So much for being a privacy conscious site /s

Anyway, my default setting is activity for the last 7 days mainly for the weekly recurring threads, otherwise I'd probably use 3 days.

(Mildly offtopic but an option to separate recurring threads from normal threads might be nice, since if you're using a shorter period for stuff to show up on the homepage, the weekly recurring threads which should be used/active for the whole week kind of don't.)

This idea is inspired (at least for me, there are probably actual forums like Tildes to draw better comparisons to and take better inspiration from) by Danbooru (P.S: This image is just SFW scenery but the site as a whole is not) , where they have meta tags for stuff like image resolution, if it has commentary, it it's translated, animated, GIF, etc.

Should we consider that but for tags like long and short read or watch, videos, reposts/duplicate posts, spoiler threads, recurring.[ ], maybe news article authors too (also appropriated from Danbooru), since these can supercede any topic or group and will rarely be suggested in any single one of them?

If it's not clear what that looks like, imagine all the normal tags being suggested/typeable at the top and all the meta tags being suggested in a separate search box just below the current one, which are displayed regardless of which group you're in, since they can apply to all the site.

Not leaving the house much these days (due to social distancing and also insane heat in NYC right now) means I've got some time to kill that I'd like to spend productively.

I took MIT 6.00.2x: Introduction to Computational Thinking and Data Science a few years back when I was refreshing my Python skills. I think it's been updated a bit since then. It was a high quality course and I enjoyed it, though there are so many Python-related courses these days, I can't guarantee it's the best.

I'm currently taking:

1. Model Thinking on Coursera from the University of Michigan. I don't know where I saw this recommended (maybe on Tildes or Hacker News?) but it's quite good so far. Scott Page teaches about how to use various models (mental models, computational ones, etc.) for breaking down and analyzing various problems and systems. I've only just started but I quite like it.

2. Testing and Monitoring Machine Learning Model Deployments on Udemy. Taking this along with a few coworkers since it's relevant to what I do. Only just starting but appears to be quite good and works through a well-documented example project on Github.

I've also come across a few that seem like they might be good courses for the future:

• Bayesian Methods for Hackers
• Probalistic graphical models on Coursera (3-part sequence, not free)
• Computational probability and inference

Now your turn: what have you taken? What did you like or not like, and why? What do you want to take?

What have you been watching and reading this week? You don't need to give us a whole essay if you don't want to, but please write something! Feel free to talk about something you saw that was cool, something that was bad, ask for recommendations, or anything else you can think of.

If you want to, feel free to find the thing you're talking about and link to its pages on Anilist, MAL, or any other database you use!