Newsfeed 2023

December 2023

The Big Cloud Exit FAQ by @dhh

This isn’t some archaic knowledge that’s been lost to the ages. We might not know exactly how the pyramids were built, but we do still know how to connect a Linux machine to the internet.

I think many of us are forgetting this.

“buy the baseline, rent the spike”

100% agree.

Constant Propagation in Compiler Design

In simpler words, we can say that if some value is assigned a known constant, than we can simply replace the that value by constant.

건강검진 검사 추천 비추천 by @BenjaminKim_

추천

  • 위내시경(위암)
  • 대장내시경(대장암, 3년~5년 마다)
  • 유방초음파(20대부터 2년에 한 번씩, 30대부터는 1년에 한번)
  • 갑상선초음파(20,30대)
  • 복부 초음파(간암, 췌장암, 신장암, 담낭암)
  • 조영증강 복부 CT(한 번쯤 해보면 좋다)
  • 뇌 MRA(40대 이상. 혈압이 안 좋은 경우만)
  • 관상동맥 CT(나이가 많은 경우)

추천까진 아니고 해도 괜찮은 것

  • 저선량 폐CT(30,40대가 되면 한 번 정도 해보고 괜찮으면 담부턴 몇년 주기로)
  • 흉부 CT

비추천

  • 조영증강을 안 한 복부 CT (얻는 것 없이 방사선만 노출된다)
  • PET-CT (절대 찍지 말것. 방사선 노출이 엄청 높다. 암환자나 찍는거다.)
  • 허리 MRI, 사지 MRI (안 아프면 할 필요 없다. 증상있음 그냥 정형외과 가라)
  • 심장 초음파 (증상 없는데 찍을 필요가 없다)

How Companies Incentivize Layoffs—A Study of Corporate Career Incentives by David Anderson

“My API is very slow, how can I make it better?” by @thegeeknarrator

“Where is your API spending most if its time? Network? Database call?”

인프랩 개발팀의 미션과 가치 by 향로(HyangLo)

배우는 관객이 필요하고 프로덕트 엔지니어는 고객이 필요하다.

완벽하지 않은 전략일지라도 이를 잘 실행하는 조직은 쉽게 지지 않는다.

말도 안되는 방향이라도 나를 믿고 따라갈 수 있을정도로 동료와의 신뢰관계를 쌓는 것이 중요하다.

S3 Express One Zone, Not Quite What I Hoped For by Jack Vanlightly

퇴사 후 나만의 세계 밖으로 나아갈 때 도움이 될 이야기들 | 페블러스 이주행 by EO

  1. Figuring out the first question.
  2. Finding the first answer to that question.
  3. Improving that answer.

I wondered if I was working at 100% of my capacity as a human being, and realized that I probably wasn't.

Graviton 3: First Impressions by Chips and Cheese

November 2023

회사와 팀의 align을 위한 키워드: 반복 & 대화 by Hoon Lee

“회사의 중요한 지표 및 프로젝트는 매주, 자주 반복하여 팀에게 설명하고 또 설명한다. 반복을 멈추지 않는다.”

“반복하다 보면, 그 내용에 대해 질문하는 팀원이 생긴다. 그러면 대화를 한다.”

반복과 대화. 회사와 팀의 align 을 만들어 주는 가장 중요한 2가지 키워드라 생각한다.

Repetition and conversation are the two most important factors in achieving alignment within your company and team.

Hash Tables by Maksim Kita's Blog

October 2023

A Survey and Taxonomy of Latency Compensation Techniques for Network Computer Games

The price of managed cloud services

There is arbitrage everywhere, and there are ways to capitalize on it.

September 2023

You can't make decisions based on a game that doesn't exist by Woojin Kim

Don not let them make decisions based on a game that doesn't exist.

Original in Korean:

없는 게임에 근거해 의사결정을 하게 놔 두면 안됩니다.

Do we need non disruptive maintenance? by Woojin Kim

While, theoretically, non-disruptive updates can be an avenue to optimize your company's revenue, there exist several costs, beyond technical aspects, that might render maintaining the current trajectory a more judicious choice. Even if the execution is technically feasible, the practicality of implementing such updates may not always align with the company's best interests. This is especially true when considering the manpower needed for verifications, the associated time consumption, the varying feasibility across different game genres, and the expenses involved in retraining the staff.

Original in Korean:

무중단 업데이트는 이론적으로 회사의 수익을 극대화 할 수 있는 방법이지만 기술적인 측면을 생략하고서도 여기에 필요한 여러 가지 비용이 있으며 이를 감안하면 현재 체계를 유지하는 쪽이 더 나은 선택일 가능성도 있습니다. 기술적으로 가능하더라도 점검 과정을 수행하는 인력, 점검 시간에 따른 손해, 게임 장르에 따른 무중단 달성 가능 여부, 스탭들에 대한 재교육 비용 등을 고려할 때 무중단이 항상 이익이 아닐 수 있습니다.

