Forget Your Small FriendFeed Wishes, We are Having Dream Coming True

Svetlana Gladkova,


FriendFeed logoRight after we were shown the upcoming FriendFeed beta featuring friend lists and some other interesting tweaks of the interface, people started sharing what else could make our experience of FriendFeed closer to perfect in the special room for discussions the team set up for everyone to share their feedback on the redesign. I believe that the team actually hoped to hear feedback only on how the latest version works and what people actually think about this version in the room. Instead they received tons of wishes big and small since users rushed to share where they wanted FriendFeed to go further.

But no matter how many minor issues have been raised in the room about possible FriendFeed improvements, the team has presented a really big thing today that is actually something no one seemed to expect but what seems to be a groundbreaking effort destined to change the way we perceive information consumption as it is now.

The news today is FriendFeed announcing plans to launch “Simple Update Protocol” some time in September. In a blog post by Paul Buchheit this protocol is described as a “simple and compact ‘ping feed’ that Web services can produce in order to alert consumers of their feeds when a feed has been updated”. And the most important thing about this new protocol is that it acts very fast, thus improving the speed of alerting subscribers of the new content available.

The new simple update protocol (SUP) by FriendFeed will actually operate based on traditional RSS and Atom feeds to query social networks and blogs supported by FriendFeed to fetch new content to users as soon as it becomes available. And the promised speed increase is really amazing: instead of checking sites that broadcast content on FriendFeed every half an hour to an hour, SUP will allow to do it every minute or even every couple of seconds (depending on the settings for each particular type of content, I believe).

So SUP is intended to serve as an addition to RSS and Atom traditional web content syndication protocols to fetch the information from feeds generated by these protocols much faster when a service that sends a query (FriendFeed in our case) uses it.

It should be understood that the new protocol will have to be implemented by the owners of those services generating feeds crawled by FriendFeed so that they could ping FriendFeed when some content is updated instead of FriendFeed crawling this service every 30 to 60 minutes. But I believe owners of such services will be pretty eager to implement it at least to avoid some of the huge broadband consumption FriendFeed crawlers cause. Besides, Paul Buchheit promises it will be an easy thing to implement.

I personally only have a few feeds that I am subscribed to using Anothr, the service that sends updated posts from RSS feeds as instant messages to Gtalk or Skype or a few other IM protocols supported. And while for the majority of feeds I am subscribed to immediacy is not actually important, there are certain things that I really need to see as soon as they are published - and for that the new SUP protocol promised by FriendFeed sounds like an interesting new solution I could really benefit from. This is why I am very enthusiastic about the announcement and believe this will bring FriendFeed to the new level of information consumption for many of us.

The only problem I see here is that out of current half a million users FriendFeed has only a small portion will actually care about the speed of updates while others will still demand minor things that they personally feel their FriendFeed experience lacks along with some major ones like eliminating constant duplicates we are exposed to on FriendFeed. Unfortunately, when you approach the moment when you reach mainstream users, introducing groundbreaking technologies and standards will hardly be appreciated by the majority of them. Besides, immediacy is not of the highest importance in information consumption for all that many web users.


If you enjoyed this post, make sure you subscribe to profy RSS feed!
4 Comments (Subscribe to rss)
  • SUP from what I have read is a serverside implementation to prevent continuous RSS client side polling. Thus moving the RSS model from pull to push.

    This is still another example of the move to the realtime web. Just as GNIP, XMPP (XEP) and SixAparts Update Stream (http://updates.sixapart.com/)

    Personally don’t think SUP will be adopted as requires too much work on the part of the server with little benefit. The benefit seems to be solely for friendfeed.

  • No GravatarSvetlana Gladkova   FriendFeed comment - August 28, 2008 at 07:13 am PDT

    Hey Sam and thanks for the comment, I don’t think I’ve seen you on Profy before.
    It is obvious that the approach is not new since the problem is widely recognized already. I have no information other than what they choose to share and if it is true that bandwidth consumption will be reduced by 90%, I believe at least some publishers will choose to have SUP implemented (especially the huge ones like Flickr, for example). But as a not very techy person myself I can hardly assess how true the claims are.

  • Svetlana –

    SUP is basically a server-side implementation of changes.xml; a changelog for a service. This has been tried before but hopefully FriendFeed will be able to make a go of it this time around, given their momentum.

    Sam –

    Thanks for mentioning Gnip. You could also add various ping services to your list, such as Dave Winer’s original weblogs.com. SUP, weblogs.com (and Gnip) still require the feed consumer to poll the API in order to get to the data. I hope you’ll check back with us next month when we launch content push. At that point data consumers will be freed from having to build any polling infrastructure at all — just tell us who you care about on which services and we’ll push the relevant data to you.

  • @Eric: Yes, hopefully FriendFeed will be able to achieve some momentum for the new protocol given the fact that currently everything done by FriendFeed gets at least some buzz. But I still think it will only be appreciated by a small number of web users that realize the full potential - which must be equally true for other attempts in the field.

Leave a comment (We support avatars from Gravatar, MyBlogLog, and FriendFeed)