I've been fooling around in some shady things. Like clit. Ahem. Convert Lit. I've also been looking at OeB - a 'standard' for the ebook industry.
Rubbishy standard. Everyone takes an OEB and them DRMs it to SFA. (Translation: OeB isn't used as plain OeB, making me think what is the point?)
Book people just don't get metadata. Librarians kind of do. But book makers, particularly ebook makers, don't really. ISBN? Use it as a what? International identifying number for my work? Why ever should I do that! BAH!
I'll use a GUID. Globally Unique Identifier. Yes, that's it.
Luckily, Amazon provides fixes for this; title based lookup for ISBN.
To that end, I've begun some work. On a Bookshelf. It's in C#, my first decent C# undertaking. There's nothing much at the moment to it, just an OEB Serializer and Unserializer class. Work enough in itself!
Browse the CVS if you care to.
On a cooler note, I found generation5.org. Generation5 is all about AI, and I found it while googling for LSystems and Virtual Plants. I was wowed at how easy it was to do much of this. The math isn't hard. Neural Nets are fairly simple to construct, and training seems the hardest part. The coolest-thing-ever was the article on ISBN reading from scanned barcodes via webcam. I'd only ever seen this mentioned before in relation to an Apple application. That is definately on the list of cool things to hack. I'm too lazy to find the link for either.
Geek, C#, Books, E-Books