Older blog entries for Bram (starting at number 59)

Mitch Kapor at CodeCon

We've got Mitch Kapor as our special guest speaker for the benefit dinner at CodeCon!

CodeCon pre-registration closes in just a few hours, and it will cost more at the door, so if you haven't yet, go sign up now.

13 Feb 2003 (updated 14 Feb 2003 at 06:59 UTC) »
CodeCon

Pre-registration for CodeCon is closing on the 15th. After that you'll have to register at the door, which costs more.

We've got a really great program this year, including a presentation on Advogato. Also notable is a panel on version control, including representatives from BitKeeper, Subversion, and OpenCM.

CodeCon is also one of the the best places to meet all the local bay area open source hackers. I hope to see everyone there.

I've created a Codeville tutorial which demonstrates its cool features quite succinctly.

Secret Project Unveiled

My secret project is now ready to be seen. It's Codeville, a version control system.

One thing no longer nagging at my brain, a million left to go.

A public service announcement

Space travel is dangerous. Please take careful consideration before boarding any space vehicles.

Abstract Games Magazine is having a game design competition with the theme of simultaneous play. I came up with the following a little while ago, but it turns out I missed the submission deadline, so I'm posting it here.

The game is called Straights and Queers (this potentially uncomfortable metaphor is by far the easiest way of understanding the rules, so please bear with me). It's played on the following board -

       __
    __/  \__
 __/  \__/  \__
/  \__/  \__/  \
\__/  \__/  \__/
/  \__/  \__/  \
\__/  \__/  \__/
/  \__/  \__/  \
\__/  \__/  \__/
   \__/  \__/
      \__/

There are two players, the straights and the queers. Each player has two pieces, one male and one female.

On each turn, both players move any pieces they have on the board and place any pieces which aren't on the board, which happens at the beginning of the game or when a piece is captured. Both players write down what their moves are without seeing the opponent's moves, then reveal what the moves are.

Pieces move to any adjacent hexagon. They cannot move to a hexagon which another piece is currently on, even if that piece belongs to the same side. Pieces can be placed on any empty hexagon, but may not be placed onto a hexagon currently containing a piece, even one belonging to the same side. A player may not move both of their pieces onto the same hexagon. When pieces move, they must move to a different spot, they cannot remain in the same place.

If two pieces wind up on the same spot, then a capture happens. If the two pieces are of the same gender, then the queer captures, otherwise the straight captures. Additionally, in the rare case where a piece winds up in a corner with all three adjacent spots occupied and hence no legal turn on the next move, then it is captured. Captured pieces are returned to the side they belong to be placed on the next move.

The first player to perform ten captures wins.

That's all the rules. I think this game is made interesting by the simultaneous play despite the extraordinarily small board. Perhaps I went too far on board smallness, rendering the game brute forceable. However, there are classic games involving some chance and secret information which have extremely simple positions, such as yachtzee and texas hold'em poker. Texas hold'em turns out to be completely out of brute force range, but yachtzee is emminently brute forceable in the solo case optimizing average score, and on paper looks just barely solveable for the two-player case trying to optimize chances of winning.

I'm curious to see if the entrants into the sumultaneous play competition generally have very limited board positions. I'd also like to know if anyone has actually set about brute forcing yachtzee. If anyone knows of such efforts please tell me.

The CodeCon 2003 program is now announced, and registration is open.

CodeCon 2003 will be February 22-24, noon-6pm at Club NV in San Francisco, California.

All presentations will given by one of the active developers, and accompanied by a functional demo.

CodeCon presentations will include:

  • Advogato - Good metadata, even when under attack, based on a trust metric

  • Alluvium - p2p media streaming for low-bandwidth broadcasters

  • Bayonne - Telephony application services for freely licensed operating systems

  • Cryptopy - pure Python crypto

  • DeepGreen - Agent Oriented investment analysis designed to be self-funding

  • GNU radio - Hacking the RF Spectrum with Free Software and Hardware

  • HOTorNOT - People submit their picture for others to rate from 1 to 10

  • Hydan - Steganographically conceal a message into an executable application

  • Khashmir - A distributed hash table library upon which applications can be built

  • Mixminion - A next-generation anonymous remailer

  • Neurogrid - Decentralized Fuzzy Meta-Data Search

  • OpenRatings - An open source professor ratings engine

  • Paketto Keiretsu - Interesting and Useful Techniques for TCP/IP Networking

  • YouServ - A communal web-hosting system for the masses

  • A panel on version control

Lance Fortnow links to some interesting commentary on P vs. NP.

I found out today the standard terminology for my favorite conjecture -

Conjecture: The circuit complexity of the k-sum problem is at least n ** k.

This directly implies that the 4SUM problem is quadratic. I believe that the 3SUM problem is also quadratic (as does everybody), but that the reasons for that are much deeper and more complex.

It also implies P != NP, so we can rest assured that noone has proven it yet.

Given the obviousness of this conjecture I'm sure circuit complexity people have already spent considerable time on it and just haven't gotten anywhere, although curiously googling for '3sum' and 'circuit complexity' doesn't turn up anyone else musing on the subject.

Here is a good list of open problems to spend time on.

I haven't posted much lately because I've been spending my blogging time on my new secret project, which will be unveiled when I get a proof of concept working.

XML people don't seem to realize that when I say I'm on the de facto standards committee I'm not kidding.

50 older entries...

New Advogato Features

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!