Picking a Google Reader Replacement (With Two Weeks to Spare)

Picking a Google Reader Replacement (With Two Weeks to Spare)

Tutorials feedly google-reader newsblur rss

So I finally got around to doing the thing I've been putting off since March: figuring out where my RSS feeds are going to live after Google kills Reader. July 1st. That's the date. As of tonight I've got fifteen days left and a folder of 143 feeds that I am, apparently, incapable of just letting go of.

I know, I know, everybody and their cousin has written a "Google Reader is dying, here's what to use" post already. I'm not trying to do a roundup here. I just want to write down what actually happened when I tried to move my stuff, because most of what I read before doing this was either way too rosy or written by someone who has twelve feeds instead of a hundred and forty.

Step one: export your OPML before you do anything else. Reader Settings > Import/Export > Export your subscriptions. It spits out a single XML file with everything in it. Do this now even if you haven't picked a replacement yet, because if Google decides to pull the plug early (they've done weirder things) you don't want to be the person who lost five years of carefully curated feeds because you were "going to get to it."

I tried three services this weekend.

Feedly was the one I'd already half-committed to because they made a lot of noise back in March about building a "Normandy" backend specifically to catch Reader refugees, and it shows. Importing the OPML took maybe four seconds. Feeds started populating almost instantly, unread counts and all. The browser extension is fine. The mobile app is actually kind of nice, though it took me an embarrassingly long time to find the button that lets you go back to a plain list view instead of their magazine-tile thing, which I hate. I don't want cards. I want a list of headlines I can blow through in ten minutes with my coffee. Give me the list view by default and I'll be happier with basically everything else about the product.

NewsBlur was more painful. The import went through fine but then it sat there "training" or "processing" or whatever they call it for something like twenty minutes on the free tier, because free accounts get throttled behind paying ones during a fetch queue. Which, fair, I get why, but when you're trying to evaluate three services back to back in one evening it's annoying to sit around waiting on one of them. Once it finished, the actual reading interface is genuinely nice, and I like that it does some intelligence around training which stories you actually read versus skip. But the free tier caps you at 12 feeds and I'm not paying $24/year (fine, that's not much, but still) for something to compare against a free option that already works.

The Old Reader I couldn't even properly test. They went so viral after the Reader shutdown announcement that they've had signups closed or waitlisted on and off for weeks, and tonight it was throwing errors when I tried to import anything over a couple hundred subscriptions worth of OPML data. I got a half-loaded feed list and gave up after the third timeout.

So: Feedly, for now. Not because it's obviously the best piece of software of the three, but because it worked without a fight and I have a day job. I reserve the right to switch again in a month once the dust settles and everyone stops scrambling. I've also heard Digg is building their own reader from scratch, but as far as I can tell nobody outside their office has actually used it yet, so that's a wait-and-see.

One thing nobody warns you about: your read/unread state and your folder organization do NOT survive most of these migrations cleanly. Feedly kept my folders. It did not keep three years of "read" history, which means every feed showed up with hundreds of unread items I'd already seen. I spent a genuinely unreasonable amount of time last night mass-marking-as-read just to get back to a manageable inbox zero, which is a very small problem to have and also somehow made me irrationally annoyed for a solid hour.

Anyway. If you're one of the people still putting this off, don't wait until June 30th like some kind of RSS daredevil. Export your OPML tonight even if you do nothing else with it.