nikhilsimha

One invaluable hack I found is to use ad block to block the recommendation section of landing page.

That way I don’t habit jump into a YouTube marathon but still search for specific topics when I need to.

I do that with hacker news too.

I am not a fan of case insensitivity either. But it must be easy to write a case insensitive grep. I wager that a nim grep would already exist.

I do agree that this is a red flag decision.

This gives me hope for webdev. Long time infra guy who dreads any front end work with the other frameworks.

That logic didn’t apply to tulip bulbs

> the best coder I ever hired taught herself to code while getting off heroin in a Glasgow slum

Those words bust the biggest bias I realized I had/have - against addicts. Thank you very much!

I used to lead teams that owned message bus, a stream processing framework and a distributed scheduler (like k8s) at Facebook.

The oncall was brutal. At some point I thought I should work on something else, perhaps even switch careers entirely. However this also forced us to separate user issues and system issues accurately. That’s only possible because we are a platform team. Since then I regained my love for distributed systems.

Another thing is, we had to cut down on the complexity - reduce number of services that talked to each other to a bare minimum. Weigh features for their impact vs. their complexity. And regularly rewrite stuff to reduce complexity.

Now Facebook being Facebook, valued speed and complexity over stability and simplicity. Specially when it comes to career growth discussions. So it’s hard to build good infra in the company.

Won't the functional variant read better than the SQL dialect

  movie_reviews
   .filter(_.actor.name.lowercase() == "Zendaya")
   .groupBy(_.title, _.credit_order, avg(_.ratings))
   .sortBy(_.credit_order)
   .take(5)

vs.

  select
    Movie {
      title,
      rating := math::mean(.ratings.score)
      actors: {
        name
      } order by @credits_order
        limit 5,
    }
  filter
    "Zendaya" in .actors.name

You are right. 4.63 million eps. My bad.

Impressive, but I don't get why folks don't normalize it to events per second instead.

It is 463k eps for anyone wondering.

I read the last sentence as "Just my 2°C". I blame it on the over use of 2°C in all these reports.

I contend that you are not a denier. If you are convinced that there is an abnormal trend, caused by human activity, but are not too concerned with the trend - you are simply not an alarmist. I too, wouldn't be concerned if cherry blossoms missed a season.

The problem IMHO is, the climate science community doesn't connect the dots between warming -> unstable weather -> widespread upheaval.

The implicit assumption you are making is that the demand increases in lock step with productivity gains. 100x faster drug discovery, 100x more drugs need to be discovered => same number of people employed.

These correlations do hold for technical fields, but logically there should be a point beyond which productivity gains outpace, demand growth / demand could even stop growing. One should either retool to solve a newer problem before this point is reached, or hope that the point is not reached in the span of their career.

Oil rig builders for example - manufacturing has been increasingly automated, but the demand for oil rig building has grown consistently. But they should probably look into solving other problems given that demand is shifting.

[prioritization] Effective engineer - lau

[systems] Designing data intensive applications - kleppman

[programming] SICP - sussman & abelson

Last one is an old scheme book. No other book (that I read) can even hold a candle to this one, in terms of actually developing my thought process around abstraction & composition of ideas in code. Things that library authors often need to deal with.

For example in react - what are the right concepts to that are powerful enough to represent a dynamic website & how should they compose together.

While we are at it, why not declare national emergency, and implement a one child policy in the high per-capita carbon emitters?

Like Australia and US?

Somalia - 0.1 Metric ton of CO2 per person per year

US - 15 Metric ton of CO2 PPPY

This blew my mind. Mostly because I haven't come across this during 8 years of c++ use for professional/competitive programming.

Always love it when I see a maintainer offer clarifications in a HN comment section! <3

What are the reasons for the perf improvements we see here?

Observed the same thing on Hindu idols in Telangana as a child. My dad had a pretty simple guess/answer.

The nose is simply the most common point of impact when you strike a face from up-front(human or idol). A face is probably likely to be the most common target when raiding.

Since we(comments and PG) are all peddling anecdotes..

IME, the dynamics in Indian schools, are polar opposite to the ones described here. "Nerds" tend to do well in both academics and extracurriculars. And in the majority case can also be referred to simultaneously as "Jocks". The separation between "nerds" and "jocks" into two ends of a (lunch table) spectrum is the biggest culture shock to me. The words that separate "nerds" and "jocks" don't exist in my native language's vocabulary even.

For the longest time, I believed that intelligence is a broadly applicable trait, that translates well to rapid spatial reasoning(sports), abstract reasoning(academics) and emotional awareness(social). I haven't seen even a single exception to the rule while going through the Indian school system.

While there is some truth to it. IMHO, it has to do with cultural ideals that "adults" adhere to. Academic achievement isn't celebrated culturally in US (relatively speaking).

I say that while also simultaneously acknowledging that the current Indian definition of academic achievement is disastrously narrow and breeds a culture of arrogance and intellectual insecurity.

Windowed aggregation is a common problem in feature engineering. We use segment trees to solve it. https://youtu.be/0HttRa2cXig

Not your ideal train system, but happens in India. Quite often.

In both examples 2 and 3, the author reads the same stream twice independently and assumes that a join is not synchronized between the transformed streams. This seems like a fundamental flaw in their offering.

Pushing in a timestamp along with the max/variance change stream[1]. And then using the timestamp to synchronize the join[2] would naturally produce a consistent output stream.

I quoted flink because they have the best docs around. But it should be possible in most streaming systems. Disclaimer, I used to work for the fb streaming group and have collaborated with the flink team very briefly.

[1] https://ci.apache.org/projects/flink/flink-docs-stable/dev/t...

[2] https://ci.apache.org/projects/flink/flink-docs-release-1.11...

Nightly to seconds is definitely great. Seconds to milliseconds is what is questionable I think.

Just to clarify slapping a queue on a database is still a bad idea IMO

Recent work from Berkeley on AnnaDB and Martin Kleppman’s OLEP stuff is a complete opposite of this idea. Something along the lines of - databases are just a queue topology with synchronization.

Queues do definitely seem like a more fundamental primitive.

Some stats from https://www.surveysystem.com/sscalc.htm Plugged in 800,000 as pop size, and 99% as confidence level and 4 as confidence interval. The number it spits out is 1039 It is a pretty good sample size.

Used to work at fb in an infra team.

Their abstraction for job scheduling (Tupperware) is about 5 years behind - something like Borg or EC2/EMR. Something like that is a fundamental reason why fb can’t do cloud as it is right now.

Plus, the infra teams do operate like product - impact at all costs. Which to most management translates to short-term impact over technical quality.

It would be a true 180 in terms of eng culture if they could pull off a cloud platform. An example is how they bought Parse and killed it, while firebase at google is doing extremely well.

Having said all that, I think focusing on impact over technical quality is probably the right business decision for what they were trying to do at the time - drive engagement and revenue.

> we aren't immune against flu, is there any reason to think we could develop (long-term) immunity to COVID-19?

Spanish flu became less lethal in subsequent generations[1]. Helps with the survivability of the virus if it doesn't kill the host. So, the strains with less lethality win out.

We don't need to be immune for the situation to resolve. Simply co-evolving with the virus is sufficient

[1] https://demystifyingscience.com/blog/2020/3/4/patriarch-of-p...

How does this compare to jsonnet or hocon? Why would someone choose this over generating configuration files with a “scripting” language like python?

^ This is why I love hackernews!

“India and China need it the most” - not true at all sir. Per capita usage is key.

https://en.m.wikipedia.org/wiki/List_of_countries_by_carbon_...