To bring this to the widest possible audience copies will also be posted on TWGS.ORG and the EISOnline forums.

In regards to the Buffer problems with TWGS, I have run tests a a series of popular servers. Here are the results:

Server Time Cycles Users in Game BBS
Coast Games 34 762 0 No
Silver Wings 70 750 6 No
Chicken Coop 71 758 0 Yes
PublicAtLarge 66 783 0 No
PLAY TW 67 763 0 No
Dragon 57 4445 0 No
Fament 327 4447 0 Yes
Deep South 57 770 0 No
Hard Coded 83 4446 11 No

I was able to demonstrate the problem on both Mission Nighthawk and Tea BBS, but not get data as to number of cycles run and time used. The results are interesting. The MBBS frontend at Fament appears to have had a moderating effect on the test. Dragon and Hard Coded may both be good connects for me, certainly the results were at variance with the usual TWGS results.

And as to Macahan's comment that in effect it's a DOS attack, tell me if you know of any TW player who would be worried about messing over his/her competition? Really?

For anyone else who wishes to run the test, here is the Qmodem script. Note that translating it to ZOC will probably note work, ZOC is compatatively slow and does not appear to be able to overload the TWGS. Telix is fairly fast, as is Procom, and both will probably be able to overload the TWGS. SWATH, TWAR, ATTAC and the other helpers may or may not be able to overload it. I have little experience with them, comments (and test results using other telnet clients would be usefull).

QMODEM Test Script
------
Dim i as integer
do while i<5001
SEND "'"
send i;
SEND " Dr. Who is a geek";
SEND " who complains that BBS frontends don't work."
SEND
i=i+1
loop
end
-----
End Script

When posting results please post the script, complete with all modifications needed to run it.

The Mad Hatter