Computer 1: Win 7 64bit
Computer 2: Ubuntu 32bit
Compilation parameters:
- Used -O2 -mssse3 -mfpmath=sse -DNDEBUG in release mode for both linux
and windows
- Used latest svn revision and:
- Applied http://developer.wz2100.net/changeset/11099
- Applied both patches from http://developer.wz2100.net/ticket/1962
Reminder: Sent / received statistics in the logs are still swapped in
newnet.
Test #1:
Map: 8c Sandcastles T1
Host: Computer 1
Additional details: Host in slot 0, computer 2 in slot 1, 6 dydo ai's in
remaining slots.
Summary: Instant de-sync. Looking a the logs it seems only p0 added
power to the AI's at the start, power seems to be the only difference in
the desync logs. I didn't let the game run long.
Notes: I tried this setup twice.
Logs: http://filebin.ca/qsgaoq/logs1.tar.gz
Conclusion: This will need to be resolved before I try this setup again.
Test #2
Map: 8c NTW2 T3
Host: Computer 2
Additional details: Host in slot 0, computer 1 in slot 1, 1 standard ai
in slot 7.
Summary: Played a ~24 minute game without any desync errors. There
wasn't any action on computer 1.
Logs:
netplay (computer 1, trimmed): http://pastebin.com/DvSinLnz
netplay (computer 2, trimmed): http://pastebin.com/n5tx79Ek
----- Start WZlog (trimmed)----
12:21:19: [recvDroidInfo] Packet from 0 refers to non-existent droid
168664, [Human : p0]
last message repeated 2 times
12:21:28: [recvDroidInfo] Packet from 0 refers to non-existent droid
173968, [Human : p0]
last message repeated 2 times
----- End WZlog (trimmed)----
Supplemental connection statistics:
Computer 1: ~12MB down, ~4MB up (IIRC)
Computer 2:
Wireshark summary:
Packets: 48408
Duration: 1537 s
Avg. packet size: 359.7 bytes
Avg. packet /s : 31.5
Bytes: 17410957
Avg. bytes /s: 11325
_______________________________________________
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev
