Check out Glinski's Hexagonal Chess, our featured variant for May, 2024.


[ Help | Earliest Comments | Latest Comments ]
[ List All Subjects of Discussion | Create New Subject of Discussion ]
[ List Earliest Comments Only For Pages | Games | Rated Pages | Rated Games | Subjects of Discussion ]

Comments/Ratings for a Single Item

Earlier Reverse Order Later
Can CVP site have a chess variants server eventually[Subject Thread] [Add Response]
Kevin Pacey wrote on Thu, Feb 8 02:54 PM UTC:

Free chess server sites sometimes even include a number of chess variants for play, such as on FICS (Free Internet Chess Server). Bughouse is even an option there, though I don't know if it's just for between two players.

I don't know if Game Courier can ever handle very fast-moving games played of CVs, but maybe a special interface like used for [even free] chess servers might eventually be possible, if it seems worthwhile. It might get more members (especially active ones) to CVP site.


Kevin Pacey wrote on Thu, Feb 8 03:33 PM UTC in reply to Kevin Pacey from 02:54 PM:

Here is a free online open source fairy chess server designed to play several CVs:

https://www.pychess.org/about


H. G. Muller wrote on Thu, Feb 8 03:38 PM UTC in reply to Kevin Pacey from 02:54 PM:

I have been running the open-source FICS code on my server for a while, mainly to organize on-line tournaments for computer programs. I did add a lot of variants on it too, such as Capablanca 10x8 variants, Spartan Chess. No one would ever play those, of course, but there were hardly any human visitors of that server anyway, and I never organized computer tournaments for those on that server.

I don't know if that old source still compiles on CentOS; otherwise it would be possible to run it on the CVP server as well. But it would have to be changed a lot for allowing a wider variety of variants, and in particular easier implementation of new ones. Furthermore, the only client capable to run the variants is WinBoard/XBoard, while nowadays people lose interest for anything you cannot play directly in your browser.


Kevin Pacey wrote on Thu, Feb 8 04:53 PM UTC in reply to H. G. Muller from 03:38 PM:

Very few (or zero) server visitors being interested in Spartan Chess I think I can understand, but I am a bit surprised that the final form of 10x8 Capablanca Chess (the variant made by the famous Cuban World Chess Champion, i.e. not all the other similar 10x8 CVs) would attract very little interest too. Of course, if few people visited that server anyway...


🕸Fergus Duniho wrote on Thu, Feb 8 05:24 PM UTC in reply to Kevin Pacey from 02:54 PM:

We have had a Chess variants server since 2001. It's called Game Courier, and it supports more Chess variants than all the others combined.

I don't know if Game Courier can ever handle very fast-moving games played of CVs

If you stay on the same game instead of checking your logs, Game Courier will check whether your opponent has moved every 7.5 seconds and update the page with an audible beep when he has.


H. G. Muller wrote on Thu, Feb 8 06:17 PM UTC in reply to Fergus Duniho from 05:24 PM:

We have had a Chess variants server since 2001. It's called Game Courier, and it supports more Chess variants than all the others combined.

This is what people call a 'turn-based server'. Which is fine for correspondence chess, but incomparable to a real-time server like FICS or LiChess. People play bullet games there, meaning 2 min per game per person, and sub-second recording of thinking time. To play such a game you need to have deposited your move about 1 sec after your opponent did. (And that should then be mostly their own thinking, not the roundtrip delay of the network...)


🕸Fergus Duniho wrote on Thu, Feb 8 07:25 PM UTC in reply to H. G. Muller from 06:17 PM:

I have made Game Courier's delay in checking for your opponent's move more variable, depending upon the time controls. I have replaced 7500 with a variable called delay, and I have set it with this PHP code. Note that the time controls are in seconds, and the delay is in microseconds.

$totaltime = $gracetime + $opptime;
if ($totaltime > 24*3600)
    echo "var delay = 7500;\n";
elseif ($totaltime > 3600)
    echo "var delay = 4000;\n";
elseif ($totaltime > 1800)
    echo "var delay = 2000;\n";
elseif ($totaltime > 900)
    echo "var delay = 1000;\n";
elseif ($totaltime > 0)
    echo "var delay = 500;\n";
else
    echo "var delay = 7500;\n";

Kevin Pacey wrote on Thu, Feb 8 09:18 PM UTC in reply to Fergus Duniho from 07:25 PM:

@ Fergus:

If the change works well, maybe some change to the documentation in the following link (or elsewhere on CVP site) might be worthwhile, to advertise that there is server-like speed on GC if desired:

https://www.chessvariants.com/play/pbm/userguide.html#timecontrols


Jean-Louis Cazaux wrote on Thu, Feb 8 09:27 PM UTC in reply to Kevin Pacey from 09:18 PM:

Well, providing one understands what is written there. It is so complex, I read this part tens of times and I still don't catch nothing at all. Not sure it will be good for advertising...


