Farewell FriendFeed. It’s been fun.

I’ve been a strong proponent of FriendFeed since its launch. Its technology, clean interface and “data first, then conversations” approach have made it a highly-successful experiment in social networking for scientists (and other groups). So you may be surprised to hear that from today, I will no longer be importing items into FriendFeed, or participating in the conversations at other feeds.

Here’s a brief explanation and some thoughts on my online activity in the coming months.

The value of FriendFeed
FriendFeed is simply an aggregator, displaying items from other online services. There’s nothing special about that: other sites do the same thing (although many have fallen by the wayside) and were FriendFeed to disappear, those items would still exist at their original “homes” on the Web.

The value of FriendFeed lies in the conversations – the comments and “likes” that the items attract. In this way, users find other, like-minded users via common interest in items of information. This, I believe, explains much of FriendFeed’s appeal to the online science community.

Ultimately though, these conversations are only valuable if we can use them. They may be enjoyable, thought-provoking, even inspiring at the time, but that’s not enough. We need to refer back to them, cite them, analyse them over time to figure out where we are going as “online scientists”. For that we need archive and search – two functions that FriendFeed can no longer provide, largely because the API can no longer retrieve a complete history of our items and conversations.

In effect we are living constantly “in the now” – feeding more and more information into a system from which, ultimately, it cannot be retrieved. Recently, Jon Udell wrote something that really resonated with me: “publish facts about yourself, or your organization, to a place on the web that you control…” I feel that the deficiencies of FriendFeed have removed too much of that control for it to be viable any more.

In deciding what to do about FriendFeed, I noted a couple of other points. First, 90% or more of the items in my FriendFeed stream are from Twitter. This suggests to me that a new “online ecosystem” with Twitter as the “broadcasting service” is a sensible strategy. Second, I’ve never been sold on the concept of the real-time Web. Receiving relevant information as it’s generated is one thing; archiving, curating and mentally processing that information is quite another. To be honest, I’m too tempted to try and process FriendFeed as it happens, when I should really be doing other things. It’s interesting to note the rise of “read it later” services, such as Instapaper, in this regard.

A new strategy
My new approach to online networking is based around two activities:

  • Publishing to services with an adequate degree of control – which essentially means the ability to retrieve or search the archive
  • Broadcasting published items to Twitter, for the benefit of anyone who might be interested


Me, the Web, 2011. Probably.

I used Dabbleboard to throw together this rough diagram, showing how I think it’s all going to work.

Final thoughts
In summary then – look out for my content on Twitter, where I’ll be looking out for yours. I’ll leave you with this code snippet (not fully tested) using a part of the FriendFeed API that still works. It fetches users to whom you are subscribed, checks to see if they have a Twitter account and if so, prints CSV format to the terminal with their FriendFeed id, name and Twitter URL. You can then go and follow them at Twitter, if you aren’t already doing so.


require "rubygems"
require "open-uri"
require "json/pure"

me = JSON.parse(open("http://friendfeed-api.com/v2/feedinfo/USERID").read)

me['subscriptions'].each do |sub|
  id   = sub['id']
  user = JSON.parse(open("http://friendfeed-api.com/v2/feedinfo/#{id}").read)
  user['services'].each do |service|
    if service['id'] == "twitter"
      puts "#{id},#{sub['name']},#{service['profile']}"

15 thoughts on “Farewell FriendFeed. It’s been fun.

  1. “FriendFeed is simply an aggregator, displaying items from other online services.”
    Wrong, that’s how you are currently using it. It doesn’t mean that others do not have communities centred on FF. Right now, I have an academic of >400 students on friendfeed. We aggregate content, but we also ask questions, post original content and have discussions on FF.

    • And I’ve been using it the same way, as you know. In fact that was my point – that the aggregation is not the key feature, since most of the items come from other sources.

      My feeling is that one day, soon, we are not going to be able to use those conversations, so it no longer makes sense for me to keep feeding information into a system that isn’t going to give it back. Twitter is not a replacement – as you say it’s even more ephemeral. I think it’s effective in directing people back to the original source, nothing more than that.

      It feels odd that in trying to streamline my online activity, I’ve culled my favourite! It also feels right.

    • That’s true and I guess, by retweet too.

      I think I’d like for more of the conversations to come back to the original source (such as these blog comments). We’ll see how that goes. It’s still evolving in my mind.

      • One more thing: FriendFeed makes it possible for me to hide a subset of the items you choose to share based on the source (e.g. BioStar comments). If you pipe everything to Twitter, this sort of filtering becomes much harder, unless there’s a client that supports quickly creating filter rules. (I already fillter “ff.im”, but this is via a command line change.)

        Could you, as an intermediate option, tell twitterfeed to add a prefix or hashtag according to the source of the item?

  2. Yes! Hurray! Agreeing all over the place.

    FF just doesn’t work well because the main thing it does it aggregate content you’ve already seen on Twitter, and split up conversations by having discussion threads on both blog posts and FriendFeed items.

    • I appreciate the arguments against shifting conversation away from the source. The counter-argument is that if the barrier to conversation is lower, a wider network is involved and there’s still some link back to the original item, then it doesn’t matter.

      And of course unlike you, I remain a huge fan of FriendFeed. I just feel that now is the time to stop, for me personally.

  3. I’ve used FF before but only to spawn from my Twitter (I haven’t really found a use for it otherwise). I agree with the ability to control information on a personal site…I think social networks are most useful when they lead back to a site you can use. It’s back to the days of tribes and personal connections. You might be able to find them elsewhere, but they should lead back to a place where they can communicate effectively.

  4. It’s almost the end of an era. As perhaps the original life science friendfeeder, this leaves me with mixed feelings, but I find myself heading in the same direction as Neil.

    We know where to find you :)

    • I wonder if the next step could be a common working space instead of conversation space. In the spirit of Biogang idea we could finally collaborate around items instead of talking about them.

      • I’ve been thinking about that too. Conversation is good but I think we need to translate more of it into action, which is much harder to do.

  5. Do not like. Selfishly, I will miss your input on FF; larger issue, I don’t see how your new scheme allows for conversation. Unless I *still* don’t get it, Twitter is even more firmly rooted in real-time, you have to be online 24/7 to get much out of it, and even then it cannot handle conversations between more than two people with any grace at all. I’ve looked hard, and I’m as concerned about the impending heat death of FF as you are, but I have been unable to find anything that even comes close to supporting a community of ideas and arguments the way FF does.

    • To clarify, I don’t see Twitter as a replacement for FriendFeed, in terms of either conversations or long-term archiving. I see it as an effective broadcast mechanism which says “I’ve just created some new content” and points to the source. Nothing more than that.

Comments are closed.