rimu

No, PieFed posts have a 'canonical' meta tag which sends search engines to the original post. The local copy of a remote post is never indexed by Google anyway.

https://ahrefs.com/blog/canonical-tags/

Deletion of unnecessary data is a cost-saving measure which makes it much less likely that piefed.social will need to shut down forever. Lots of fedi software doesn't do this and it has bad outcomes. Check this out:

mastodon.cloud has 1,300 active users:

Their running costs are $800 per month and now they're shutting down.

I don't know all the reasons why they're closing down but clearly that is an insane amount of money for a service supporting that many users.

PieFed.social has 1600 users and costs whatever the electricity for a 10 year old computer is. $5 per month? On a Hetzner VPS it was about $50 per month.

This isn't big tech where more data means more share price or more AI training data or deeper ad-targeting user profiles. This is smol tech where more data means more problems. If we want to preserve any content we need to delete some content.

Hey, welcome back.

piefed.world hasn't really invested much into populating their Explore menu, to make finding communities easier.

I've just added a few feeds though, check out https://piefed.world/feeds

Also https://feddit.online/topics is an amazing directory of all the communities. It's a separate instance to piefed.world though so when you find good communities you'll need to manually copy their address into your piefed.world account. Or just move yourself to feddit.online...

Cool diagram.

If you removed hexbear and lemmygrad (nearly everyone defederates them so they just clutter the place up) then the other information would be clearer. For example I didn't notice that beehaw has defederated both lemmy.world and sh.itjust.works until my third viewing.

There are 500 Lemmy instances and 80 PieFed instances which are too small to show on the graph :)

Stage 4: code your own social network and run it on the ThinkPad under the stairs with a DB server in a wardrobe.. Fight scrapers and spammers incessantly.

Hey @organicpu@lemmy.organic-server.org did you use a LLM like Ollama to write your post?

No shade if you did, I just want to make sure we're categorizing things correctly at our end.

Months of work to redo one super basic dialog box.

It blows my mind how slow large companies are.

How would people feel about a feature that lets them filter their feed by post type? You could choose not to see any videos or any images, etc.

Oh, this is about the login wall?

lol that was just a temporary measure to fend off the scrapers. They've been really bad the last couple of weeks, especially.

I'm testing out a custom-built Anubis now, that should mean we can do without a login wall.

No problem!

BTW in PieFed you can report a user account, not just individual posts. So if you go to https://piefed.world/u/stermy4u@lemmy.world then click on More -> Report that'll ping the piefed.world admins about it. It looks like in this case they've been responding to individual posts by removing them and not seeing the bigger picture but they will see it once they look at the user profile.

It varies by instance. PieFed has "spam detection" (not really) which is just a list of the posts and comments that have been downvoted a lot in the last 3 days so it's easy for admins to spot unreported badness and take action. I don't know how often other admins do but I check that report multiple times per day.

For example I nuked unlogopoco from piefed.social a month ago

Looking at the modlog for stermy, I'm seeing a fair bit of moderation action being taken on them and very few posts since then:

On piefed.social I banned townflex.com entirely weeks ago, which must be why I'm not seeing their latest spams.

Basically Lemmy has really poor tools for this which makes mods and admins do a lot more work then necessary, causing them to burn out. If they used PieFed they'd have a much better time.

Ahh yep, that'll be it. Thanks.

That I regret link isn't working for me.

I enjoyed that book and it gave me a lot to think about. It changed some of how I do things.

But no way am I going to adopt hexagonal architecture wholesale everywhere all the time. It's pretty over-engineered for the kind of projects I do.

https://github.com/anomalyco/polaris does not exist. Is the url wrong?

Ah, I just found a bug here - the instance blocking wasn't checking the instance of the author, only the instance of the community. It checks both now.

Thanks for bringing this up!