Dave Winer’s Micro-blogging Experiment

Dave Winer is running a little micro-blogging/link-blogging experiment. We like the direction Dave is heading and would like to enhance feedly to better support it.

Here is an example of link-blog generated by Dave:

An item in this feed looks like this:

<item><description>A second reactor building at the Fukushima 
Daiichiatomic power plant has exploded.</description>
<pubDate>Mon, 14 Mar 2011 04:15:08 GMT</pubDate>

Question to Dave Winer: What would be your reaction to using a richer description to capture the same metadata information? For example, the description of the item could include a <a> which itself could have data-XXX=”YYY” decorations if we want to encode more metadata regarding the resource being linked? This would have the benefit of graciously flowing through all of the existing RSS infrastructure while allowing us to capture more semi-structure micro-messages.

<description>A second reactor building at the Fukushima Daiichi 
atomic power plant has exploded.
See:<a href="http://www.guardian.co.uk/..."
data-XXX="YYY">Explosion at Japan nuclear plant</a>
<pubDate>Mon, 14 Mar 2011 04:15:08 GMT</pubDate>

Author: @feedly

Read more. Know more.

5 thoughts on “Dave Winer’s Micro-blogging Experiment”

  1. First, I really appreciate your interest and will work with you in whatever way I can to help feedly be part of the network my software is on. This is more than an experiment, I’ve put a lot of effort and time into the software, it’s my first product shipment since 2002. Lots of determination here.

    Anyway, I’m not sure what you’re trying to accomplish and what the connection is between Google’s API and my feed. Are you saying that your only way of accessing RSS is through Google? If so, good luck — you’re as stuck as the developers who are depending on Twitter to be the center of their universe.

    The beauty of RSS is that no one controls either you or me. So I’m not going to do anything to trick Google into carrying my data through their servers. Either they want it or they don’t. I don’t see Google as any more important than Feedly. And I sure trust them a lot less than I trust you.

    Hope this helps…

    1. Hi Dave,

      Thanks for the reply.

      As you point out, the reason we love RSS is that you do not have to trust any central authority. Our back end can process RSS feeds directly or using Google Reader as an intermediary – which offers a bunch of value added services: searching, tagging, cross user liking, grouping etc. *with no lock in or trust*: if one day google reader miss behaves, we can simply export the OPML to a new feed processor and voila.

      The reason I was looking at the Google Reader processed version of your link-blog feed is that I want to understand how they handle not title and the additional elements.

      I would be interested in hearing your opinion on extending the RSS item vs. enriching the description with semi-structured micro-format. The way I look at it as a user of RSS is that the semi-structured micro-format provides more flexibility and expressiveness without requiring changes to the core of the RSS format and infrastructure. But your perspective as the creator of RSS is probably different.

  2. I really like the feedly service. And i’m an admirer of Dave’s perspectives and software.
    I’m an RSS advocate and always support the format in my own code.

    It would be great to see feedly continue to be open to adopting some of the new innovative work being done with RSS.
    In doing so, I would encourage you to focus on direct handling of RSS feeds without Google parsing dependence. it sounds like your software does this and that you are just documenting how the Google Reader processor deals with unknown elements and namespaces etc.

    I don’t like the idea of sneaking in html semantics into an already verbose xml document. Clever it might be. But it does not seem right. Almost like inserting JSON inside of CDATA inside of the description element. Useful, maybe…. but strange and not “expected data”.

    I’d rather see feedly take the high road and do a better job than Google with feed parsing and being aware of namespace declarations…. at least the widely used ones and new ones that are being born out of the efforts to create alternative open (micro)blogging networks on top of RSS. Dave has been using and proposing some interesting RSS extensions and i’m also working on one called inReplyTo (for socializing RSS with replies/comments/mentions). Their is also a bitTorrent extension that could get popular etc etc.

    Feedly is a brilliant canvas for experimental RSS aggregation so i’m hoping to see this momentum continue in this direction. I would not bother writing a comment here if I didn’t believe that your service was important in the scheme of things. Great to see your interest and look forward to your thoughts moving forward.


    1. Hi Michael,
      Nice to meet you. From my perspective, this is not an encoding/format war. The reason I was suggesting the html micro-format approach is that I believe that it is more flexible: My understanding is that Dave is trying to evolve the existing RSS infrastructure to allow it to support short messages. These short messages can include links but could also evolve in the future to include people, places, products, barcode, hashtags, etc.. Therefore I think that finding an extensible mechanism to represent semi-structured message is key. I hope this clarifies.

Comments are closed.