2019/Utrecht/discovery
Discovery on the indieweb was a session at IndieWebCamp Utrecht 2019.
Notes archived from: https://etherpad.indieweb.org/discovery
IndieWebCamp Utrecht 2019
Session: Discovery on the indieweb
When: 2019-05-18 16:35
Participants
- Neil Mather
- Jeremy Cherfas
- Rosemary Orchard
- Frank Meeuwsen
- Sebastiaan Andeweg
- Djoerd
- Martijn van der Ven
- Dylan
- Dora remote
- David Shanske remote
- Add yourself here⦠(see this for more details)
Notes
- how do we make this the year of discovery?
- how do we build an indieweb version of Nuzzel?
- Discovery: micro.blog has hand curated feed - just 1 person, community manager of micro.blog
- some downsides - 1 person, 1 timezone, can mostly serve on US times
- but lots of positives - potentially avoid negativity, nsfw content
- new are the emoji-feeds: feeds that work based on emoji people use. These tagmoji are restricted; the decision to have a tagmoji lies with micro.blog
- also a photos feed at /discover/photo; does not include photos that have come in from a non-hosted site via a feed.
- muted people are filtered out
- aaron has a page - aaronpk.com/emoji - profile pic changes based on emoji in the post
- seb also has this on his site - seblog.nl/emoji
- stream.indieweb.org - public indieweb feed - comes from aaron's aperture, curated by aaron
- similar to micro.blog in that it is curated - except aaron curates who goes in to the feed, and only for #indieweb hashtag (or link to indieweb.org)
- indiewebring is another way of discovering people - clicking on the links in the webring takes you to a new person in the ring
- belong.io - an example that needs adding to the discovery page (andy baio of xoxo festival)
- links from people that he follows that are talked about or interacted with a lot
- ( Ton Zijlstra: I have a Twitter harvester running for hashtags of interest, get the links from them rank them by usage today / last week, and resolve shortened links to indicate types (video, presentation, pdf, blog etc.)
- similar to Nuzzel approach
- would you rather be addicated to facebook/twitter or the indieweb?
- on facebook you know you are being played with
- another example: bulletin.p3k.io/indieweb
- similar to belong.io - links mentioned in the indieweb channel - sorted by recency and how many people are talking about it
- discovery vs discoverability - how you want to be discovered
- can you follow a particular hashtag?
- indieweb.xyz
- people can submit based on a particular topic
- it could be combined with curation
- kickscondor.com/hrefhunt
- personal websites that he thinks are cool and worth sharing
- automationorchard.com
- python scripts pulling from rss feeds
- manually built list of websites to be included
- https://seblog.nl/2018/10/22/13/three-things-about-readers
- graph db - neo4j
- looks at feeds and pulls the likes from it
- captures who likes different posts
- both aaron and seb like a particular tweet
- the db can be queried to pull out posts with greater number of tweets
- seb discovered joschi through seeing a like from someone else
- you can see likes from the feeds that you follow
- could also work on links, bookmarks, etc - doesn't have to just be likes
- would like to let you 'choose your own algorithm'
- github.com/sebsel/leesmap
- one in laravel and one in elixir
- aaron: separating readers in to smaller projects helped build it
- same could be done for discovery
- split the problem into smaller projects
- algorithm is one of the hard parts
- could existing building blocks help?
- maybe websub - could be used as part of building a crawler
- Ton Zijlstra publishes an opml list of the feeds I follow, so others can explore that.
- Chris Aldrich does the same, I believe.