(I'm rather new to bitzi, so please excuse me if this is a FAQ or suchlike.) It occours to me that it might be useful, in many curcimstances, to be able to lookup (via URL) on data other then the full SHA1 or bitprint (sha1+tigertree). For example, when searching for video or audio files, it would be useful to search on only the data and not the internal metadata, which often changes, and thus produces false multiple tickets. Also, it would be useful to search on, say, the sha1 of the first n seconds, so you can find the bitprint of the full versions of partial files, and thus (hopefuly) the rest of the file. Also, it would be helpful to be able to search on the sha1 of the straight (rasterized) image data of a variety of formats. This would allow the tickets for the correct data to be found, even if you have a PNG and the original poster had a BMP of the same image. Basicly, it would be nice if you could use the lookup/urn:>>foo<< for more values of foo (or, if you already can, for all allowable values of "foo" to be listed).

You can't currently use lookup/urn:>>foo<<, but we do have plans along those lines. For example, we may allow lookups via ed2k and/or FastTrack hashes.

I suspect that rather than hashing the first n seconds (or bytes) of a file we'll someday start collecting more tiger tree values (currently we only collect the top of the tree, that's the second part of a bitprint).

I'd love to have a way to identify clusters of images (and other media) from the same source, via your "sha1 of the straight (rasterized) image data of a variety of formats" or other mechanisms. Hackers are more than welcome to add such features to the appropriate bitcollider plugin. :)