Author: Robert Hyatt
Date: 09:10:51 11/08/01
Go up one level in this thread
On November 08, 2001 at 07:57:37, Vincent Diepeveen wrote: > >Aha now someone with experience using auto232 player is going to >open his mouth. There are many OS-es which get linked on the auto232 player. > >laptops with win98 versus old 'servers' with NT. and they all use auto232. > OK.. first, lets get on the same _page_. I have said this many times before, but again won't hurt: "Auto232 is a piece of trash." I'm not talking about auto232. I am talking about everyone using either winboard/xboard, or at least following that protocol, or else they use their own custom interface that can communicate with a chess server (IE what Bruce and I both did years ago). >now the auto232 protocol has bugs, but whatever your protocol is, what are >you going to do in case that 2 programs stop playing each other because of >disturbance? > using a local chess server, if a program hangs, it doesn't do a thing. The machine reboots, it reconnects, and the game continues. >I mean it could be the com port of a laptop. Some laptops have *bad* comports. > >It could be the network card. Some network cards are a joke as we all know. > Ok... it should be easy to get a local vendor to supply a box full of Intel EtherExpress pro 100 cards. They sell for $35 bucks apiece. >It could be the windows OS which blocks input for a while or whatever or >throws away packets. Doesn't matter. TCP/IP is error-free. Lost packets are re-sent with no problems. > >It could be program A > >it could be program B. > >What you do here? If a program can't operate correctly, it forfeits that game and is withdrawn from the event if it can't be fixed. Very simple. >Operate by hand again? >Most likely the libraries fritz uses are disturbing the most because >the interface it's using is written in MFC. With tiger i have hardly problems >auto232 playing. Idem Shredder. Both are seemingly non-mfc. But fritz is >mfc, so game after game one gets problems. Perhaps not even a bug from fritz >itself. Most likely simply MFC. For the last time, "FORGET AUTO232, IT IS A PIECE OF TRASH." > >Now one tournament goes ok. Second tournament suddenly a company C says: >"we no longer support this protocol we have developed our own similar protocol" Tough. If they can connect to the local chess server running FICS code, then they can participate. If they can't, they don't participate. Period. > >So they force you that they can act as server, so they determine who forfeits >and who didn't. There will be _one_ tournament server that _everybody_ connects to. This server will run FICS code. It will look _exactly_ like FICS. > >Suddenly 10% of all games they win by forfeit. Other games are seemingly >shorter, like the opponent gets 10 minutes less a game *somehow*. You are stuck on auto232. Forget it. I am using FICS code. > >Even worse, the *only* way to see the game is to watch their display, because >they dissallow the opponent saving the game (as they changed protocol). Nope. Using FICS code, we can have a group of PCs against the wall, all running xboard, using a thin-film projector to show the games on the wall as they progress, independent of the monitors of the two participants. Just log on a third PC and type "observe Crafty" to show my game on that machine. > >Also they delete the learning files on the harddisk of the opponent, or simply >kill the process and restart it. > >Would you accept that? Not with FICS they won't. > >SSDF did. > >Vincent
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.