68 Bits of Unsolicited Advice by Kevin Kelly

Always demand a deadline. A deadline weeds out the extraneous and the ordinary. It prevents you from trying to make it perfect, so you have to make it different. Different is better.

Why an Engineering Manager Should Not Review Code

  • Technical Lead and Engineering Manager are two roles that partner to make the teamwork. Therefore, they should have clear roles and responsibilities and cooperate in a mutual partnership.
  • Teams hit a size inflection point of ~4 where someone must dedicate their time to communications for the team to work. Otherwise, the team collapses into a black hole of thrash.

Delivering 80% of a project as soon as possible can be more beneficial for your team and your own development than striving for 100% completion by the deadline.

  1. 팀장님을 잘 활용하자,
  2. 100% 를 on-time 에 전달하기 보다는, 80%를 한 템포 빠르게 공유하여 피드백을 빨리 받고 반영하여 완성도를 높이자

Link: https://www.linkedin.com/posts/seunghoon-lee-30935520_%EC%A3%BC%EB%8B%88%EC%96%B4%EB%A1%9C%EC%84%9C-%EC%9D%BC%EC%9D%84-%EC%9E%98%ED%95%98%EA%B8%B0-%EC%9C%84%ED%95%9C-%EB%B0%A9%EB%B2%95-activity-7108709928395055104-WVkp?utm_source=share&utm_medium=member_desktop

Non-blocking sequential processing in Go using infinite (unbounded) buffered channel

Turbo 8 is dropping TypeScript

Most programmers find themselves drawn strongly to typing or not quite early in their career, and then spend the rest of it rationalizing The Correct Choice to themselves and others.

Explaining The Postgres Meme

Join ordering is NP Hard

How I Built Three Billion-dollar CompaniesㅣFrank Slootman, CEO at SnowFlake

OK is terrible.

Two days max. I prefer actually tomorrow morning.

August 2023

print(“lol”) doubled the speed of my Go function

ASCII Pronunciation Rules for Programmers

Getting Friendly With CPU Caches

Link: https://www.ardanlabs.com/blog/2023/07/getting-friendly-with-cpu-caches.html

July 2023

The Most Exciting Phrase in Science Is Not “Eureka!” But “That’s funny …”

Link: https://quoteinvestigator.com/2015/03/02/eureka-funny/

Why Refactor?

We are doing this in order to be able to deliver more functionality more quickly.

Link: https://youtu.be/vqEg37e4Mkw?t=1513

How I built a $1M business on 2 hours a day.

  1. do deep work
  2. consistently
  3. over a long period of time

https://twitter.com/thepatwalls/status/1674432696695783431?s=20

Impossible?

BTW, fyi, I don’t remember when I used ‘impossible’ or ‘unrealistic’ on my commitments. Wording is important because that makes difference. Our job is tough but never be pessimistic.

Link: https://m.facebook.com/story.php?story_fbid=pfbid0fmAVRt7X74NQqd2S91V7sXVXpn2Md6K5qmxZWT8qUUFTeUzpQwTm6Lp4DhPiR6Gfl&id=100001518988037

June 2023

Two tricks to increase your company's stock price by at least 10x

  1. Rename the game to [Your MMO Game name Here] Manager [Release year number Here]

I am strongly convinced that this comment will actually work.

Link: https://blog.woojinkim.org/two-tricks-to-increase-companys-stock-price-by-at-least-10x/

You need to start by setting challenging goals that are more than the sum of their parts.

Link: https://twitter.com/lqez/status/1670975453765701632

All of us, without exception, are replaceable. Therefore, if you use irreplaceability in negotiations, it could lead to trouble.

Link: https://twitter.com/_woojinkim/status/1587981325059129345?ref=blog.woojinkim.org

On Wall Street, many employees viewed the company as a pain, and in big tech, many viewed the company as a family - neither of which is a healthy relationship.

Original full korean: 월가에서는 많은 직원들이 회사를 고통으로 여겼고 빅테크에서는 많은 직원들이 회사를 가족으로 여겼다. 그 어느 쪽도 건강한 회사와 관계는 아니다. 그 중심 밸런스는 본인이 잡아야 한다. 나의 목표와 회사의 목표에 접점이 있으면 서로 최선을 다한다. 회사 목표에 나의 모든 것을 꾸겨 넣을 이유도 없고 회사가 나에게 맞추기를 어리석게 바랄 이유도 없다.

Link: https://twitter.com/eyesofjenny/status/1669639792399880192

Where the beautiful person left off, the unit tests remain.

Link: https://twitter.com/sm_park/status/1668631492917096451

“Why do you study?", “To not love something meaningless.”