Kevin Pacey wrote on Thu, Feb 8 10:02 PM UTC in reply to Jean-Louis Cazaux from 09:27 PM:

Maybe a [new??] suggested GC blitz time control of 5 minutes per player might be offered, if possible? Game Courier already allows for something like Increments, so Fischer or Bronstein Delay blitz time controls might be options, too (I don't know if chess servers offer such these days - I only ever used a chess server once, briefly, decades ago, after a friend gifted me a chess server's gift card he won as a [door?] prize at a chess event, that he didn't really want, and my internet was dialup then, and I constantly lost my connection...).


🕸Fergus Duniho wrote on Thu, Feb 8 10:15 PM UTC in reply to Kevin Pacey from 10:02 PM:

Maybe a [new??] suggested GC blitz time control of 5 minutes per player might be offered, if possible?

Do you mean for the whole game?

Game Courier already allows for something like Increments, so Fischer or Bronstein Delay blitz time controls might be options, too

Game Courier has complex, configurable time controls, but I'm not familiar with the time controls you mention.

I suppose for games with Blitz time controls, I should also make the invitations expire much sooner. It wouldn't do for someone to accept an invitation to a blitz game when the person who issued it was asleep or at work or something.


🕸Fergus Duniho wrote on Thu, Feb 8 10:21 PM UTC in reply to Jean-Louis Cazaux from 09:27 PM:

It means that it normally checks whether the opponent has moved every 7.5 seconds, but when tighter time controls are set, it will check at smaller intervals, ranging from every 4 seconds to every half second, depending upon how quickly the game is supposed to go.


Kevin Pacey wrote on Thu, Feb 8 10:29 PM UTC in reply to Fergus Duniho from 10:15 PM:

Yes, I meant for the whole game (standard meaning of blitz time control in chess). For Fischer increments a 2 second bonus can be added after a player makes a move in a blitz game, for example. Bronstein Delay I'm less familiar with, but it can be Googled I guess.

Yes, having blitz games (the usual choice of chess server players, though they can use longer time controls on at least some servers, as an option, say for a tournament) might be a serious problem if there is no way to ensure both players are online and on Game Courier when their game begins.

edit: A Google blurb:

Bronstein Timing is similar to Fischer Timing. Players have some time at the beginning of game and then after each move they receive a bonus of a few seconds (let's assume 3 seconds for now). The difference is that if you use less then 3 seconds (the bonus), you receive back only the time you used for your move.

edit2: I think 'receive back etc.' is a poor phrase for it. Instead it should just say you receive no bonus on that turn whatsoever, if you use less than 3 seconds before making your move.


Kevin Pacey wrote on Thu, Feb 8 11:12 PM UTC in reply to Kevin Pacey from 10:29 PM:

I've edited my post I'm replying to, in case anyone missed it.


H. G. Muller wrote on Fri, Feb 9 06:38 AM UTC in reply to Kevin Pacey from Thu Feb 8 10:29 PM:

I am not sure what you mean by that last sentence. If you would think 2 sec in a 3 sec Bronstein game, your clock will not advance. So I would say you did receive a bonus of 2 sec in that case.

In Bronstein TC the clocks just start running after a delay, each move.


Kevin Pacey wrote on Fri, Feb 9 10:56 AM UTC in reply to H. G. Muller from 06:38 AM:

No H.G., I think I am right about my interpretation (which is just making sense of Google's poor wording). Otherwise there would be no difference essentially between Fischer Increments and Bronstein Delay. Long ago I played both at my chess club, before I stopped going there, and vaguely recall the difference. Maybe you can find a link that supports your viewpoint(?) [edit: apologies H.G., it seems you are correct, says the wiki - I missed the shrinkage of the bonus effect that happens at times:]

https://en.wikipedia.org/wiki/Chess_clock


🕸Fergus Duniho wrote on Fri, Feb 9 11:54 AM UTC:

Although I can decrease the delay in checking whether one’s opponent has moved, the server-side nature of Game Courier still slows it down. Ideally, fast real-time games could be handled by a JavaScript interface that works without needing to reload the whole page. But because rule enforcement is handled with a server-side language, JavaScript could not handle that part. But with better modularization of the code, it might be able to pass a move through the script with just the rule enforcement code and change some JavaScript variables. But there is still the matter that when someone makes an illegal move, Game Courier normally handles this with the die command, which exits the script with a printed message. I can see what I can do, but before I begin on this, I want to complete its support for navigating through past moves with JavaScript. This feature is currently not available for hexagonal, circular, and spherical games, but I can fix this by adding css rendering methods for them that place individual piece images on generated empty boards.


17 comments displayed

Earlier Reverse Order Later

Permalink to the exact comments currently displayed.