Really excited to get OTLP and PromQL support into CloudWatch Metrics. This is just the latest step in a long journey, but an important one that opens a lot of doors to the future.
aws.amazon.com/blogs/mt/int...
Hopefully cardinality won't be the same pricing pitfall it has been in the past.
Once you know that it can happen it will happen!
Big fan of TYDYE, TRUKN, GRTFL, and DEDHD around SFO
Really going to town with this one. 😁
Begging to be forked.
This is how we ended up with 17 USC 117(a) allowing the making of copies of a computer program so that the program can actually be executed and so that it can be backed up. I don't think the law has really caught up with how copies of works need to be created to do things not protected by 106.
Copyright law was largely written without an understanding of how computers work. To the extent that "looking at things ... when you're a computer" requires making a copy of those things then it necessarily involves exercising one of the exclusive rights enumerated in 17 USC 106.
@was.tl was wondering about commissioning something recently.
It's that selection and arrangement that is copyrightable, not simply the words themselves. It's also why a model trained on every word in the English language wouldn't make every subsequent written work a derivative work. The copyright covers the expression, not necessarily its constituent parts.
brb getting ralph tattooed
Just go straight to the main course? 😜
But when you go online do you log in or log on?
Talk to me if you have an interesting profiling use case. I'm doing some explorations :)
Code-level Telemetry Instrumentation: from "Oh Hell No" to "Worth It": whitneylee.com/2025/10/30/codelevel-tel...
Hello, friends! After 10(!) years of #AdventOfCode, I've made some changes to preserve my sanity: there will be 12 days of puzzles each December (still starting Dec 1) and there is no longer a global leaderboard. Read more:
adventofcode.com/2025/about#f...
adventofcode.com/2025/about#f...
So many bad habits and outmoded beliefs to stomp out. Thankfully I'm not feeling like Cassandra very much these days.
Some love from my wife for #AdventOfCode and @was.tl (cc @aneurysm9.com): elly.town/d/blog/2025-...
I think that's one of the things that makes Midnight such a fantastic Doctor Who episode. Also the growing sense of dread at being locked inside that box with Something that might be even worse than what's outside...
I grew up there and can confidently state it is a wonderful and beautiful place to visit but I'd never want to live there again.
Reminder: code review is not "to catch mistakes", it is a tool for understanding to limit bus/lottery factor first and foremost. When code is written (if you're lucky) by one human, having the second human abdicate their judgment to the machine is actively harmful to your team's code understanding.
Even better is when it says "let me simplify your failing test" and proceeds to rip out the test entirely. I guess "no [test], no problem" is a philosophy...
But I thought French wasn't real!
Secret Level
Say Nothing
Disclaimer
Q CLI is great. Sadly there are ~37 other Q products that are poorly differentiated and a bad experience with one might sour people in the whole "suite".
When I've been successful in using LLMs to generate prototypes to convince a team to take, or at least explore, a new path that opens up opportunities within that team for the senior folks to work with junior folks to get things done. Shit doesn't have to be the only thing that flows downhill.
Now I'm the most junior person on my team and the rules of the game have changed. I'm no longer guiding individuals but teams and organizations and I'll take all the help I can get in doing so.
I absolutely do when I can. I was able to do a lot of this in my last position and I hope to be able to do more of it in the future. Some grew and left the team, others are now running with the product I helped conceive and launch.
LLMs that can shit out prototypes that help me achieve those goals _are_ the new tools that I get to experiment with. I still get to be creative with what I ask it to do and how I refine the results until they're what I need. I find that more enjoyable than trying to program humans with prose.
