2009/6/23 Christian Nentwich <christ...@modeltwozero.com> > Hi Don, > > thank you, that is very useful. Definitely food for thought. I am probably > going to have time to take a look at this tomorrow. If you have a developer > copy of the server running on a private port, and can mail me that off-list, > I'd like to give it a try. >
I'm in the middle of implementing the t2 stuff, it currently doesn't work. When I have it working I will set up a test server on CGOS - not for general use quite yet but enough to test your client. > > > A few questions: > - What is a good way of referring to this new version of CGOS? CGOS 2? > Haven't given that any thought. Let's go with 'CGOS 2' for now. > > - What happens if the old server is sent "e2"? Does it drop the > connection? > d...@greencheeks:~$ telnet cgos.boardspace.net 6867 Trying 67.202.75.88... Connected to cgos.boardspace.net. Escape character is '^]'. protocol e2 joe Error: invalid response Connection closed by foreign host. > > - Probably more of a user question, but what is the default venue > exactly? You said 9x9, I guess that's 9-300-7.5, as it is now? > Yes, that is my plan. > > > And unrelated: > - I've sometimes wanted to see who is currently playing on the server. > Will this be possible (e.g. through a web pages that gets updated every few > minutes)? > The web page will be in PHP, so when you refresh the web page you will get the exact and current state of the server. In other words if I log on and 2 seconds later you hit the site, you will see that I am logged on. I am also considering the idea of sending info message to every client when someone logs on or off and perhaps even to broadcast the result and pairings of each game. What do you think of that? If I do that, then a sophisticated enough client could track the state of the server without requiring the user to constantly refer to a separate web page. Of course this does not change the protocol since "info" is already defined as part of the protocol, although to be most useful I would have to standardize the exact format of those kinds of messages. - Don > > > Christian > > > On 23/06/2009 14:29, Don Dailey wrote: > > Christian (and anyone else interested) in the new CGOS protocol: > > There are 2 minor change to the protocol. I'm hoping to do a test drive > today as the new server is functional, though not 100% complete. > > Here are the two changes: > > First of all, when the server asks for the protocol, you can continue to > use the "e1" protocol and it will default to a default venue (which will be > 9x9 for now.) In this way the old clients will continue to work. > > But if you specify the new 'e2' protocol, such as: 'e2 cgosPython 0.2 > beta' you will next get something like this: > > venues 9-300-7.5 11-420-7.5 13-540-7.5 15-660-7.5 17-780-7.5 19-900-7.5 > > This is a list of available venues and could be anything the server is > configured for. Each venue is separated by a space. A hyphen separates > the 3 elements within a venue which are boardsize, time in seconds, and > komi. So 15-660-7.5 is: > > boardsize: 15x15 > time control: 11:00 (11 minutes, zero seconds.) > komi: 7.5 > > You may want to provide some mechanism for informing the user of the > software which venues are available, even if it is as simple as just > displaying it. If the user tries to configure a venue that is not > available the connection will abort. You can also respond with a 'quit' > if the client notices that the user did not specify an available venue. > Of course all of that can be handled any way the client sees fit to do or > you can just ignore the issue and require the user to configure a proper > venue (the available venues will be posted on the web page of course.) > > The proper response is ONE of those protocols, such as "13-540-7.5" > > After this, everything remains the same as in the old protocol and the > server will next get the username and password as usual. > > The second (optional) change is this: > > If you respond to the protocol message with something like this: > > "t2 cgosNoGo 0.1 alpha - Java engine client for NoGo" > > The client will be placed in test mode. Nothing else changes about the > protocol, it is the same as "e2" but this triggers the server to play test > games with the client which will not be rated or recorded. You will then > recieve the venues message which you must respond to just as in the e2 > protocol. > > Another minor change, which does not directly affect the protocol, is that > the server will abort any connection that seems unduly delayed. For > instance if the client does not respond to messages after a few seconds > (such as gameover or venues) then the server will abort the connection. > This is to avoid a needless buildup of bogus connections. > > > > FYI here is a list of clients GGOS has seen for the e1 protocol. I'm > leaving out the clients I have provided: > > e1 CGOS soket engine client 1.3 windows7 by WangManDong > e1 CGOS pike client 0.9 by Gunnar Farneback > e1 CGOS tcl engine client 1.2 [platform] by WangManDong > e1 MROZTEST 1 > e1 NaruGoTest > e1 Tesuji Software CGOS Client > e1 cgosNoGo 0.1 alpha - Java engine client for NoGo > e1 cgosNoGo 0.1 alpha - Java engine client for Nogo > e1 cgosPython 0.1 beta > e1 cgosPython 0.2 beta > e1 cgosPython 0.3.0 beta > e1 joe > e1 myCtest > e1 test client > e1 testing > v1 "0.1 NaruGo client for CGOS." > v1 NaruGoTest > > - Don > > > > > > On Tue, Jun 23, 2009 at 8:20 AM, Christian Nentwich < > christ...@modeltwozero.com> wrote: > >> All, >> >> FYI, the Python version of the CGOS client has moved into the main CGOS >> sourceforge area: >> >> http://cgos.sourceforge.net/client-python/ >> >> There is also a new release (0.3.0) out that supports multiple engines. >> >> Christian >> _______________________________________________ >> computer-go mailing list >> computer-go@computer-go.org >> http://www.computer-go.org/mailman/listinfo/computer-go/ >> > > ------------------------------ > > _______________________________________________ > computer-go mailing > listcomputer...@computer-go.orghttp://www.computer-go.org/mailman/listinfo/computer-go/ > > > > _______________________________________________ > computer-go mailing list > computer-go@computer-go.org > http://www.computer-go.org/mailman/listinfo/computer-go/ >
_______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
