Complete.Org: Mailing Lists: Archives: freeciv-dev: September 2002:
[Freeciv-Dev] Re: new connect dialog (ver 2) (PR#1911)
Home

[Freeciv-Dev] Re: new connect dialog (ver 2) (PR#1911)

[Top] [All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index] [Thread Index]
To: bugs@xxxxxxxxxxxxxxxxxxx
Cc: Freeciv Developers <freeciv-dev@xxxxxxxxxxx>
Subject: [Freeciv-Dev] Re: new connect dialog (ver 2) (PR#1911)
From: Christian Knoke <chrisk@xxxxxxxx>
Date: Mon, 2 Sep 2002 20:54:15 +0200

On Sun, Sep 01, 2002 at 08:59:42AM -0700, Mike Kaufman wrote:
> 
> attached is version 2 of the new connect dialog.
> 
> Changes:
> 
> fixed ALLOW_NONE typo in civserver.c
> 
> mem.c: char *dest is declared before str==NULL check.
> 
> don't close stdout and stderr on the fork anymore
> 
> translated skill level
> 
> Save Game... -> Save Game as...
> 
> Disconnect -> Disconnect from Game
> 
> Quit -> Quit Freeciv
> 
> Start multiplayer... -> Join multiplayer...
> 
> increased length of array holding "Human"/"AI"
> 
> remove xaw connect dialog since it was buggy.
> 
> added Christian's waitpid patch. (Reinier, could you compile this on Sun
> and see if all is well?)
> 
> remove ALLOW_HACK requirement from the packet_player_list_request

This works very good.

Two small problems still: 

- When clicking "Start saved game" and "Next", and then "Previous",
  the file load box stays open.
  This might (probably will) confuse some newbies. You load a game
  and nothing happens.

- In some rare cases, the civserver doesn't get killed, and leaves
  the starting terminal in a strange mood. You get a shell prompt
  *and* civservers output mixed. Especially happens, when you end the 
  civclient by clicking on the "close" X window corner.

> Non-changes:
> 
> didn't worry too much about Christians ALLOW_HACK concern since anytime 
> the server is started from the client, the client immediately connects.

Maybe you should. Just do

#define AUTOCONNECT_INTERVAL            10
#define MAX_AUTOCONNECT_ATTEMPTS        10000

and you're in.

> didn't kill off server on disconnect (want other people's opinion on
> this)

Well, this can be discussed. I would say, kill the server. Just warn the
user that this stops the game also for the remote players. Can you detect
this (remote players) in an easy way?

> it might have some bearing on the HACK problem though which is a pro for
> doing it.

It has. The remote player just types /firstlevel to get hack access.

A comfortable gui solution for the disconnect case can be found later.
For now, restricting the -c option to *local* connections should be
sufficient AFAICT.

> didn't add Daniel's flags since they required that rulesets be sent to 
> on the client on game load which is somewhat problematic.
> 
> still has the gtk_notebook tabs issue on GTK2 (vasc, help!)
> 
> 
> Comments welcome.
> 
> -mike

Christian

-- 
Christian Knoke     * * *      http://www.enter.de/~c.knoke/
* * * * * * * * *  Ceterum censeo Microsoft esse dividendum.


[Prev in Thread] Current Thread [Next in Thread]