Using an image classifier seems bulky.

Doing a less expensive, more naive approach could be a fun side project.

Would it be enough to allow the DOM chunk to arrive, but just filter it client-side, in users' browsers, so they never see it?

I'm new to BlueSky, but it looks like you can search "mushroom" at bsky.app/feeds to find feeds like bsky.app/profile/hark....

spotify.com uses React.js.

In my spare time on weekends, I'd like to build whatever would make people's lives significantly easier.

It's hard to know what that is right now, though.

Do you all have any ideas?

I do - I'm new here.

I like to build things during weekends and evenings if I have time.

How have you been finding this platform so far?

What are some of the things you dislike about it?

If the original tool had just been marketed and advertised more, I could've just used it from the beginning, instead of creating a worse version of something that already existed.

I spent a lot of time building a tool to auto-complete text.

Then, when I thought about monetizing it, I did some research, and it turns out there was already a free, open-source tool that already did what my tool does, plus so much more, all for free, and with an excellent UI.

Marketing isn't inherently evil.

filter(lambda x: x in nums, nums) is "lazy" and doesn't run immediately.

When we do nums.pop(), nums becomes [1,2,3].

It's only once we do

print(list(f))

that

f = filter(lambda x: x in nums, nums)

gets evaluated and we get [1,2,3].