Ellie Huxtable
@ellie.wtf
2 months ago Riding into the desert today, shooting some Ektar 100 - very excited to see how it turns out 🥰
💬 0
♻️ 0
❤️ 18
Riding into the desert today, shooting some Ektar 100 - very excited to see how it turns out 🥰
Mapping SwiftPM traits to Xcode for Tuist highlighted the growing gap between them. Traits = cleaner SWIFT_ACTIVE_COMPILATION_CONDITIONS + conditional deps. But Xcode gets neither convenience. These inconsistencies pile up and hurt the ecosystem. hit the ‘relearning auld lang syne with classtab.org and worrying that nobody will show up to the party’ phase of nye Steal this alert sanity check:
• errors up + traffic flat = regression
• latency up + traffic up = load
• errors up + traffic up = scaling or dependency
• errors flat + latency up = slow backend
Boring rules. Fewer fire drills.
docs.netlify.com/manage/monit... Thanks for the feedback, Remy. Totally agree: the bandwidth CSV is super useful, and having a similar export for web requests would make spikes way easier to trace. Shared this with the team.
fuckingapproachableconcu... made it to HackerNews and people are loving it! 🎉
Also just added SKILLS.md so you can include it as a resource for your coding agents:
https://github.com/pepic... It's never been easier to move a MySQL database to PlanetScale using our import tool.
We've added better observability, per-table copy selection, and other advanced migration features.
www.youtube.com/watch?v=N1HC... Which one do you want more of: observability, security, or “how it all fits together”?
3️⃣ Create → deploy → run:
How the newer pieces fit together into an “AI that ships” workflow.
www.netlify.com/blog/create-... 2️⃣ AI-generated app security checklist:
A quick sanity-check before you ship “vibe code” to real users.
www.netlify.com/guides/secur... 1️⃣ Observability guide:
Validate deploys + debug prod faster (what to check first, where to look next).
www.netlify.com/guides/runni... ICYMI: 3 Netlify bookmarks if you’re shipping fast (and want fewer “why is prod on fire” moments). 👇🧵
Get to know more about the GitButler MCP server!
blog.gitbutler.com/using-gb-mcp If you want a dead-simple Observability demo:
“Show me the top erroring routes right now.”
Click the route. Trace the request. Tie it to a deploy. Ship the fix.
Find it. Understand it. Fix it. 👇
www.netlify.com/guides/runni... From an empty folder to a full Prisma 7 app running on Prisma Postgres in 5 minutes.
You can follow the same steps here 📹👇
pris.ly/y/p7-quicks... The difference between panic and progress:
Before: “Something’s wrong. Anyone know what?”
After: “Errors spiked on /checkout right after deploy 1429.”
Observability is how you get answers without a 47-message Slack thread.
www.netlify.com/platform/obs... we've actually priced in all the externalities of car ownership but instead of having car owners pay for asthma treatment and environmental cleanup, it's a money transfer from idiots to local cartels also we've learned that one cannot use capitalism to limit consumption of cars
Liquid impressed me. Forgiving by design: parser accepts anything without errors, undefined vars render empty, user templates can't run insecure code.
I'm porting it to Zig. Goal: Liquid without Ruby, including Wasm. More themable software with that safety contract.
github.com/pepicrft/liquidz This is a great suggestion! We have a bunch of improvements for observability on our plate, and this is definitely one of them. How would you want it to work?
i think the zod equivalent might be like using codecs a lot more zod.dev/codecs - which already has some useful ones, but could be expanded a lot parsing with rich types starts to overlap with unmarshalling, which can be treacherous but maybe going halfway - rich types but strict schemas - is utopia? blog.rubygems.org/2017/10/09/u... ✨That’s the highlight reel. ✨ It’s all about tightening the loop from idea → live → fixed. What do you want to see Netlify ship next for agent workflows in 2026?