SQL or Death? Seminar Series – Spring 2025
11 comments
·January 30, 2025tanvach
I dunno, personally when I teach SQL to engineers, it’s not the syntax that is the main difficulty (yes it can be improved) but the relational data mental model that trips people up. They all want sequential data processing and loops. Translating that to vectorized code is the closest thing I got to help non-data people grok.
fxttr
I used to hate SQL when I was a backend engineer. I had difficulties understanding exactly this mental model and tried to avoid writing SQL by using ORMs. At some point I “accidentally” switched to data engineering and was forced to get to grips with SQL. I think that in the meantime I warmed up to a functional programming style which helped me to some extent.
cyberax
For me, it's not the syntax, and not even the relational model. They are fairly easy to explain.
No, it's the morass of the interactions between GROUP BY/ORDER BY/HAVING. Like, why isn't there FIRST statement to select the first element in the group?
helge9210
To have the first element, you have to describe a relation of order. It's not ordered by default.
benzayb
"make sql as fast as possible" is a misnomer.
you cannot make a language fast, only its underlying implementation (i.e the specific dbms).
and this is the error I see 90% of the time. there is a confusion between conceptual (model) vs implementation concerns.
the problem with SQL the language is its conceptual foundations -- that it is not currently relational and yet it is used as the language of "relational" DBMSes in the market.
ayhanfuat
It says “making sql go as fast as possible”, referring to the talks about optimizers.
brikym
I choose kusto query language
TheCleric
I choose death. Thanks.
DannyBee
I'm surprised they didn't name it SQL or Death Or 1?
treebeard901
DROP [TheCleric];
Can I take death?
For me, it's the lack of SQLs expression composability. And yes, I'm well aware of various libraries and plug-ins that purport to yield "composable SQL". It's never gonna' happen at-scale due to decades of investment in query optimizers for the popular relational engines.
Just make peace with all of SQL's warts or go NoSQL and make piece with code-bloat and eventual consistency.