Getting tastier each minute, 0.1.6 will be out later today when I finish implementing examination mode.
If I don't get a Un*x-related job soon I'll grab my guitar and start playing it for money. You _don't_ want me to do that, the music scene will never recover from the damage. (I am known for having performed the Free Software Song worse than RMS -- the guitar performance was decent but I sing much worse than he)
1. The only tarball, for 5.02, is 30 MB long. There are no older versions at the site. I look to the 33.6 kbps modem. It looks to me.
...4 hours later...
2. uncompressed the thing, as expected there is a 103 MB PGN file for the book. Those who can't write chess engines add huge game databases. No documentation, Makefile or configure script in the toplevel directory.
3. Entered the 'doc' dir. So I must go to the src directory, run autoconf, configure and make. Sounds right.
4. When I try to compile I get an error from GCC about trying to use sp as general register for register variables. Peek the code. Peek the Makefile. Removed the -O3 switch. Ok, it compiled this time.
5. I was almost su'ing and 'make install' when I had a bad feeling. A voice says 'check the Makefile before'. This is the install target:
cp gnuchess.exe /cygdrive/c/winboard/gnuchess.exe
GNU Chess assumes you are running Microsoft Windows. (scream) How did this crap got to bear the GNU on its name ????????
6. The doc file said I should build the binary book file from the 100-MB PGN file. I follow the exact instructions and... nothing happens. No file is written, no error message, no crash, no nothing. It just does nothing. The 100 MB file that made the whole thing so slow to download is completely useless (I can use it for Crafty, but it's useless for GNU Chess unless it builds the damn binary file)
Well, GNU Chess (in fact, any engine that talks the xboard protocol version 2) support is added to eboard, the changes are already on CVS. Wrote a harsh document about the whole experience, also on CVS by now.
I must now look for older versions of GNU Chess, maybe they suck less. I must also start looking for GPL'd (or at least free in the Debian Free Software Guidelines sense) engines. If you know of any, drop me a note.
Similar comments apply to UML and Rational Rose enforcement. If one can't picture the structure of the software by himself without the aid of a diagram-drawing, code skeleton generation program (pencil and paper is ok) or agree on extensions by talking with the workmates, he shouldn't be developing software in the first place.
wrote a tutorial on the usage of automake and autoconf, should be published on an "official SEUL URL" soon (as soon as Roger comes back from the 6th dimension), one of the copies I sent for review is here: autotut
Since last entry, released gBib 0.1.1, a Gnome BibTeX editor, quite a nice program (the main developer is Alejandro Sierra), had some nice fun hacking the bibfile parser.
should have spent some time playing around with chessd
meanwhile, in a job interview last wednesday...
...we work with the latest and greatest of technologies: Microsoft's leading line of products, such as ASP, IIS, SQL Server and we also are a major player with open source technology, using Java, JSP and Oracle extensively...
That's self-control: I managed not to fall down laughing when the guy called Java and Oracle open-source.
While still upset at sgmltools' lack of documentation, I'm more upset with pkgtool failing to tell me "this tgz file is incomplete".
BTW, GNU Chess 5.0's tarball is 30 MB and GNU removed all the previous versions of GNU chess from the site...ack! Better than getting a chess engine to win out every GM in the world, a chess engine programmer's challenge should be building efficient/"smart" search algorithms - searching the good positions and discarding fast the bad ones. Including a (required) 20 MB book (I haven't downloaded the beast to check the size yet, but I can only presume the huge size is due to a huge game book) looks dumb - what people want when playing chess with a computer ? That the computer reproduces every move from a GM game held in 1987 it has in the database or to challenge the CPU's transistors to outthink him ?
Fixed two nasty bugs in the rgpsp protocol (the remote gPS polling), version 0.10.2 is out.
I spent the weekend coding eboard, after some weird GtkText behaviour I spent the last night modifying xchat's xtext widget to display ICS output -- a little line wrapping oddity still remains but it looks good. This week I'll hopefully focus on writing the chess-related code and have an interface capable of playing a regular chess game on FICS by friday. :)
Someone posted an entry recently about getting the Crafty chess engine to run in a MOSIX cluster and was searching for strong players to try with. The FICS (Free Internet Chess Server) provides computer accounts, so that you can login with a chess program and get it pplaying with either other computers or human players. Notice 1: one of the fics boxes (particularly the one that is the web server) is down again this week, the link above will probably only work again by the end of the week or even next week. Notice 2: the Free in FICS is for free beer, not free as in freedom. Notice 3: to reach FICS while the web server is down, telnet to freechess.org, port 5000, login as guest user.
Meanwhile... still looking for a job, preferrably one I don't have to see Windows again.
New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.
Keep up with the latest Advogato features by reading the Advogato status blog.
If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!