I've gotten pyrex to compile a wrapper around wmatch though I need to figure out how to return a numeric array. Once I've got that I can dump the command line wrapper and use the python extension module.

Though I'm still wondering how we can break free of the JPL code and have something that's fully open source.

Other than that I do need to finish getting it ready for OS X

I hope we win the apple bioinformatics cluster tomorrow. Though I also hope that if we're featured on the apple web page we don't get too badly slashdotted. Our pages are pretty simple so hopefully it'll be okay.

I'd like to be able to update blog entries with meta data like FOAF entries and todo entries. The example http://www.w3.org/2004/lambda/Sites/index.html has some suggestions of how to embed RDF metadata in html.

I'd also like to get some photo viewier working correctly (with rdf enhancements of course)

last question is how to divvy up personal entries, work entries, and private entries.

• programming async tile asembly model
• can you use tiles sets to building square?
• start with N+4 tiles
• using binary counter use 2^k tiles
• resulting square could be used to define another large square @ 2^2^k
• design looks like a demultiplexer circuit, if we could attach and gates we could build a memory addressing structure
• error correcting reversable tileselts
• what rate do tiles connect and disconnect? (rate reaction)
• Gmc concentration of tiles
• growing r_r=k[tile]=ke^-Gmc
• Gse defines the free energy of a strength 1 bond
• r_r,b = ke^-(b Gse)
• reduce error rate by lowering concentratoin.
• error correction
• use error correcting tiles - complex
• go slower (error rate proportional to sq rt - slow
• invent new tile technology - hard
• combine techniques
• use larger tile set that encodes redundent information
• if it makes an error it slows doesn giving time for the reverse rate to knock a title off.
• using this redundant tile set doubles the error slope of, giving a 10^4 speedup
• back of envolope calculation shows that increasing tileset, changes the rate by sq rt, though simulations show that it improved but didn't increase at expected rate
• better 3x3 tile set uses an algorithm using strong, standard and null bonds to define a "snake-like" path through the error correcting block
• using this technique one can define a self-repairing tileset

look on winfree's site for xgrow