[DiscordArchive] is the entire phase treated like another map?
[DiscordArchive] is the entire phase treated like another map?
Archived author: Foe • Posted: 2023-09-30T10:12:30.959000+00:00
Original source
Unless you want to parallelize the per-map update task and update world chunks in parallel
Archived author: Foe • Posted: 2023-09-30T10:12:40.554000+00:00
Original source
This would still not scale as well as sharding
Archived author: Pandaros • Posted: 2023-09-30T10:13:55.968000+00:00
Original source
what I was thinking is map has these maps, or anything in map update that is
```cpp
MapRefMgr m_mapRefMgr;
ActiveNonPlayers m_activeNonPlayers;
```
I will add a 2nd dimension to these maps with the phase
pseudo code:
```cpp
map<uint, MapRefMgr> m_mapRefMgr;
map<uint, ActiveNonPlayers> m_activeNonPlayers;
```
Then in update modulus all the known phase IDs by thread count, if there is more than 1 thread count. pass those to the update threads that do the work.
Archived author: Pandaros • Posted: 2023-09-30T10:15:53.106000+00:00
Original source
to effectively virtualize the existence the phase
Archived author: Pandaros • Posted: 2023-09-30T10:16:01.966000+00:00
Original source
without having to refactor all of map