Damon Cortesi

Last week I dove into Hudi, Iceberg, and Delta Lake. Take a peek into my brain as I see how they work under the hood. dacort.dev/posts/modern-d…

Logs back in after a year...wth is all this crap?

@biblebeltman Me too!

@dacort I miss TweetStats. Thanks for the memories!

@rattus Until next year... dacort.xyz

@rattus Son of a.

Last week I dove into Hudi, Iceberg, and Delta Lake. Take a peek into my brain as I see how they work under the hood. dacort.dev/posts/modern-d…

With @bluesky opening up, curious how many orgs will adopt the DNS verification approach. If only there were some stats about this... 🤔

Not the O(log n) I came for, but the log scale I stayed for.

The cool thing is I can just force close the app and keep using it. 🎉

That’s because I haven’t updated it since November of 2022. 😜

Since TweetDeck is back as X Pro I just wonder if @dacort will bring back some of the old twitter analytics for old time sake! LOL ;)

@MichaelDrogalis For long-running (>15m) definitely. Wasn't sure how long the "beam down" would last.

@dacort Hmm, you tell me! I heard ECS is better for somewhat more long-running tasks. True?

This experiment is starting to get really fun: What if, for any public infrastructure tutorial, you could curl a URL and beam down live data into your instance?

@MichaelDrogalis Do you even need ECS or could you just run the container/jar on Lambda?

Lambda + ECS is shaping up to be the right combo. - It's nearly zero complexity for users - It lets the infrastructure provider control costs

No, seriously, what is happening? Why can't I see anything?

@cortesi Not just me then, eh? (re: code seeming slow)

Visual Studio Code has recently started having odd, irritating performance issues, so this week I decided to go back to my roots and try neovim. Took a while to configure, but I'm incredibly happy now it's set up, and there's literally nothing I'm missing from Code.

💥 🍿

@aaronshunter 👋 🏔️

I made it to Seattle!

@chadfowler @github @adamloving Were you or somebody else at PSL working on that?

Who has experimented with using LLMs and other ML tools to read @github issues, understand the code, and post pull requests to resolve/implement those issues? I'm imagining a manual process for now of tackling tasks in a large open source codebase using nothing but AI as a constraint and documenting the process. Surely I'm not the first one to think of trying this experiment, so I'm looking for pointers to who has alraedy done it.

@TimHaines Wow, that's some inefficient code compared to what's needed today. lol github.com/dacort/dm-whac…

@dacort No, but also very not surprised.

So uh, Twitter doesn’t have any mechanism to bulk delete DMs?

@gwenshap @screenstudio Good to know...been feeling jealous of the folks that can use screen studio.

PSA: @screenstudio is excellent for short screen recordings of ~5 minutes or less. But it is unusable for longer recordings. I made the mistake of trying to use it with a 25-minute recording, and after a few hours of fighting with it, I gave up and re-did everything with ScreenFlow. It's more manual, but at least usable.

@Mause_me Yep, can repro. Opened github.com/duckdb/duckdb/…

@Mause_me Will double check this weekend in a less-hurried state. :)

Alright, frens. It's Friday and I've got some spare time at THE OFFICE so I'm going to see if I can use the @duckdb Spark API to whip up a local example of my favorite "extreme_weather.py" script. Let's go.