Async-mmocore

[Alvaro Vasconcelos]

Хроники

1. Prelude
2. Chronicle 1: Harbingers of War
3. Chronicle 2: Age of Splendor
4. Chronicle 3: Rise of Darkness
5. Chronicle 4: Scions of Destiny
6. Chronicle 5: Oath of Blood
7. Interlude
8. The 1st Throne: The Kamael
9. The 1st Throne: Hellbound
10. The 2nd Throne: Gracia
11. The 2nd Throne: Freya
12. Chaotic Throne: High Five
13. The Chaotic Throne 3: Goddess of Destruction
14. Goddess of Destruction Awakening
15. Goddess of Destruction Harmony
16. Goddess of Destruction Tauti
17. Goddess of Destruction Glory Days
18. Goddess of Destruction Lindvior
19. Goddess of Destruction Epeisodion / Valiance
20. Dimensional Strangers
21. Infinite Odyssey
22. Helios
23. Grand Cursade
24. Salvation
25. Fafurion
26. Prologue: Prelude of War

Исходники

Присутствуют

Сборка

Async-mmocore

The Goal

The Async-mmocore is primary designed to Massive Multiplayer Online (MMO) Game Servers. The Goal of the Async-mmocore is to provide a easy way to handle MMO connections to a server abstracting the networking layer complexity.

More infos:

Вы не можете просматривать ссылку пожалуйста воспользуйтесь следующими ссылками Вход или Регистрация

 

[Rozhek]

Hello, i saw your project several months ago. Great work, but:
As i know it is a not a best way to use non blocking IO for low load applications.
When you trading job between threads(steal job) you lose performance in notifications on every select, read, write operation; on locks; on selector weaking up or simply when you use to much threads by your CPUs. And when quanity of thread start rise these problems will scale and possible you can get 30-50% worse result in ops\sec than blocking IO
What do you think about it? Did you have benchmarks in comparasion with blocking IO? Did you introduce some practice to reduce this overhead?