Link: https://twitter.com/sungwookim/status/1666264671677190144

I don't think you learn from mistakes, I think you learn from the process of fixing your own mistakes. You get humiliated, you get unexpected help… that's how you learn and grow.

Original Korean: 실수로부터 배운다는 말보다는 직접 수습하는 과정에서 배울 수 있다 생각한다. 그 과정에서 치욕당하고 의외의 도움도 받으며 … 그렇게 배우고 성장한다.

Link: https://twitter.com/lqez/status/1664567682904252416

Most people die at 25 and aren't buried until they're 75”

Link: https://twitter.com/george__mack/status/1659217629469609984

Well, this is NOT a bug, it is a feature. A CEO's job is to focus on profit, indeed.

Link: https://twitter.com/SergioRocks/status/1656282517060681728

May 2023

Essential end-of-month practices to elevate your performance

Link: https://www.youtube.com/watch?v=NY4I82RwKNM

Our Mad Journey of Building a Vector Database in Go - Weaviate at FOSDEM 2023

Memory allocations are the big bottleneck in Go.

avo(Generate x86 Assembly with Go) is good first step to use SIMD(Single Instruction, Multiple Data) in Go.

April 2023

A professional is trying to win any match.

Your competitors won't wait for you to finish studying.

It's admirable to listen and learn, but it's not enough.

Link: https://www.youtube.com/watch?v=GlSvRpmOcAQ

There are hundreds or thousands of reasons why things don't work. It's the job of startup founders and their colleagues to make it work “in spite of” them.

Link: https://mobile.twitter.com/jooddang/status/1635429650636877826

Refined representation of technical debt

By enhancing the visibility of our technical debt, we can potentially create an opportunity to tackle it effectively.

February 2023

Excerpt from 이번엔 성공할 줄 알았지? [스머프 시즌2 ep.06 스타트업이 망하는 이유]

The odds are in your favor. The surviving companies don't even do very well themselves, and the competition is gone.

The Next Grass Is Greener - A Green Labs Retrospective(더 푸릇푸릇해질 다음 여정을 향해 - 그린랩스 회고)

Failed to create a successful product.

January 2023

What does the engineering team do?

Daily Execution that measures business metrics, finds components of them, and improves probabilistically.

Interesting idea from Hoon Park's writing 엔지니어링 팀은 무슨 일을 하는가?.

He also add four action items for the engineer.

  1. You must not do anything that cannot affect your business. At least it can't have a higher priority.
  2. No matter what you do, the results should be measurable.
  3. Every moment we have to think about what can be improved to increase the Y value the fastest and most significantly.
  4. Individual engineers should feel weird about not measuring and ask questions.

2022 센드버드 오픈 오피스 아워 ㅣ글로벌 B2B 스타트업을 만들기 위한 20 Quesetions 를 보고 인상깊었던 답변 내맘대로 정리

Q. 고객사의 커스텀 개발 요청을 어떤 기준으로, 어디까지 수용해야할까요? 제품 이점이 명확하게 강화되고 추가 고객사 확보가 가능하다면 수용하는 편입니다. 피드백 바탕으로 제품 개발하는 것은 중요한 지점인데 팀에서는 SI라고 생각하는 경우가 있어서요. 비슷한 맥락에서 고객의 목소리/요구사항에서 무엇을 반영하고(우선순위화하고), 무엇을 안 할지 정하는 기준은 무엇일까요? 그대로 다 만들 수도 없을 것이고 그렇게 하는게 좋은 제품을 만드는 길은 아닐듯해 어떻게 취사선택할지 고민이 됩니다.

A. 너무 고민하지 말고 해봐. 말은 우아하지만 현실은 그렇게 우아하지 않다. 고민하고 실망하고 하는 과정에서 조직에 직관이 생긴다.

링크: https://youtu.be/CewBydKrHWw?t=2642

Q. 한국 및 미국에서 요금제 구성 및 가격 최적화는 누가(콘트롤 타워) 어떤 방식으로 진행했는지 궁금합니다.

A. 창과 방패의 싸움, 영업을 위해서는 고객에게 할인을 해주어야하고, 그러면 서버는 최적화하고, 최적화하고 나면 할인할 여유가 생기고… 유지하기 힘듬, 실제로 회사 밸류에이션이나 캐시플로우에 많은 영향을 줌. 지지부진한 싸움이긴 하지만 버텨야 함. 화이팅!

링크: https://youtu.be/CewBydKrHWw?t=3700

Q. 미국으로 플립할 것을 전제로 했을 때 국내에서 VC 투자 유치시 고려해야 할 사항들이 무엇인지 궁금합니다. 한국 기업이 글로벌 VC에게서 투자를 받으려면 어떻게 해야 할까요?

A. 미국법인을 만들어 한국법인을 인수하는 것을 고려해라. QSBS

링크: https://youtu.be/CewBydKrHWw?t=3895