuously quoting ask/bid at the best price, i.e. a lot of trades, capturing a very small spread each time. Given that it's a random walk, buy and sell don't always occur simultaneously either, e.g. in a trending regime where the HFT MM may start pulling quotes due to risk checks and/or inventory/hedging concerns.

So if I am an MM that quotes a wider but gets fewer executions at larger time intervals, shouldn't I be able to also make profit? After all, my queue position is in front of the HFT MM because I put in orders earlier (since I am quoting wider), latency here is irrelevant.

Can you explain where my logic is faulty?

gas9S9zw3P9c··on Ask HN: What programming skills are required by HFT or Trading firms
Out of curiosity, why do MMs have to operate on HFT time horizons? Couldn't they operate on a somewhat lower frequencies by quoting wider? Sure, they won't see the same profits, but why is it impossible?
gas9S9zw3P9c··on Show HN: Margin – A lightweight, flexible markup language for structured thought
How do you deal with multiline data? For example, what if I wanted to add a multiline code snippet?
gas9S9zw3P9c··on Ask HN: Why is Bitcoin crashing massively right now?
Exactly. This thread describes perfectly how humans love making up stories in their own mind to get the illusion they "understand" something. There's nothing to understand here, and no single simple reason, except that a large volume of BTC was sold.
gas9S9zw3P9c··on Ask HN: Is Web Search Broken?
For me, yes. These days, pretty much all my searches are prefixed with specific domains, such as `site:news.ycombinator.com`. It's a small whitelist of domains like, reddit, HN, stackoverflow, twitter, wikipedia, etc.

If I don't do this, I usually don't get what I'm looking for. Instead, I'm overwhelmed by low-quality self-branding content on Medium, dev.to, towardsdatascience, Quora, etc. There are too many of these now.

Overall, my search has shifted to largely social sites, especially semi-anonymous ones like HN/reddit. The upvoted content (and comments) I find there tends to be miles better than anything Google surfaces on their first page. This has definitely changed over the recent years, it used to not be like this.

gas9S9zw3P9c··on Show HN: Axiom – No-code Browser Automation
I had the exact same thought. I would suggest to take out the .ai because it IMO makes you look disingenuous. If your name is already clear marketing BS, then someone may extrapolate that to the rest of your product.

That happened to me, you already lost credibility with me by naming it .ai because, in my mind, you either 1. Have no idea what AI means, or 2. You know very well but you are lying.

I'm not saying either of those are true, but that's what was going through my mind when I first opened the website.

gas9S9zw3P9c··on Ask HN: Is Rust worth it compared to Golang?
I used to write everything in Golang, but recently ported many things to Rust, so I feel like I have a good idea here. I'd consider myself quite proficient in both. I see Go and Rust as almost opposites.

Go is "easy" and makes you immediately productive. Like, productive on the first day. It's a great language for that reason. Compilers don't complain, no complicated generics, easy concurrency, loose interfaces, GC, etc. But this comes at a cost - it's easy to write sloppy code. Because of its flexibility, my code often suffered from concurrency bugs (don't get me started on deadlock debugging), bad abstractions, or performance issues. I felt like I was productive, but I often had to re-factor my codebases multiple times because of these issues.

Rust is the opposite. I didn't feel immediate productive, and the fight with the compiler was tough at times. But most of the time, the compiler was right and I was wrong. This typically resulted in "good" code once it compiled and I rarely had to go back and refactor my code. And of course there were no pesky concurrency bugs due to the compiler checks.

These days, I would choose Rust over go for my own projects. If I consider the time spent on debugging bugs in Go, I consider myself more productive in Rust, even if the initial code takes longer to write. However, if I was leading an engineering team, I would consider Go instead. It's much easier to pick up, and it's easier to read someone else's Go code. Reading someone else's Rust code can be hard.

gas9S9zw3P9c··on Ask HN: Best Resources to Learn Japanese?
First, you should be clear what your goal is. "Learning Japanese" is ambiguous. For example, I know many people who passed the JLPT N1 exam, but cannot have a simple conversation with a native and have a horrible accent nobody can understand. It's clear they've just "hacked" the exam by memorizing stuff. And that's totally fine if that's the goal to e.g. get a job or a visa. And I know people who are absolutely fluent in conversation, but would never pass an N3 exam and cannot read simple articles. My point is that these are very different skills that need a different approach. What do you want? What is your end goal?

Then, everyone learns differently, so there are no generic "best resources". I don't know if this works for you, but I can tell you what worked for me. For each of the step below, use Anki [0] or any other SRS system to memorize things.

1. Start with a formal textbook like Elementary Japanese (that's what I used) or Genki. This will give you a curriculum for the basics and you don't need to worry about what to learn next. Simply following these textbooks will bring you to a lower intermediate level (JLPT N4 to N3).

2. Do something like "Remember the Kanji" [1] to memorize more characters and get a feeling for them. There are many variations on this you can Google. You can do this in parallel with step 1 too. Learning Kanji in isolation is IMO overrated and not super useful, but just having the familiarity will pay dividends later.

3. This IMO is the tough part. You are now ~lower-intermediate level. You are past beginner books, but probably not good enough to read real-world stuff. For me, it was a mix of resources here. Find people to practice with, listen to podcasts like JapanesePod, read more advanced textbooks, read some online websites you are interested in, etc. I don't have a good answer on how to get past this level. Also, if your goal is to speak conversational JP you really need to talk to people here because so far you've only read textbooks.

