No server process and a single file per DB which I can put wherever I like.
No server process and a single file per DB which I can put wherever I like.
"In a CREATE TABLE statement, if the "STRICT" table-option keyword is added to the end, after the closing ")", then strict typing rules apply to that table. ... The STRICT keyword at the end of a CREATE TABLE statement is only recognized by SQLite version 3.37.0 (2021-11-27) and later.
sqlite> create table x (a int);
sqlite> insert into x values ('hello');
sqlite> select * from x;
hello
sqlite> drop table x;
sqlite> create table x (a int) strict;
sqlite> insert into x values ('hello');
Runtime error: cannot store TEXT value in INT column x.a (19)Basically, AFAIK, SQLite becomes problematic once you need more than one computer to handle requests.
They're probably not as battle tested as the PostgreSQL ones, but they are around, have users, and are actively developed.
The ones I remember off the top of my head:
* https://github.com/rqlite/rqlite <-- more of a "distributed database using RAFT" type of thing
Honestly though, if I need these sorts of distribution features, I would probably prefer the database to have them built in. I don't really see the point in using SQLite at that scale.
https://www.sqlite.org/whentouse.html
https://www.sqlite.org/quirks.html
Full-scale RDBMSs, especially Postgres, have lots of goodies that either SQLlite doesn't have (or which it does have, but which aren't so richly featured). Once you've gotten hooked on a few of these, the distinction will feel a lot more clear.
Meanwhile the tipping points in favor of SQLite seem to be embedded systems, and its whole "service-less" architecture and plain ease of use. Which is why it still gets lots of love, for those contexts.
It's a good reminder to give some thought to whether one actually needs MySQL|Postgres. If not, SQLite is the way to go. Most of my code that uses a DB is using SQLite.
But obviously, if you actually need MySQL|Postgres then SQLite is not an option.