I'll plead ignorance on this, but I've always thought that the torrent approach would be useful for discovering microformats... I mean, you could sort of have seeded mF data that would point to clusters of paged marked up with mFs... Sort of like massive, distributed databases.
I don't know what the mechanisms would be or what this would really mean, but rather than have an explicit "contain-microformats" link system, you could point to a warm of sites that are likely to contain data related to various types of data... events, address, addresses, etc. I dunno, maybe the cranberry sauce is getting to me. ;) Chris On 11/24/05, Brian Suda <[EMAIL PROTECTED]> wrote: > i just want to re-itterate that i'm not trying to design a 'discovery > format' i'm just trying to find examples of others that have already > come across this problem and how they have attempted to solve it. > > in my previous Avon example, i was not suggesting that the 'discovery > file' re-listed all the 40,000 microformats. Instead (similar to the > sitemap.xml) it simply listed the pages that contained microformats, > and what formats were encoded. Therefore, it took some of the weight > off the spider to determine these things and head down paths that had > no data. Instead all it had to do was collect one file, and queue up > just those listed and possibly even distribute that load to > specialized spider-decoders for each microformat type. > > It is not ideal, it is just an example of how other systems do discovery. > > -brian > > -- > brian suda > http://suda.co.uk > _______________________________________________ > microformats-discuss mailing list > [email protected] > http://microformats.org/mailman/listinfo/microformats-discuss > _______________________________________________ microformats-discuss mailing list [email protected] http://microformats.org/mailman/listinfo/microformats-discuss
