Wherein I muse about Google Reader past, and what it might have been. And, wherein I describe what I hope springs up in the aftermath of its closing.
Reader’s been long gone already
So, they’re finally shutting down Google Reader, huh? It’s sad, but unsurprising. It’s obvious they didn’t really have a strategic place for it in the Google+ universe, and it just was just neglected since the Sharepocalypse. Kill off the key social synergy of hosting a centralized news reader, and it’s no wonder you’ll see usage decline.
I’ve been running Tiny Tiny RSS on my own server since Google killed off in-product sharing. So, I won’t be too terribly affected by the shutdown personally. I think my wife still uses Google Reader, having moved after the demise of Bloglines. If she likes the looks of TT-RSS, I’ll set her up with an account too.
But, Marco and Dave have it right: This will probably be a good thing for RSS. The problem has been that Google Reader was just good enough to lull me out of scratching my own itch. This is coming from the guy who wrote a 600 page book on RSS and Atom out of love for the tech. So, I’m sure I won’t be the only one poking around code archives and blowing dust off old repositories.
Social novelty filtering
So, this is the itch I’d love to scratch in the post-Google-Reader age: Reader-before-Plus offered social novelty filtering. That is, fast sharing within the product fueled reciprocal feeds of novelty, filtered by my “friends”, presented in the same news reading interface as my other bazillion feeds from the web at large.
Yeah, yeah, I know: You can “share” to Facebook, to Twitter, even to Google+—that’s why Google killed in-product sharing on Reader, after all. I do that on occasion, and that’s how a lot of people get their streams of novelty. But, it’s nowhere near the same thing, neither in quality nor in quantity.
The thing about these “friends” on Google Reader was that we never interacted directly. It was refreshing, it was great. No small talk, no conflicts, no getting in each others’ virtual faces—we just mutually harnessed slices of each others’ minds to build intelligent streams of novelty.
That might sound cold or mechanical or exploitative—but the thing is, there are plenty of other outlets for interpersonal exchange, and I even met up with some of my Reader “friends” out-of-band there too. But, those channels are about you and me, we’re the objects of interest & the stars of the show. There are very few channels that are about shared novelty as the object of interest, where you and I can get ourselves out of the way and conspire to surface cool things.
(And, of those channels that do exist—Pinterest, for example, maybe Tumblr—I’m not a big fan of the UI vs ye olde Reader. Still, magic & strange loops can emerge from recursion & re-entrant flows; your mileage may vary. See also: Google’s Lost Social Network.)
So anyway, I’m probably going to play around with the machinery of feed aggregation again. But, one of the things I really would like to see as a thing out there is distributed social novelty filtering. How do you do that? Well, the first half of it is pretty simple:
- In feed readers, implement a one-click “share” button on every item. Maybe offer an optional field for comment.
- Offer a public feed of every “shared” item, easily discoverable from a public profile.
For example, my installation of TT-RSS offers a feed of my “shared” items—albeit not in a very discoverable way. I also funnel craptons of material into my pinboard.in feed, but mainly for personal archiving and search. Even Pinterest has a feed for me, though I barely use it.
The other half is where some innovation could stand to happen: The obvious thing is to just get yourself a new RSS reader and subscribe to all the shared-item feeds of your “friends”.
But, there’s a lot more to be done here:
- discover and follow “friends” across services,
- de-duplicate shared items by URL,
- rank items by counting shares,
- construct discussion threads from shares
Go nuts, please! Someone form a startup and take my money to solve this, so I don’t have to. Even better, release some open source so I can host it myself and maybe even contribute some code. I’m thinking about doing some of this, but my lack of sustained attention span for projects is well documented.
But, the important thing here is that it’s distributed and a thing that’s conventionally done out there on the web. This shouldn’t be constrained to a single vendor’s silo ever again, because that allowed a single vendor to kill it dead and I miss it terribly.