Well, it is the end of a fun little era.
Appreciate you @bluebottleroast, and Josh, V, and the other fine members of the crew of the South Park BB who I got to know because I was there (almost) every day.
Had some incredible conversations with people working on sandboxes for companies large and small, Cloud Infra, investors, Consumer Founders, users of related domains, and a few random run-ins that still were all amazing.
And now, for a long break from that specific SF city block.
However, if you'd like to understand roughly how cloud computing will change for the next encroaching market category, and want a sneak peek of the new stuff being worked on in virtualization, hypervisors, isolation and conf computing, DM me here, or ping me wherever, and I'd love to chat.
When we're back in SF in a few weeks, we will find a new place to hold open conversational hours around low-level infra like this, and as usual, everyone is welcome.
@bilskirnir3124@thsottiaux Linux was created as a rejection of robust software architecture and tools. It's a monolithic kernel with no internal isolation written as memory unsafe code. A typo in an obscure driver for unused functionality can easily create a fully privileged code execution vulnerability.
Designed a 3 months agentic memory internship. Launched a company this year focused on agentic backends as we as memory.
I have to say that a company needs lots of skills that are not necessarily technical.
Finally finished reading the Postgres 14 Internals book. Some general takeaways after reading this long, technical book:
1. Postgres is wonderful but not perfect. I started this book thinking that Postgres was the holy grail of databases (at least from everything I'd heard), but some of its architectural decisions probably wouldn't be made today if it were built from scratch. Things like transaction ID wraparound, its implementation of MVCC, and the process-per-connection model.
The book specifically does a great job explaining how these things work and why they can be problematic. Understanding these limitations also gave me more appreciation for projects like OrioleDB and PgRust that are reimagining Postgres
2. Reading groups are very very helpful. I probably wouldn't have finished this book on my own. And even if I had, I don't think I would've learned nearly as much if I hadn't been able to talk about each chapter with other people
3. Everything I learned has been useful in some way. There were definitely moments when I wondered how much of what I was reading would actually be helpful, especially some of the implementation details. But I've already found myself connecting things I learned in this book to other papers I've read and conversations that I've had. That being said, there were some chapters that I found a little boring and could've skipped in hindsight
In need of some book recs now!
@matthewmorek We are [nearly] SOC2 compliant. We've been working for almost half a year with consultants and auditors to approve this workflow and we're really, really close. Established, legacy auditors too (not sketchy startups).
Iβve been saying this. Leetcode was hated, and it will die sooner rather than later. What will come after is no standard interview, more subjectivity, and no definite preparation advice
Eng hiring will turn into the type of hiring PMs, designers etc know all too well!