just met an engineering team that fired their manager after installing a slack bot that asks "so how we looking on the project guys" every 124 minutes
Ted Nyman
2,384 posts
Ted Nyman
@tnm
computers at openai. prev ceo at cased, cto github, eng banksimple. github.com/tnm. "sf’s guardian angel"-ken jennings
- Fun story about perceptions of scale. GitHub ran VERY lean on infra for a long time. In 2013, visited Baidu HQ in Beijing, and met with an infra team. They asked “so how many servers do you run?” We said “about 150?” & they said “oh cool 150,000. Decent.” No no lol. 150!
- I would love one fucking demo of "computer use" that doesn't involve booking travel or a hotel or a restaurant at some mid place in Hayes Valley.Perplexity is reimagining hotel discovery — more like having a personal concierge. You can ask detailed, complex queries, and get tailored hotel recommendations in seconds. What once took over 30 minutes and required juggling multiple tabs and sites now happens with a single
00:00 - anthropic putting out http status codes no one has even seen before
- you can also go home, spend time with your parents while they are still young, make out with your ex girlfriend, enjoy the warmth of summer, and just be 20. we need to stop with this forced fomo, this un-needed angst, this stealing of youth; which is ultimately a manipulation.summer is 1/4 of the year you can spend it at an internship with the best FAANG or you can take a break & relax while the best time to build something life changing slips away 10 startups (amazon, stripe, airbnb) we’re all started in the summer by college kids & drop outs who
00:00 - story time. redis has a command called KEYS, which goes through every key. it’s a debug command, only to be run against an offline copy of your database. one day many years ago, a rails-lad at GitHub was on a machine trying to do some analysis via the ruby redis client..Redis started timing out randomly. 1-2 times per hour. Operation that should take 2ms was taking 5+ seconds. What we checked: - Memory: 40% used (plenty available) - CPU: 12% average - Network: normal - No slow queries in logs - Replication lag: 0ms Enabled Redis slow log
- Truth is 90% of backend (web) perf engineering is adding indexes and fixing N+1s. Should be app dev 101 but it's not. If you want to stand-out as an engineer at a fast-growing company: just find these and fix them.> I am completely blown away by how amazing this is and how dramatically it impacted performance. Creating a simple index dropped database queries from ~8s to ~20ms on a production DB which containing around 20mil rows. PostgreSQL index only scans FTW. stackoverflow.com/questions/1323…
- Extremely excited to announce kit, Cased's open source code intelligence toolkit for building AI devtools. kit shines with precise, relevant context to LLMs. Code graphs, tree-sitter symbol extraction, file ops, simple & semantic search, chunking, summarization, & more. ⤵️
- Replying to @dangugel~5 big MySQL servers, one primary. Handful of Redis servers. Web was ~30 servers running unicorn. The rest were git file servers, & then we had CI machines, etc. Search was on AWS.
- GitHub wanted more of our distributed team to move to SF, we spent 3M+ on an office buildout. Not only did it not encourage people to move; locals didn’t even come in. I used to describe it as “empty as mattress store,” which it was. Build your business.I spent $1.5M building our office after raising a seed round. My co-founder thought I was crazy. Here's what changed his mind... 𝐓𝐡𝐞 𝐂𝐨𝐧𝐭𝐞𝐱𝐭: After our seed round, I looked at our team. Mostly immigrants. Working 6-day weeks. Building something incredibly hard. The
00:00 - Replying to @BenScottStevenseven just hearing “pac-10” nowadays makes me sad
