Random sort for items?github.com/stackernews/stacker.news/issues/832
2717 sats \ 4 comments \ @k00b 14 Feb meta freebie
zaps forwarded to @btcbagehot (100%)
Magnificent idea from @btcbagehot in service of evergreenness.
We've been discussing adding a "discover" page to surface old, fantastic content, but I never use the discover page on any site.
Benefits over discover page:
  1. it doesn't need a custom UI and the UX is familiar to existing stackers, ie they would be much more likely to actually use it.
  2. One of the reasons I think discovery pages don't get used is because their UI is bespoke, unfamiliar, and frankly have a "random" UI
  3. @ekzyis notes discover pages also have "hidden/unclear biases (users pretty much expect some hidden agenda now thanks to web2)"
We probably don't want this to be exactly random. Meaning we don't want it surface sub-standard content, but we do want it surface content above a certain threshold of quality without regard for time.
A bunch more fantastic ideas: https://github.com/utanapishtim/stacker.news/blob/snip-0006/snips/snip-0006.md
write
preview
related posts
455 sats \ 1 reply \ @kr 15 Feb
one counterpoint…
something that a discover page does better than a random filter is it shows content that is relevant to the current time, regardless of the created time.
when the world is buzzing about a certain topic, i want to see posts related to that topic on that specific day.
it might be interesting to scroll a random feed if i was bored, but if i really wanted to get more background info on current topics, it wouldn’t be of much use.
reply
235 sats \ 0 replies \ @k00b OP 15 Feb
Fair! I think the are pretty distinct things
reply
455 sats \ 0 replies \ @kr 14 Feb
i like the idea of a sorting system that highlights somewhat random content (with a quality threshold).
maybe this filter could be called the “rabbit hole” or “explore” since “random” doesn’t really describe the filter’s function.
reply
422 sats \ 0 replies \ @WeAreAllSatoshi 15 Feb
I love SNIP - it makes me smile every time.
reply