Re:Use SQL subqueries to count distinct 50x faster
by lvh on Friday January 24, 2014 @08:38AM
(#7114812)
The declarative language promise was that it would be possible to perform such optimizations. Indeed, by-and-large, query planners do a pretty great job. I don't think it was ever a promise that query planners would be able to always find the fastest query for what you want.
[ Parent
| on HN ]
Re:Use SQL subqueries to count distinct 50x faster
by jerf on Friday January 24, 2014 @09:38AM
(#7115024)
I read the declarative language promise as you will never have to know what's going on under the hood. This is never true in any non-trivial declarative language being used for non-trivial work. I have developed a very negative initial reaction to anything that claims to be "declarative".
Re:Use SQL subqueries to count distinct 50x faster
by Pxtl on Friday January 24, 2014 @04:12PM
(#7118069)
Likewise. The more declarative a platform is, the more I'm trusting its abstractions to never leak, because when I do I'm going to have to learn every agonizing minute detail of how it works under the hood.
SQL is a tiny kernel of beautiful relational mathematics wrapped up in a 3-foot ball of ugly implementation hacks.
NoSQL is what happens when you throw out the beautiful kernel and just keep the hacks, so it's not really better I guess.