Is it easier to add functions and polymorphic type system to SQL, or to add relational operations to a functional language?

@pamelajhyde Wordle 1,290 2/6* 🟨🟨🟨🟨🟨 🟩🟩🟩🟩🟩

Uber's Neutrino project uses @ApacheCalcite in combination with @ApachePinot. Nice!

@rolandbouman Time you came over to 🦋 :)

I have started posting daily summaries of the Golden Gate Raptor Observatory hawkwatch to (I used to post to @GGRO_raptors but due to management changes and/or incompetence I am currently unable to post to that account.)

Bluesky is good

Bluesky is the online equivalent of moving to Canada. The maple syrup is great and the people are nice.

@arntzenius Welcome! It’ll be even better when we’ve had some rain. Not long now…

@medriscoll Yup. I went to the Farallons a few years ago. Up close there’s a lot of seabird guano, loud seals/sea lions, and jellyfish.

@arntzenius The hard part for me is crafting the narrative: the order in which I introduce topics, how I motivate what I'm just about to say. When I have the narrative, the slides remind me what to say, and give the audience something to look at. The presentation software is not important.

@KarlKFI @rakyll Or you could be seeing good engineers adapting to their environment. E.g. If you're not allowed to change the API of a service once it has been deployed, you make very sure you get the API right first time.

@KarlKFI @rakyll In my experience it's a dialectic, not a divide. Good engineers can see both points and find a compromise. If you find yourself arguing an engineer you think is type A, it's likely that they think you are putting too much weight on the type B argument and are pushing back.

@MaineFrameworks (Sorry, I don’t mean to rant at you. I am gradually realizing I have been conned by this so-called AI revolution.)

@pamelajhyde Wordle 1,224 5/6* ⬜⬜⬜⬜🟨 ⬜🟩🟩⬜⬜ 🟩🟩🟩⬜⬜ 🟩🟩🟩⬜⬜ 🟩🟩🟩🟩🟩 One of my guesses was “froyo”. I’m so glad that wasn’t the answer. I was desperate.

@pamelajhyde Wordle 1,222 4/6* ⬜⬜⬜⬜⬜ ⬜⬜⬜⬜⬜ ⬜⬜🟩⬜🟩 🟩🟩🟩🟩🟩

@erikdarlingdata @DejanSarka In the second tweet of the thread I introduce “WITH DBSTATE”, which can pass a database snapshot - potentially large number of rows - by reference. Not sure how you learned about my experience. I have never used SQLite.

@unclouded I agree, but I just wanted to hear you say it.

@DolanRyanne And that was the case with Hadoop jobs too. Because each job was just a program (jar file), and there was no schema except that baked into the program. But I'm saying you should pass each job not just the schema but also the data it needs.