Lao
February 3, 2023, 5:28pm
1
I preferred the old system of working on a notebook and having it appear on when we publish it. When I look at recent now, there is a bunch of unfinished notebooks and blocks of probably a class all doing a tutorial.
Before, I felt like I had a chance of receiving a few likes and getting on trending if I made a great notebook. Now I feel like it will get buried and no one will ever see it.
I love that we can subscribe to people now, so in the long term, maybe that kind of social network will help, but only after you’ve established a following.
It would be really nice for notebooks to only appear on recent if the content creators wants to promote it.
9 Likes
Cobus
February 10, 2023, 5:02pm
2
Thanks for the feedback, @Lao . We are looking at making improvements to the /recent page. We think it is advantageous for people to see work in progress also instead of only seeing the beautiful finished products. We are looking at mechanisms and ideas for users who want to promote their notebooks. If you have some suggestions, please submit them to the feedback repo so we can follow up there.
1 Like
Lao
February 22, 2023, 7:44pm
3
Here is a link to the github post with some possible solutions
opened 05:59PM - 22 Feb 23 UTC
**Is your feature request related to a problem? Please describe.**
This is a fo… llow up to this thread on the observables forum: https://talk.observablehq.com/t/exploring-recent-notebooks-is-not-working-well/7594
In short, I don't think the current system is good for searchers or publishers. Before, recent notebooks was a proving ground for new notebooks. If you made something good, there was a decent chance people would see it and like it enough for it to appear on trending.
Now, it is way more work to sift through recent and people are disincentivized from even looking. If there are not enough reviews because people don't want to look or are spread thin, it is not a fair discovery process. It favors people who have an established following or who work with other people who use Observables, who can give them the initial likes needed to get on trending. This creates a barrier to entry for new people. By contrast, trending is consistently high enough quality to attract community attention, and because enough people are participating and rating, the signal is above the noise and trending is meritocratic. Fixing this problem would also improve trending, because more of the good notebooks were discovered and gathered in one place. It would concentrate quality on the high end through natural competition, and it would increase the amount of useful niche diversity on the low end.
With the previous system, it was clear when and why things were happening. I could take as much time as I wanted to make a notebook and know it would only go public when I published it. Like the news, there is a spike in attention when something is published which trails off when everyone who was interested in it has seen it. I knew where I was in the narrative arc, which was fun and exciting. It also taught me a lot, because a few notebooks I didn't think were that big of a deal got a big reaction, and some that I though were really cool were duds. That kind of feedback helps me know what people like and be a better designer. With the current system, I have no idea how anything works, what is happening or how to adapt to it. Instead of going on an exciting adventure when I publish a notebook, I just create in the black vacuum of space and hope something bumps into it eventually.
**Describe the solution you'd like**
1. Go back to the old model where we publish things when they are finished. Have unfinished notebooks visible on that user's notebook page, people you follow or on search, but not on recent. Alternatively, give all users a finite slot or token where they can promote any one notebook at a time.
2. Create a team of editors that review and monitor every new notebook people create. It wouldn't be fair to promote those notebooks with likes, but make a new 'editor's choice' page which would be like a curated recent that would deserve people's attention.
3. Create an algorithm that suggests similar notebooks as the one you are viewing. Like the YouTube sidebar.
4. Be able to see the likes of people you follow or allow people to make 'playlists' collections of other people's notebooks. This would allow a kind of reviewer user who doesn't create any notebooks themselves, but gains a following by having good taste. You could even allow people to follow specific collections while not following the whole user.
5. Have a monthly event, competition or theme to give people a motivating goal and a common interest. This would make it easier to discover new people and allow them to develop the peer to peer following networks.
1 Like
Cobus
February 22, 2023, 11:48pm
4
Thank you for the detailed suggestions in the feedback repo. Much appreciated.