Author: Roberto Waldteufel
Date: 03:32:54 11/11/98
Go up one level in this thread
On November 11, 1998 at 02:47:21, Frank Schneider wrote: >On November 11, 1998 at 01:30:24, Oliver Y. wrote: > >>Here's a quote from PCWorld's report on the convention at Santa Clara, CA: >> >>'Moritz and Dyson assert that the Linux operating system could prove to be one >>of the most important innovations of all as "hordes of programmers" work to >>develop applications for the so-called open source operating system, which is >>more accessible than Microsoft Windows since its source code is freely >>available. "Open source turns your customers into your developers," says Dyson. >>"It's a fundamental change."' >Recently two internal papers from Microsoft were leaked to the internet. They >are called the halloween papers. It seems Microsoft thinks that Linux is a >serious competitor in the server market. See >http://www.tuxedo.org/~esr/halloween.html >http://www.opensource.org/halloween2.html > >But now for chess-related thoughts > >>How 'difficult' would it be to have competitive software appearing in Linux? >I think it is quite easy to port engines. Many good engines are already >available on unix-platforms or started there. This includes some of the >engines that are now commercial. > >>When, in a meaningful way? (not including those that are platform independent >>already) > >The problem is the userinterface. It is not easy to develop portable GUIs >and developing a GUI for unix-machines only is questionable since most >users use windows. Porting GUIs from windows to unix is very hard (and that >is intended by microsoft). > >I see some possibilities for getting good GUIs under Linux: >a) There is already XBoard which cooperates with many engines. Possibly someone > develops something similar to XBoard that looks more like commercial > programs. >b) There are Linux projects that try to offer the Windows32-api on Linux (Wine). > *IF* they become better porting a program to Linux would become easier. > About a year ago I tried to run Gromit1.2 under Linux (using a early > version of Wine) and it worked. Not perfectly, but one could play chess. >c) Java may become the language of choice to write GUIs. A commercial vendor > could develop a Java-GUI with pluggable engines (a protocol like XBoard > or 'Chessbase'). The GUI would be portable because of Java and porting > a (C/C++) engine is not too difficult. >d) Today 90% (my estimation, what do others suggest?) of the programmers that > write free chessprograms (including myself) concentrate on engines. Some > of them may become bored one day and start writing GUIs and databases. > As machines become faster the need to have the best engine is decreasing for > users who want to play chess themselfes. >e) The number of Linux-users increases very fast (almost doubling every year). > When there are enough users there will be a market that is interesting for > commercial developers. > >IMHO c) is most likely, probably supported by b) and e). Some time ago I had >an idea how to support d) and I published a 'paper' >(see http://home.t-online.de/home/hobblefrank/index.htm and there > http://home.t-online.de/home/hobblefrank/cidleng.htm) >but there was only a little response. Anyone interested now? > > >Frank Hi Frank, Option (b) comes as something of a revelation to me! Do you suppose that 32-bit Windows Console Applications (ie "DOS box" programs with text-only screen) would run readily under linux using this technique? I ask because my program is precisely that, a console application. Although it is compiled as a 32-bit Windows application, it uses no fancy graphics, no mouse support, and runs in full screen mode (ie no window) with a 50 line x 80 character display. Strangely enough, I manage to display a reasonable chess diagram in this restricted medium (not as good as Gromit's nice GUI though). I would hope that this would make my program an ideal application for WINE, since it uses only a simple subset of the Windows API, only requiring the API for screen, keyboard and file I/O. I wrote the program with PBCC, which is specifically a console compiler for 32-bit Windows. The compiler manufacturers are intending to develop a version of the same compiler for linux, but I have no idea how long this will take. My general dissatisfaction with Windows has lead me to await the arrival of a linnux version of the compiler with considerable interest, but I have held off trying linnux so far because I thought that niether my program nor my compiler could run without the Windows API. If I was mistaken in this impression, I think I would like to give linnux a try without waiting for the linux version of the compiler. I hear linux is very stable :-) I only have to look sideways at Windows95 and it crashes.....:-( Best wishes, Roberto
This page took 0.01 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.