[DiscordArchive] so any C++ class instance at all is always invalidated as soon as the trigger is over ? so each trig
[DiscordArchive] so any C++ class instance at all is always invalidated as soon as the trigger is over ? so each trig
Archived author: jackpoz • Posted: 2024-03-09T23:28:47.532000+00:00
Original source
so any C++ class instance at all is always invalidated as soon as the trigger is over ? so each trigger is its own little sandbox ?
Archived author: jackpoz • Posted: 2024-03-09T23:29:24.983000+00:00
Original source
that's actualy why I stopped with ArcSole or what it was
Archived author: jackpoz • Posted: 2024-03-09T23:29:45.179000+00:00
Original source
there was no way for taxipaths to ever be safe
Archived author: Rochet2 • Posted: 2024-03-09T23:29:46.687000+00:00
Original source
Only raw pointers.
If you push some lua owned data (Like LuaVal, .. some other type), then there is no invalidation needed.
Archived author: jackpoz • Posted: 2024-03-09T23:30:45.997000+00:00
Original source
ideally, there should be 0% chance for a buggy script (no matter the language) to cause issues/crashes, but that is rarely easy
Archived author: jackpoz • Posted: 2024-03-09T23:31:05.243000+00:00
Original source
you sort of need an abstraction layer between scripts and game
Archived author: Foe • Posted: 2024-03-09T23:31:38.601000+00:00
Original source
The only really reoccurring crash I have seen in forever is invalid sql queries
Archived author: Foe • Posted: 2024-03-09T23:32:15.448000+00:00
Original source
But when you raw dog some sql then it is what it is I guess
Archived author: Foe • Posted: 2024-03-09T23:32:56.450000+00:00
Original source
I'm also guessing that is deliberate with the query implementation in the core
Archived author: Foe • Posted: 2024-03-09T23:33:04.257000+00:00
Original source
(for data safety and all)