• 0 Posts
  • 10 Comments
Joined 1 year ago
cake
Cake day: June 12th, 2023

help-circle

  • Agreed on your point. We need a way to identify those links so that our browser or app can automatically open them through our own instance.

    I am thinking along the lines of a registered resource type, or maybe a central redirect page, hosted by each instance, that knows how to send you to your instance to view the post there.

    I am sure it is a problem that can be solved. I would however not be in favour of some kind of central identity management. It is to easy a choke point and will take autonomy away from the instances.


  • That should just work. You view the post on your own instance and reply there. That reponse trickles to the other instances.

    It may take a while to propagate though. The paradigm is close to that of the ancient nntp news groups where responses travel at the speed of the server’s synchronisation. It may be tricky for rapid fire conversation, but works well for comments of articles.





  • The exciting thing about this space is that much of it is undefined. It is all about the protocols and the main features at the moment. The 2nd generation tools will be born out of what we discuss now and think about now.

    How do you make sure a user is not trapped in his special interest bubble and still gets to see content that has everyone excited? How will we make use of the underlying data, on both posts and users to suggest and aggregate content.

    I think there will be more than one solution eventually, different flavours of aggregators running on the same underlying data.

    So much possibility. And we control it. If you don’t like the way your lemmy instance or kbin aggregates, choose another site or build your own. The data is there.



  • Edit: Wrote this on mobile. The mobile U/I is not always clear as to the source magazine where the post came from, so I missed the Linux in there. Things are not as dire on Linux as on Windows for AMD, so my assessment may be a bit pessimistic. With AMD’s focus on the data centre for machine learning, the linux driver stack seems fairly well supported.

    I spent the last few days getting stable defusion and pytorch working on my Radeon 6800 XT in windows. The machineml distribution of stable diffusion runs at about 1/4 of the speed of raw rocm when I compare it to the shark tooling, which supports rocm via docker on windows.

    Expect tooling to be clinky and that you will need to compile everything yourself on linux. Prebuilt stuff will all be for Nvidia.

    Amd is pushing hard into the ai space, but aiming at datacenter users. They are rumoured to be building rocm for their windows drivers, but when that will ship is anyone’s guess.

    So right now, if you need to hit the ground running for your academic work, I would recommend NVidia, as much as it pains me, a long time AMD user.