4. Once you get to something like ~N2 Level, just immerse yourself while you keep adding stuff to Anki. Do you like videogames (like me)? Just play them in Japanese. Like movies/anime? Watch in Japanese. Tech news? Read in Japanese with browser plugins. Really, just do whatever you enjoy, but do it in Japanese. It will be hard at first, but you'll get used to it.

5. If your goal is to pass an exam (like N1), get books specific to that exam.

[0] https://apps.ankiweb.net/

[1] https://kanji.koohii.com/

gas9S9zw3P9c··on Amateur programmer to paid professional in 60 days. What lang/stack?
Probably Javascript/Node and a bunch of frontend frameworks like React/Vue/etc. You'll have lots of competition and pay is on average lower than for more esoteric technologies, but the market for that is huge. Almost every company needs frontend devs.
gas9S9zw3P9c··on Ask HN: What's your biggest pain point in your life?
Not knowing what I want to do with my life. I'm in my 30s now, have hopped jobs a lot of times, and have some savings, but I feel more lost than ever.
gas9S9zw3P9c··on Ask HN: What's that one thing you would do, if you got to rewind your career?
This is a great question. I spent the last ~10 years hopping around jobs, working at both FAANG and small startups, but I never stayed anywhere more than 1-2 years. I also moved around a lot. My biggest mistake was focusing too much on salary, job title, technologies and self-development, and not enough on forming deep relationships with others. That was true even in college where I got good grades, but I rarely went to lectures, study groups or parties.

Now I'm here with a relatively broad network but few deep connections. I don't have anyone I can ask for career advice, have beers with, chat with about work, or discuss for new opportunities. Looking at my resume you'd think I'm "successful" - but it's the opposite. It feels kind of lonely and at times I'm sad.

No matter what you do in life, in the end it's all about people. Don't waste your opportunities to get to know your fellow coworkers/classmates/bosses. Make some friends.

Also, don't burn your bridges. There were times in my career when I sacrificed relationships so that I could move on. For example, leaving a co-founder behind when I no longer believed in the company. Looking back, I was kind of an asshole and too immature to realize it. I wasn't capable to fully empathize with the people around me. I truly regret that now.

gas9S9zw3P9c··on Ask HN: What are the most enjoyable language/framework you build projects with?
How is that different from JS, which you say is your favorite? Sure, you can have Typescript, but you can type your Python code as well and use something like mypy. I would understand disliking Python if you said your favorite was something compiled like Go or Rust, but it's surprising to me you dislike Python so much while still liking JS.
gas9S9zw3P9c··on Ask HN: How do you manage self-study?
I think you are describing "how to learn effectively, manage time, and stay motivated [once you've picked a topic]" - which is important and useful, but IMO slightly different from what the OP is describing (and a problem I have), which is "how to pick what to learn next when there are 1,000 interesting choices with uncertain payoffs"

For example, you say "Learn one thing at a time" - Sure, but how do you pick that one thing when there are 1,000 things on your list that all seem equally useful and interesting? What I am looking for is a proper system for picking that one thing, not using my gut feeling.

gas9S9zw3P9c··on Ask HN: How do you manage self-study?
That's an interesting approach, but I feel like you are describing a slightly different problem than the op is talking about, or at least I have.

Once I pick something to work on I have no problem staying motivated or sticking with it. I've also never felt bad about myself running into roadblocks. My problem is that there are too many interesting things to work on, I have a hard time picking from them, leading to FOMA for all the other things I can't pick. What I'm looking for is not so much a way to stick with a task, but rather a system to help me "manage" my life, deciding which of 1,000 potential thing has the highest payoff, and how to balance them.

gas9S9zw3P9c··on Ask HN: How do you manage self-study?
I have exact same problem. But for me it's not just about learning. It extends to other aspects of my life - side projects, health, work, travel, etc. I have huge lists of things I want to do, but no real way to manage them. I'm having an especially hard time balancing them.

I've been through various apps and approaches, and even tried to write some tools myself. So far I have been unsuccessful in finding the right abstractions and solution to solve the problem. Everything I've tried ran into edge cases it couldn't handle. I always came back to listing goals and daily schedules in raw text files, sometimes using org-mode.

One approach that has worked okay-ish for me is to have a hierarchy of personal OKRs. Quarterly -> Monthly -> Weekly. I found anything longer than quarterly to not be very useful - life changes too quickly. Even quarterly may be too much. You create these as-you-go, e.g. each Sunday you review your past week and create OKRs for the next week, possibly adjusting some of your monthly goals. Each day is then managed with a simple TODO list and you count tasks towards your weekly OKRs. At the end of each period (day, week, month) you have a review.

This approach still has a lot of shortcomings (not being flexible enough, not incorporating habits, some things are difficult to measure and can't be expressed as OKRs, etc) and I've tried several other things I could talk about, but the time period in which I used this approach was one of the more productive ones.

Regardless of the technique, one thing I've come to realize is that people tend to spend not enough time on "meta" - figuring out what to spend time on. If you think about it, spending 1-2 full days a month making sure that you are working on the right things aligned with your long-term goals is reasonable, but very few people spend this amount of time (me included). Instead, we tend to keep ourselves busy with the micro - tasks right in front of us.

← PreviousPage 3 of 4Next →