Author: Dann Corbit
Date: 21:33:15 09/29/03
Go up one level in this thread
On September 29, 2003 at 23:58:00, Steven Edwards wrote: >On September 29, 2003 at 23:23:22, Jon Dart wrote: > >>Re having an alternate chess server: sure, maybe. But there's a critical mass >>problem. ICC and FICS succeded because they have a large pool of players and >>some suitable opponent is almost always available, whatever time of day or night >>you are online. If you can't get enough participants to achieve that, then it's >>not likely to catch on, IMO. > >Point taken. At the moment my machinery is generating over 20 SETI@home work >units a day and that can be trimmed considerably to provide some computing power >to run a few "always connected" programs, like my old program Spector, my new, >upcoming program Symbolic, and a local version of Crafty. > >Re comments on pool size: my idea is to go for quality vs quantity. > >Because the proposed FICS is for authors only, it can be customized to better >support program players than is the case at other servers. Examples: > >1. Multiple accounts for different program versions by the same author. > >2. Account attribute data includes hardware and software configuration items. > >3. Local copies of opening books and tablebases could be used. > >4. Availability of a more program-friendly formatted set of commands and status >I/O; possible elimination of FICS features unneeded for program usage. > >5. Program authors' suggestions for feature set enhancement are taken seriously. > >6. A facility for periodic mailing of the rating list; also, periodic mailing >all PGN game scores for those accounts that allow disbursement of the score >data. > >7. The possibility of automated pairings, matches, and tournament organization; >programs could check the server periodically to see if they've been paired and >when and with whom the game is scheduled. Post-event tourament score >crosstables could be automatically mailed. > >8. The possibility of the server taking an active role in initiating a session >with a remote progarm if the program's account has authorize such usage. > >>On a technical note: a version of FICS source was released years ago (on the U. >>of Pittsburgh site I think) but IIRC it was BSD-based and at the time I couldn't >>easily get it to run on a commercial UNIX. So you might need to do some hacking >>on it. Maybe FICS itself has something cleaner, but if so they haven't released >>it, as far as I know. > >I've seen the FICS 1.7.4 sources. I might wind up writing my own version using >my toolkit for all the chess specific items. An open source chess server that also handled correctly the timing stuff would be nice to see. I suspect a really good chess server could be composed from ACE (for the client/server/communication parts) and your toolkit for the chess stuff. ACE: http://www.cs.wustl.edu/~schmidt/ACE.html
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.