• 0 Posts
  • 11 Comments
Joined 1 year ago
cake
Cake day: July 11th, 2023

help-circle

  • I’m attempting this migration on an instance that has been running for about a month, is federated with the top 10+ instances and has synced a lot of data.

    The steps I’m using are as follows:

    stop docker: sudo docker stop domainname_pictrs_1

    run docker-compose to open a session in the stopped container: sudo docker-compose run pictrs sh

    run the cmdlet to migrate pictrs via https://git.asonix.dog/asonix/pict-rs/#filesystem-to-object-storage-migration

    When this runs, it appers to be trying to sync like… all of the lemmy fediverse… to my object storage:

    2023-08-13T17:55:44.426301Z WARN pict_rs: Running checks

    2023-08-13T17:55:45.188984Z WARN pict_rs: Checks complete, migrating store

    2023-08-13T17:55:45.275403Z WARN pict_rs: 56963 hashes will be migrated

    Most of these fail, and I’m trying to run it again with --skip-missing-files , but based on what I’m seeing I don’t know if this is really something that can be done once an instance has federated with a lot of other instances.

    Am I missing something?

    Edit: with --skip-missing-files its telling me that it’s going to take 23403 seconds (6.5 hours) to complete this migration.

    When I look into the bucket, I see all kinds of random images being migrated over, so it’s definitely storing pretty much every image that my instance has ever synced. Is there a way to just migrate content that originated on my instance?





  • I don’t really see a problem with this. Is it so much different from making a good 3D model?

    We’re talking about assets that will be used for generating massive crowds. That’s already done with CGI. These scans aren’t even “AI”… they’re just like metahumans in Cryengine.

    This guy just put the term AI on it because it freaks everyone out.

    If you take the $200 for a motion and body scan and you sign your rights away, that’s what you get. This isn’t a change to how Hollywood already operations. Fear-mongering for nothing.




  • Google made a huge mistake shutting down Google+. If they had built it out to integrate with Youtube, where people could have a space to Tweet, have a Main Page feed like Facebook, and post videos all in the same platform, they would have dominated the market.

    I still have a hard time believing that no-one has created a platform that encompasses all of those things. Meta is doing it piece-meal but it’s all disorganized. It should be one unified platform.

    That’s why I hope some developers start working on a way to integrate Lemmy and Mastodon and like… PeerTube together into a single frontend. I’d love to be able to manage my Mastodon posts and BS on Lemmy in the same website.


  • I think one of the major issues is how poorly we’re doing at directing people to individual instances.

    Lemmy works fine if we have a bunch of good / stable instances created for a variety of different topics and users spread out. All the kinks and things do need to be worked out, but at the same time there needs to be a better way of load-balancing people to different instances. Either that or the entire backend needs to be re-written to allow better load-balancing. I can’t imagine lemmy.world can survive another major influx of users.

    We’re just a small small portion of the reddit userbase. Lemmy will explode if there’s ever a mass migration.