Forums WoW Modding Support Archives Azerothcore Discord Archives [DiscordArchive] why is there a thread handling mail expirations?

[DiscordArchive] why is there a thread handling mail expirations?

[DiscordArchive] why is there a thread handling mail expirations?

Pages (4): Previous 1 2 3 4 Next
rektbyfaith
Administrator
0
11-11-2024, 11:23 AM
#11
Archived author: andriuspel • Posted: 2024-11-11T11:23:07.950000+00:00
Original source

that "self destruction" call is most dangerious for now
rektbyfaith
11-11-2024, 11:23 AM #11

Archived author: andriuspel • Posted: 2024-11-11T11:23:07.950000+00:00
Original source

that "self destruction" call is most dangerious for now

rektbyfaith
Administrator
0
11-11-2024, 11:29 AM
#12
Archived author: Takenbacon • Posted: 2024-11-11T11:29:54.686000+00:00
Original source

sounds about right considering it was just discussed that the current AH threading model is bad and unsafe. slapping a shared_ptr on items is attempting to address a symptom and not the actual problem
rektbyfaith
11-11-2024, 11:29 AM #12

Archived author: Takenbacon • Posted: 2024-11-11T11:29:54.686000+00:00
Original source

sounds about right considering it was just discussed that the current AH threading model is bad and unsafe. slapping a shared_ptr on items is attempting to address a symptom and not the actual problem

rektbyfaith
Administrator
0
11-11-2024, 11:35 AM
#13
Archived author: andriuspel • Posted: 2024-11-11T11:35:42.359000+00:00
Original source

its not like i'm trying to hide, simply shared_ptr was designed for such cases, to avoid deallocation, memory corruption cases easier, since shared_ptr has atomic counter and when counter hits 0, object memory is removed
rektbyfaith
11-11-2024, 11:35 AM #13

Archived author: andriuspel • Posted: 2024-11-11T11:35:42.359000+00:00
Original source

its not like i'm trying to hide, simply shared_ptr was designed for such cases, to avoid deallocation, memory corruption cases easier, since shared_ptr has atomic counter and when counter hits 0, object memory is removed

rektbyfaith
Administrator
0
11-11-2024, 11:36 AM
#14
Archived author: andriuspel • Posted: 2024-11-11T11:36:08.414000+00:00
Original source

also shared_ptr is meant for shared ownership
rektbyfaith
11-11-2024, 11:36 AM #14

Archived author: andriuspel • Posted: 2024-11-11T11:36:08.414000+00:00
Original source

also shared_ptr is meant for shared ownership

rektbyfaith
Administrator
0
11-11-2024, 11:36 AM
#15
Archived author: andriuspel • Posted: 2024-11-11T11:36:49.569000+00:00
Original source

unique_ptr differs mainly that it can hold only single object entry and can be only one unique object
rektbyfaith
11-11-2024, 11:36 AM #15

Archived author: andriuspel • Posted: 2024-11-11T11:36:49.569000+00:00
Original source

unique_ptr differs mainly that it can hold only single object entry and can be only one unique object

rektbyfaith
Administrator
0
11-11-2024, 11:38 AM
#16
Archived author: andriuspel • Posted: 2024-11-11T11:38:35.395000+00:00
Original source

there is possible to use mutexes or other methods, or maybe even same lambda with item capture events
rektbyfaith
11-11-2024, 11:38 AM #16

Archived author: andriuspel • Posted: 2024-11-11T11:38:35.395000+00:00
Original source

there is possible to use mutexes or other methods, or maybe even same lambda with item capture events

rektbyfaith
Administrator
0
11-11-2024, 11:39 AM
#17
Archived author: andriuspel • Posted: 2024-11-11T11:39:17.494000+00:00
Original source

value locking methods are not safe in general
rektbyfaith
11-11-2024, 11:39 AM #17

Archived author: andriuspel • Posted: 2024-11-11T11:39:17.494000+00:00
Original source

value locking methods are not safe in general

rektbyfaith
Administrator
0
11-11-2024, 11:41 AM
#18
Archived author: andriuspel • Posted: 2024-11-11T11:41:11.410000+00:00
Original source

and weak_ptr is like, shared_ptr locker, it can lock shared_ptr pointer before actual object removal (aka, before counter hits 0)
rektbyfaith
11-11-2024, 11:41 AM #18

Archived author: andriuspel • Posted: 2024-11-11T11:41:11.410000+00:00
Original source

and weak_ptr is like, shared_ptr locker, it can lock shared_ptr pointer before actual object removal (aka, before counter hits 0)

rektbyfaith
Administrator
0
11-11-2024, 11:50 AM
#19
Archived author: walkline • Posted: 2024-11-11T11:50:07.729000+00:00
Original source

Well, I think Takenbacon and I are not against smart pointers in general. Refactoring to use smart pointers in some places isn't a bad idea. But if we are talking about this in the context of fixing things with smart pointers, there are not that many issues (at least not that I'm aware of) that are actually broken. This auction issue is more like an exception and it's definitely broken, and using smart pointers alone isn't enough to fix it.
rektbyfaith
11-11-2024, 11:50 AM #19

Archived author: walkline • Posted: 2024-11-11T11:50:07.729000+00:00
Original source

Well, I think Takenbacon and I are not against smart pointers in general. Refactoring to use smart pointers in some places isn't a bad idea. But if we are talking about this in the context of fixing things with smart pointers, there are not that many issues (at least not that I'm aware of) that are actually broken. This auction issue is more like an exception and it's definitely broken, and using smart pointers alone isn't enough to fix it.

rektbyfaith
Administrator
0
11-11-2024, 11:54 AM
#20
Archived author: andriuspel • Posted: 2024-11-11T11:54:33.580000+00:00
Original source

there could be other way around - refactor item removal queue
rektbyfaith
11-11-2024, 11:54 AM #20

Archived author: andriuspel • Posted: 2024-11-11T11:54:33.580000+00:00
Original source

there could be other way around - refactor item removal queue

Pages (4): Previous 1 2 3 4 Next
Recently Browsing
 1 Guest(s)
Recently Browsing
 1 Guest(s)