find www -follow -type feed -print | sort
find process seeds itself. When I read something I like I follow promising links; and if I do like what I then find, I click the feed icon in the browser address bar, and my preferences arrange for the feed to be filed by my news reader.
So I’m recursively walking a massive directed graph. The further I explore, the more there is to be explored. I haven’t yet added the same node twice (I do at least remember where I’ve been and what I’ve already added) but it’s only a matter of time (and now I think about it, I have added planets which syndicate feeds I’ve already subscribed to).
Should I tire of this auto-discovery, my feed reader will happily, and with unnerving accuracy, offer its own recommendations.
sort isn’t too bad. My reader supports various filing and filtering and viewing operations.
What I’m starting to want, though, is
find www -type feed -follow -print | sort | uniq
Last week half a dozen feeds I subscribe to reported that Python3.0a2 had been released; none added much news to that headline. Similarly lots of people told me about the google chart API launch — though in in this case, some had something original to report. It’s fuzzy
uniq I need, a filter which kills hollow echoes and keeps thoughtful commentary.
My feed reader shows me what’s been updated. It can’t show me what’s new.