About my net lag,

Don asked, "What do you do when you play on KGS?"

When playing against humans, there is hardly ever a problem because
humans either resign quickly, pass normally, or quit the game (ugh!) -
humans don't notice the net lag.  In the kgs computer tournaments,
it's still my bot that ends up resigning anyway ;(

>From what you say Don, the server is already using slight Fisherish
rules - 5 min + .25 sec ?  I like your idea of making it even more
Fisherish.  I think it would help normalize the ratings in the
1300-2000 bracket where my bots play, if it were increased to 5 min +
1 sec, or 3-4 min + 2 sec, or something like that.  It sounds like an
easy-enough change and one that doesn't seem to have bad side-effects.

Thanks again for the great work in putting up the CGOS server & to all
the other programmers who put bots up there to play.

Peter

p.s.  This is my network setup - ethernet from laptop to a tower on
the roof of my house. Parabolic Directional antenna wireless B, to an
intermediate tower 5km away receives it on a 180 degree nondirectional
antenna, sends it via directional antenna over wireless b to a
directional antenna at my ISP it-outlook another 5km away. From the
ISP I think there are several more wireless hops, eventually to
Manila, Philippines, then across some cable (I assume a cable).

The real dog, as you can see, is the hop across the ocean, about 80%
of my ping time.  I assume that from Japan they have shorter ping
because they have better infrastructure but I'd be curious to hear a
confirmation of that.  I know that in China some technical
universities even have terrible networks (I was trying to get Yu Ping,
7d chinese pro pursuing his masters in go programming, to put his java
bot on cgos but to the best of my knowledge I failed to get him to do
it. maybe china blocks the cgos port.  don't try to read wikipedia
from china either, it's blocked along with anything else that has
certain unofficial accounts of supposed history.)

1 [EMAIL PROTECTED] tracepath 208.100.19.102                                ~
 1:  192.168.1.51 (192.168.1.51)                            0.242ms pmtu 1500
 1:  192.168.1.1 (192.168.1.1)                            asymm 36  14.985ms
 2:  172.16.0.254 (172.16.0.254)                           33.483ms
 3:  router.it-outlook.noc (192.168.1.254)                 27.493ms
 4:  222.127.77.143 (222.127.77.143)                      asymm  3
34.011ms pmtu 1452
 5:  203.177.68.197 (203.177.68.197)                       84.384ms
 6:  203.177.31.89 (203.177.31.89)                         84.382ms
 7:  203.177.254.185 (203.177.254.185)                    369.732ms
 8:  POS3-1.IG4.LAX1.ALTER.NET (157.130.214.193)          asymm 14 337.252ms
 9:  0.so-5-0-0.XL2.LAX1.ALTER.NET (152.63.112.254)       asymm 12 342.739ms
10:  0.so-5-0-0.XL2.CHI13.ALTER.NET (152.63.64.14)        asymm 13 362.669ms
11:  POS7-0.GW1.CHI13.ALTER.NET (152.63.69.181)           asymm 13 378.225ms
12:  wbsconnect-gw.customer.alter.net (65.207.236.126)    asymm 19 378.225ms
13:  61.216-86-149.static.steadfast.net (216.86.149.61)   asymm 19 540.291ms
14:  vswitch3.steadfast.net (216.86.146.24)               asymm 15 363.857ms
15:  boardspace.net (208.100.19.102)                      378.845ms reached
     Resume: pmtu 1452 hops 15 back 15

There is also about 15% packet loss (I think all in the first leg from
my house to the ISP) that depending on the protocol probably doesn't
make much difference because it's only a 30ms trip to get it resent.

0 [EMAIL PROTECTED] ping  208.100.19.102                                    ~
PING 208.100.19.102 (208.100.19.102) 56(84) bytes of data.
64 bytes from 208.100.19.102: icmp_seq=1 ttl=50 time=333 ms
64 bytes from 208.100.19.102: icmp_seq=2 ttl=50 time=340 ms
64 bytes from 208.100.19.102: icmp_seq=4 ttl=50 time=327 ms
64 bytes from 208.100.19.102: icmp_seq=6 ttl=50 time=344 ms
64 bytes from 208.100.19.102: icmp_seq=7 ttl=50 time=332 ms
64 bytes from 208.100.19.102: icmp_seq=8 ttl=50 time=425 ms
64 bytes from 208.100.19.102: icmp_seq=9 ttl=50 time=362 ms
64 bytes from 208.100.19.102: icmp_seq=10 ttl=50 time=350 ms
64 bytes from 208.100.19.102: icmp_seq=11 ttl=50 time=337 ms
64 bytes from 208.100.19.102: icmp_seq=12 ttl=50 time=375 ms
64 bytes from 208.100.19.102: icmp_seq=13 ttl=50 time=346 ms
64 bytes from 208.100.19.102: icmp_seq=14 ttl=50 time=384 ms
64 bytes from 208.100.19.102: icmp_seq=15 ttl=50 time=342 ms
64 bytes from 208.100.19.102: icmp_seq=16 ttl=50 time=398 ms
64 bytes from 208.100.19.102: icmp_seq=17 ttl=50 time=335 ms
64 bytes from 208.100.19.102: icmp_seq=18 ttl=50 time=339 ms
64 bytes from 208.100.19.102: icmp_seq=19 ttl=50 time=327 ms
64 bytes from 208.100.19.102: icmp_seq=20 ttl=50 time=365 ms
64 bytes from 208.100.19.102: icmp_seq=21 ttl=50 time=336 ms

--- 208.100.19.102 ping statistics ---
22 packets transmitted, 19 received, 13% packet loss, time 21134ms
rtt min/avg/max/mdev = 327.380/352.887/425.192/25.698 ms
_______________________________________________
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/

Reply via email to