a lot of learning is rabbit holes. you hit a concept you don’t know, disappear for 10 mins, then try to come back without losing the thread. if @OpenAI had a finder-style workspace with conversations side by side and attention switching between them, it would 10x productivity.
@emirbeganovic What a story, Emir. From BHLD CDs in Sarajevo to Head of Infrastructure at Omarchy. Congrats! 🎉 Isorun's "computers as a resource" idea resonates deeply.
We just launched DigitalOcean MicroVMs on the same belief, and we'd love your take: https://t.co/oyZS4MElJz
Introducing DigitalOcean MicroVMs: isolated VMs that pause when idle & resume instantly. AI agents work in bursts. Stop wasting cycles building custom virtualization layers for fast startup & isolation.
We handle the infra, so you can build!
https://t.co/S0VcpoetYL
Fascinating HN comment from someone who's company built a custom distributed data warehouse using compressed SQLite DB files in S3 that were queried using Lambda functions orchestrated by PostgreSQL running a custom foreign data wrapper https://t.co/gvRviN8D9R
Yesterday at @Retailic, we transitioned our REST API running: Ring, Next.jdbc, Pathom3 from regular servers to AWS Lambda on ARM64 Graviton 2 processor by adding a 110LOC of code to adapt the Ring request/response model to AWS Lambda model. 🧵
#serverless#AWS#Oracle#Clojure
Productionization new languages at Google is a lot of work. As a person who witnessed two attempts, this thread is what I learned. Google has an incubation policy for new languages. You start by being experimental, then become incubator and then graduate to a canonical language.
Papers published by Facebook on traffic engineering over the years are really phenomenal.
- Kraken, 2016 - https://t.co/dOjDI4PMVT
- Maelstrom, 2018 https://t.co/BOWMUdvTMk
- Taiji, 2019 https://t.co/tgfU2IY2CN
- Zero downtime deploys, 2020 https://t.co/6NzFj90WL5
@chetanp Post deployment Sima aunty be like “Everything in life is compromise.”😂😂 There’s a Sima aunty version for every phase of the product lifecycle..
Here's an amazing mental model I learned recently for building scalable systems:
You can never be confident of reducing the probability of mean time to failure (MTTF) at a big scale (esp AWS), but you can put all your weight behind reducing the mean time to recovery(MTTR)!