Author: Robert Hyatt
Date: 12:13:46 07/08/04
Go up one level in this thread
On July 08, 2004 at 13:33:08, Andrew Wagner wrote: >On July 07, 2004 at 23:18:19, Robert Hyatt wrote: > >>On July 07, 2004 at 22:35:34, Andrew Wagner wrote: >> >>>On July 07, 2004 at 22:22:20, Robert Hyatt wrote: >>> >>>>On July 07, 2004 at 22:04:41, Andrew Wagner wrote: >>>> >>>>>On July 07, 2004 at 21:57:14, Robert Hyatt wrote: >>>>> >>>>>>On July 07, 2004 at 19:27:13, GuyHaworth wrote: >>>>>> >>>>>>> >>>>>>>There is clearly a lot of interest on this bulletin board in the ICGA WCCC 2004 >>>>>>>event in Israel. It is clearly repaying us with some interesting games, >>>>>>>incidents and a close contest. >>>>>>> >>>>>>>Like G-CP and others who would very much like to be there in person, I regret >>>>>>>that I have not been able to attend, and have offered my apologies to Omid and >>>>>>>colleagues who are doing a first class job with the organisation there. >>>>>>> >>>>>>>[ In my case, I would not have been able to attend anytime in the last 10 weeks >>>>>>>if the event had been anywhere but my home town, but that's life. ] >>>>>>> >>>>>>> >>>>>>> >>>>>>>I am happy to receive suggestions for the requirements of an ICGA infrastructure >>>>>>>to support ICGA events, and also happy to hear of good examples of such >>>>>>>infrastructure, technologies involved, and expertise willing to be involved. >>>>>>> >>>>>>>I say this not only with my ICGA V-P hat on, but as a Univ lecturer happy to >>>>>>>engage suitable CS students in suitable, interesting projects which are required >>>>>>>to solve 'real problems' for 'real customers' by the authenticating body (the >>>>>>>BCS) in the UK. Email is more convenient for me than CCC but I don't wish to >>>>>>>stifle discussion on CCC if that is preferred. >>>>>>> >>>>>> >>>>>> >>>>>>This is easy. >>>>>> >>>>>>1. Set up a LAN at the tournament site, with a local machine running a FICS >>>>>>server. >>>>>> >>>>>>2. Require _all_ entrants to be FICS compatible, so that moves are >>>>>>automatically sent between programs with no human intervention, pairing is done >>>>>>electronically and games started automatically, clocks managed by the FICS >>>>>>server, games recorded by the FICS server, tournament crosstable provided by the >>>>>>tournament manager software on the same server, etc... >>>>>> >>>>>>3. No exceptions allowed. See note 2 above. >>>>>> >>>>>> >>>>> >>>>>For the record, I don't think there is any currently available version of the >>>>>FICS software which is that sophisticated. Heck, even ICC doesn't start the >>>>>games automatically. My guess would be that this entire server program would >>>>>have to be written from scratch. No small project, but certainly worthwhile. >>>>> >>>> >>>>Have you played in an ICC tourney lately? IE the "pear" manager does it _all_. >>>>Does the pairings. starts the games. Collects the results. Etc. >>>> >>>> >>> >>>Umm, Bob? Did you forget who you're talking to? I _run_ tournaments on pear. >> >>My error. I was thinking of "mamer" although I think there is another one that >>can auto-start games... >> >> >>>First off, it doesn't start the games automatically, it requires at the least a >>>command from one of the two operators. It also doesn't automatically collect >>>PGN, for example. Not to mention, the ICC software is far more advanced than >>>FICS (though I don't know specifically much about FICS tournaments), and is not >>>open source. >> >>ICC collects the PGN just fine. Pear and others report standings, crosstables, >>etc as the games are finished... >> >> > >ICC collects the PGN, but pear doesn't. In other words, there's no way to say >"give me all the PGN scores from such-and-such a tournament". Anyway, the real >point here is that what you're suggesting isn't a simple process of loading an >existing server program onto a box and letting it rip. There's quite a bit of >work involved. > There are multiple issues. 1. playing games and making them visible to the world. Any FICS server on the planet will do that just fine. Automatic move transmission, automatic clock management, automatic draw claim verification, automatic _everything_ as far as playing a single game goes. That would be a _huge_ step in the right direction. 2. Pairings. Multiple tournament bots do this just fine also. While they might not all handle the automatic game start stuff, one or more do, so that the event could be managed via a bot and eliminate pairing errors as well as getting the games started as soon as possible. 3. PGN. It would not be difficult to write a simple telnet program that would connect to ICC and download the PGN games for a selected set of players. Or, using a local FICS server where only WCCC games are played, loading the PGN for all participants would be an easy thing to do. I even have the basic code to do this already as I have a custom interface that will talk to an FICS server and send commands or whatever. This program could easily be modified to grab the PGN and stuff it into a web server directory so that as soon as games end, the PGN is available for anyone to view. 4. results and standings. The bots already do this. I don't see a "quite a bit of work" effort level here. And once whatever effort is required is done, it doesn't have to be re-done next year. It's all free after that, a do it once type of project that would clean up these events for the forseeable future. I'll even volunteer to modify my code to be a "pgn grabber/poster". Can't get any easier than that... >> >> >>> >>>>>> >>>>>>> >>>>>>> >>>>>>>However, you will note that Chessbase are included in WCCC2004's sponsors, and >>>>>>>are indeed webcasting the event with Boris' commentaries. Chessbase offered to >>>>>>>support the ICGA with their technology at Maastricht in 2002 and that offer was >>>>>>>welcomed by the Association. We are very grateful for Chessbase's continued >>>>>>>interest in the ICGA WCCC and their commitment to webcasting. >>>>>>> >>>>>> >>>>>> >>>>>>Fine. It is a proprietary interface, with a proprietary protocol. Let 'em >>>>>>webcast all they want, but do the main event in a publicly available protocol to >>>>>>prevent one company from "taking over". At ICC, we have had thousands of people >>>>>>looking at games at one time, such as the Deep Blue vs Kasparov match. ICGA >>>>>>events can't even be seen publicly unless you only run windows and download a >>>>>>(what is now free but might not always be since it is commercial) interface for >>>>>>that proprietary protocol. >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>>> >>>>>>> >>>>>>>Finally, I thank Omid again for performing his many roles in parallel with >>>>>>>exercising FALCON very creditably in the event. He has the support of the ICGA >>>>>>>personnel there. >>>>>>> >>>>>>> >>>>>>>Guy Haworth >>>>>>>VP, ICGA >>>>>> >>>>>> >>>>>> >>>>>>Fix the black hole. Nothing is getting out in a timely manner except for >>>>>>participants logging in to CCC as they have time. That is almost criminal >>>>>>behavior for the "International COMPUTER games association" to not make the >>>>>>event available to a COMPUTER network... >>>>>> >>>>>>It just defies all logic. IBM did it for the DB match, in a way that _anybody_ >>>>>>could watch, unix or not. Why can't the ICGA do the same? Why hasn't it _done_ >>>>>>it already. It's not like this is a new and novel idea...
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.