Replies: 3 comments 1 reply
-
Beta Was this translation helpful? Give feedback.
-
Another use case for curated collections: I frequently link to this article as an example of "offering a free service is risky": https://baremetrics.com/blog/freemium-saas-implode I'm sure there are other articles like it out there though. So what I'd like to do is start a list like 'freemium + startup-lessons' with this one article, just to begin soliciting other suggestions. |
Beta Was this translation helpful? Give feedback.
-
Having pondered this for a while, I’ve decided that the ranking system (distinct from ordering, i.e. recommend read/view/listen-order) that most appeals to me in the context of Linkblocks is a tier list, as popularized in games: There are dedicated apps for these types of rankings, such as https://tiermaker.com/ and https://www.smashtierlist.com/ Prior art on github:
What I like about tier lists is that they leave more room for multiple top favorites. It’s easier to name your top 3 favorite movies, as opposed to ranking those three movies from best to “worst” which doesn’t feel right. Unnumbered rankings also avoids the temptation to explicitly score our stuff with a number, as if works of art can be reduced to a mathematical score. Link sharing is first and foremost about sharing your favorite stuff, so listing stuff you don’t recommend is a low priority use-case. Therefore I’m imagining a tier list along these lines:
Here’s an example tier list based on my favorite tv shows (that I’ve recalled thus far): Best stuff
Great stuff
Good stuff
The result is that we get more substantive, less numerically judgemental scores such as: Succession
Parks and Recreation
|
Beta Was this translation helpful? Give feedback.
-
Once a rudimentary import is in place, the primary use case I have in mind for linkblocks is curated collections: omnivore-app/omnivore#1845
In short, I’d like the ability to create curated lists of roughly the same kind as what you’ll find on IMDB:
https://www.imdb.com/list/ls055592025/
One way to do it would be to allow lists based on tag combinations, e.g.:
burnout + open-source
The key difference from having a bunch of articles with a certain tag is that a Collection has a ranking, with (1) being the top recommendation. That way you’ll have two key data points that can be used to arrive at a global list of the top3/top5/etc. #burnout+#opensource articles.
I’m making this a discussion because there’s still some details to settle about how to design this. My current thinking is that for any group of articles with a certain label, you can optionally designate 3-5 articles of that label with a ranking, whereas the rest would be unranked.
Beta Was this translation helpful? Give feedback.
All reactions