Searching OPML 

During Dave's talk in Toronto (Thanks Dave!) he talked about people building OPML search engines. Then it dawned on me the other day that iPodderX has an OPML search engine! We spider through the ipodder.org OPML directory and cache it in our database so it can be searched by a user to find what they want. We even go a step further and cache what the OPML directory is pointing to. Pretty cool huh?

Bandwidth Solutiosn 

Tom Morris responds to Adam Curry's cry for ideas to solve the podcast bandwidth issues, and one idea that struck me as interesting was this one:

How about this? Have a whole bunch of RSS feeds. Every time someone looks for your RSS feed, they get a different feed - so, for instance adamcurry01.xml, adamcurry02.xml, adamcurry03.xml etc. There would be a smalll time difference between each file. So, for instance, if 01.xml publishes at 1pm, then 02.xml publishes at 1.30pm. The time variation could be varied by the bandwidth used. That way, the difference between n.xml and n+1.xml is dependent on how much bandwidth is used. Perhaps each n.xml would be mapped to a different server if you wanted to reduce the difference between n.xml and n+1.xml. This way, each podcatcher picks up the file over a longer period.

I think at one point Doug Kaye was looking into something similar, not sure why it didn't happen.. Maybe one issue with the whole idea is that the directories are only going to have adamcurry01.xml and not support the dynamic linking to the other time-shifted RSS feeds. Maybe that is the reason Doug didn't follow through. Anyways its an interesting thought that might need to be looked at a bit more closely.

Categories 

I think I need to start to use the categories feature... Things are starting to get a tad messy around here.

changes.opml.org 

I'm still trying to get changes.opml.org to show up on my right sidebar and its not dynamically refreshing :( Looks like Dave is going to do the same thing on his blog, maybe he'll let us know how he did it :(