3 ms·
How about instead of making AI wrappers to over 50 years old SQL, we’d make a database query language that’s easier to read an write?
by miohtama 3y ago
How about instead of making AI wrappers to over 50 years old SQL, we’d make a database query language that’s easier to read an write?
- marginalia_nu 3y agoIn general, if something has been around for a very long time and nobody apparently seems to have thought to improve it, then odds are the reason is it's pretty good and genuinely hard to improve on.
- aae42 3y agoin other words, SQL is a shark, not a dinosaur
- realanswe91 3y agoThe Shark Query Language
- fbdab103 3y agoCodd himself had loads of complaints about SQL.
- oblio 3y agoWell, if you think about it, a huge percentage of dinosaurs survived (birds), and for the rest, can you really fault them for going extinct when a humongous boulder hit the planet?
- runlaszlorun 3y ago> pretty good and genuinely hard to improve on I think that might be a bit more positive than I would be. Broadly speaking, I think you could say that the downsides of the legacy technology in question aren’t larger than the collective switching costs. But I’d definitely agree that when something has been around that long, it’s prob not all bad.
- SoftTalker 3y agoBecause SQL is a good query language, if you bother to really learn it.
- neodymiumphish 3y agoMy fear with this approach is that the first implementation would be severely handicapped compared to SQL, amd it'd take years to support some one-off need for any organizational user, so it'd never be fully utilized.
- drittich 3y agohttps://prql-lang.org/ https://prql-lang.org/ might be an answer for this. As a cross-database pipelined language, it would allow RAG to be intermixed with the query, and the syntax may(?) be more reliable to generate
- aitchnyu 3y agoI'm watching Edgeql, I feel it could end the career of AI query generators, ORMs and low end BI. https://www.edgedb.com/ https://www.edgedb.com/