News for hackers. Stuff that matters.
Re:PostgreSQL reconsiders its process-based model by idiomaticrust on Monday June 19, 2023 @03:35PM (#36395590)
He is right. Such rewrites cause a lot of problems if your compiler doesn't help you with avoiding data races.

But there is another way.

Re:PostgreSQL reconsiders its process-based model by hnarn on Monday June 19, 2023 @03:39PM (#36395636)
> But there is another way.

Ok?

Re:PostgreSQL reconsiders its process-based model by mycall on Monday June 19, 2023 @03:42PM (#36395660)
Microsoft SQL Server has SQLOS which is another way [0].

[0] https://www.thegeekdiary.com/what-is-sql-server-operating-sy...

Re:PostgreSQL reconsiders its process-based model by carstenhag on Monday June 19, 2023 @03:43PM (#36395673)
The person probably implied that Postgres should switch to another toolchain that guarantees more things at compile time, so probably Rust.
Re:PostgreSQL reconsiders its process-based model by blincoln on Monday June 19, 2023 @04:35PM (#36396151)
If the existing code is old-school enough to use thousands of global variables in a thread-unsafe way, seems like changing it enough to compile as safe Rust code would push the "non-trivial" envelope pretty far.
Re:PostgreSQL reconsiders its process-based model by bb88 on Monday June 19, 2023 @04:54PM (#36396328)
You can take a chunk of code and just rewrite it in Rust. You'll learn a lot quickly by this.
Re:PostgreSQL reconsiders its process-based model by tylerhou on Monday June 19, 2023 @05:22PM (#36396653)
The boundaries within database code are not clear. There are too many interlocking parts to take a nontrivial chunk and rewrite it Rust.
Re:PostgreSQL reconsiders its process-based model by steve_adams_86 on Monday June 19, 2023 @05:17PM (#36396595)
It’s sort of like the inverse of the Matrix when Neo learns kung fu. You realize that you actually don’t know how to program :)
Re:PostgreSQL reconsiders its process-based model by chc on Monday June 19, 2023 @03:43PM (#36395674)
I think it's meant to imply the solution given in their username ("idiomatic Rust").
Re:PostgreSQL reconsiders its process-based model by lelanthran on Tuesday June 20, 2023 @03:34AM (#36400824)
> I think it's meant to imply the solution given in their username ("idiomatic Rust").

I think "Idiom: a tic (Rust)" can also fit if I squint hard enough and decide it looks like a definition from an online dictionary :-)

Re:PostgreSQL reconsiders its process-based model by avgcorrection on Monday June 19, 2023 @04:18PM (#36395986)
Don’t mind the gimmick gallery (username).
Re:PostgreSQL reconsiders its process-based model by zilti on Monday June 19, 2023 @07:45PM (#36397875)
Indeed, Zig is a nice language for this