3 ms·
I see your point, but I'm very hesitant to change any configuration of SQLite. It kinda feels like one step too close to doing devops - which one wants to avoid
by fforflo 8y ago
I see your point, but I'm very hesitant to change any configuration of SQLite. It kinda feels like one step too close to doing devops - which one wants to avoid by using SQLite I guess.
Having said that, I do play around with PRAGMA statements when it's really needed, but usually tweaking the code usually works fine - even increasing the timeout is probably enough :D
- BoorishBears 8y agoI read that first sentence like 3 times and still don't get it.
- sansnomme 8y agoOften people use SQLite for sheer convenience instead of perceived performance or resource efficiency i.e. because SQLite is zero config, works out of the box. Sure spinning up a docker container with PostgreSQL is easy enough but why do that when you can use the default standard library with SQLite already embedded and linker configed?
- BoorishBears 8y agoSQLite is an alternative to fopen not a full blown DB... And it’s performance and reliability are a huge part of why it runs on millions of devices everywhere. Your browser was never going to embed a build of PostgreSQL and Docker. Enabling WAL can hardly be called config, it’s a one liner in every driver I’ve ever seen. It’s like saying specifying the file directory SQLite uses is config.
- groovy2shoes 8y ago> It’s like saying specifying the file directory SQLite uses is config. Or, to continue the fopen analogy, like saying that the file you've given fopen is to be opened with mode "a+". The only real difference is that fopen's mode argument is required whereas SQLite's mode has a useful default. > Enabling WAL can hardly be called config, it’s a one liner in every driver I’ve ever seen. Even in the raw libsqlite3 C API (as long as you don't need any error checking ;) ): sqlite3_exec(db, "PRAGMA journal_mode=WAL", NULL, NULL, NULL); presumably because it's just a one-liner in SQLite's dialect.