2013/UK/Readers

Archived from https://etherpad.mozilla.org/indiewebcamp-readers, please clean.

IndieWebCampUK 2013-09-07 readers session
Is polling good enough for the indieweb?
 * No - lol
 * Polling might be good enough if you're just following 50 indieweb sites

The use case of celebrity was discussed, but for purposes of getting quickly off the ground, getting the basics done is more important.

Note: in 2003 all blogging UI/feature progress stopped - pingback was the last thing that was created.

Why did this stop? May have to look into this on archive.org. One thing might have been that there was no clear next steps to make the systems interoperate, and later on the siloed social networks. Perhaps the people that might have implemented feeds on their own sites and interoperability on their own blogs joined or started working on siloed projects.


 * 2002 Friendster opened to the public. (grew to 3 million users in three months).
 * 2003 Myspace launched. Linden Lab and Linkedin also launched.

In 2003, there was no Twitter, no Facebook, and no Friendster API. Blogs were all alone and just had to communicate together. Pingback was enough then.

The arguments in 2003 were RSS or ATOM. It wasn't whether or not you had access to a feed, but what format it was in.


 * 2004 - Facebook launched.
 * 2006 - Twitter launched, Myspace led social sites.
 * 2007 - iPhone released
 * 2008 - Facebook overtook Myspace as leading site.

(More: http://www.uncp.edu/home/acurtis/NewMedia/SocialMedia/SocialMediaHistory.html)


 * Twitter shut down their feeds (due to cost?)
 * Maintaining feeds is just enough more code that it can be an issue and have errors.
 * JSON as a format is interesting, but it isn't fair to require people to provide another way to get data. Twitter has their data provided in a proprietary JSON format.
 * Microformats and Microdata is great for SEO. Not many use Microdata.
 * SEO (and spam) is what killed ping/trackbacks. If it's great for robots, you know it will be great for spammers.

There used to be ways to get custom data out of a page or set of pages. Yahoo! Pipes helped. One could edit OPML files. You could use static to RSS feed generators. Could not publish the feed but a list of URLs.

What's the difference between IndieReader and an RSS reader:


 * The same difference Twitter/Facebook and a Feed Reader.
 * There's a place to post on Twitter/Facebook, and none on Feed Reader.
 * There's also the ability (through an integrated UI) to make a comment or post on any of the content published. This singlehandedly killed the feed reader as we know it, as the UI was integrated.

YAGNI (you ain't gonna need it)

Making a Reader
Alf: To make a reader, we need, something to get:

use-case: Filtered search
 * all the content
 * ordered by the date it was created
 * changes
 * deletes, either incorporated into changes feed or its own
 * jump into that stream, and navigate up or down
 * aggregating everything
 * filtering it by some criteria, e.g. search terms, date ranges

Planet aggregators use-case:
 * want to consume feeds that are articles-only, feed with parameter only
 * shortcomings
 * no way to send a planet a delete
 * no way to update

Making sharing easy?
 * export sharing list on RSS
 * publish sharing list as a site with mf for better export

Planet notification of aggregation Similar to a follower notification.

What happened to Salmon?
 * Too hard to consume

Barnaby described how his Twitter shim works in the PHP microformats2 parser.

Conclusion: we can implement a lot of what readers need with HTML + next/prev posts. Is that optimized enough to build more on top of it?