@erikdarlingdata, seems like a good day to ask the SOTU on how the latest Claude Opus handles advanced query tuning challenges, where the right answer is some obscure technique described only on top-end blogs (your/Obbish, Paul White's, etc)
@erikdarlingdata Understandable, tho I was asking b/c that LLM ability/inability represents the gap between your skillset and a bright/persistent non-specialist trying to figure out a tough tuning problem
@SQL_Kiwi@erikdarlingdata (Win11 sucks on my dated laptop so I'm wiping and putting XFCE Linux on it and playing around with some open source DBs... but would still love to experiment with my fave DB)
@SQL_Kiwi do you know if the optimizer code and behavior is essentially the same on Linux containers? (Not that you'd ever waste time on that :-) )
As in, if I follow along with your and @erikdarlingdata 's blog posts, should I expect to see the same things?
(I'd been watching some old Paul Randal Pluralsight courses and he demonstrates a tail-of-the-log backup when the MDF has been deleted, and I got curious if ChatGPT would get this right.
No. It just lost your user data pursuing forensics...
)
I know we're all using the cloud now so this doesn't matter anymore (:-P), but ChatGPT does not understand tail of the log backups:
https://t.co/AomZKQq4W1
@erikdarlingdata@SQL_Kiwi This is confirming my theory that the prime audience of these apps are dudes like me who are not only middle-aged, but also over-the-hill and feeling it.π
@SQL_Kiwi@erikdarlingdata , random Q, but do either of you use "PKM" (Personal Knowledge Management) tools for managing the vast amounts of obscure details you accumulate in your careers?
(e.g. Logseq, Obsidian, Roam Research)