They just need to sell it already in a whiskey flask for treatment on the go
17.01.2025 12:38
👍 1
🔁 0
💬 0
📌 0
They just need to sell it already in a whiskey flask for treatment on the go
There are two kinds of people in the world:
1. Those that measure out their Robitussin the right way
2. People like me
Tableau was so so good. I loved the dev interface so much.
I learned tableau way before learning SQL and I’m really glad I did. It was much easier that way.
The closest thing I’ve found to that is the experience from Sigma computing, but they don’t offer a single user tier.
#databs
This guy’s living in 9999
THE BEST
Raycast is my favorite tool of all time. You gave me my original invite 🙏
I think they’ve got the best wrapped around. I lost half my history this year moving to a databricks laptop but still so fun to see.
Spotify was my top app for sure haha.
Nothing like a "Weekly Huddle (Mandatory)" to set the right tone on a Monday 👌
This is not my team, thank goodness hahaha.
What?? Snowflake iceberg is parquet in s3
What are the two sides
I knew I had to be missing something. Thank you, Danny.
Even the example records weren’t making sense the way I read the problem
I bet the min isn’t needed as well. The days difference will be null for it and that can be turned into a true in the case statement assuming no null dates in the input.
Then you do a case statement:
- If date is the min date, true
- if days diff is greater than 90 (or any arbitrary cooldown), also true
- otherwise false
What if you add a window function computing the days of diff between the current and previous entry sorted by date.
Also grab the min date while you’re at it. This is more useful if you need to run this for multiple partitions instead of for just the entire table.
Then…
I’ve been thinking about this.
I think an annoying computer science approach can do this without recursion.
I think it’ll run nlogn as well.
I need some paper
This comes back to the core missing piece in duck imo. No catalog support.
That’s the beauty of it! The answer is always sql hahaha
I think you can run duckdb in python. I’ve installed it for a demo but I’m not sure if that is what you mean.
Trash. Offensive
SQL
Bigquery and databricks sql both have it on lock too 💪
Wake me up when @duckdb.org supports iceberg catalogs.
Big SQL doesn’t want us to have this.
#databs
Start:
I don’t know how we can make this valuable without a semantic layer
Fixed:
I don’t know how we can make this valuable with a semantic layer
Fixed more:
I don’t know how we can make this valuable
Final fix:
I don’t know
#databs it’s all pointless. Just disable sso and lose your login
Let’s smoke about it. King sized
1 - Poetry just because it works. I don’t package and distribute Python libraries so I have pretty basic package management needs. No opinion on uv either way.
2 - Nothing fancy like that, I just wanted a minimal prompt. It was just trickier than I expected haha
I’m a poetry guy, but this would work so nicely there too. Thank you for sharing.
My zshrc is filled with my nonsense language of aliases and a surprisingly long chunk to make my terminal prompt just be a battery emoji 🔋
Gotta be at least 100 bucks a month
They just don’t get it
It fits!!!!!! This rules. 10/10
