Post #2466643
2025-08-19 07:41 UTC
@ludicity@mastodon.sprawl.club recently was interviewing a data engineer who was speaking of, believe me, importance of B-tree depths in Index Joins performed by Spark.
That was some gooooooooood crack.
I couldn't tell if he was speaking from the outside source, like LLM or if it was just panic before the interview and scrambling all DB related knowledge last night without sleep though. Given the speed he was speaking about all the things I think the latter, he didn't look like reading.
My bet would be that
Replies (2)
-
@ldvsoft@fosstodon.org 2025-08-19 07:42
@ludicity@mastodon.sprawl.club behavoiur would be hardly distinctive from a bit more sophisticated candidate reading hints a couple of times from LLM verbatim.
-
@stevel@hachyderm.io 2025-08-19 15:24
@ldvsoft@fosstodon.org @ludicity@mastodon.sprawl.club got to respect their knowledge of B+ trees though. And as for what Spark does internally-I'd have to look at the source to make sure it wasn't doing it either directly or through Scala collections