Former Diaspora core team member, I work on various fediverse projects, and also spend my time making music and indie adventure games!

  • 8 Posts
  • 10 Comments
Joined 5 years ago
cake
Cake day: November 29th, 2019

help-circle

  • Sean Tilley@lemmy.mlOPMtoFediverse@lemmy.mlIFTAS is in a Funding Crisis
    link
    fedilink
    English
    arrow-up
    1
    arrow-down
    1
    ·
    2 days ago

    I’m just saying, there’s tangible things to point to which explain the current situation, and how we got here. At the end of the day, compromises had to be made to have a working thing in the first place.

    We can sit and wring our hands about a piece of software not being open source, but ideological purism doesn’t always get things made. Perfect is the enemy of good.

    Besides which, a larger problem is that FOSS devs of critical projects aren’t really making much money, either. You could advance the argument that FOSS isn’t about money, but funding sure helps the longevity of FOSS projects. The Fediverse is practically anemic in this regard.


  • Sean Tilley@lemmy.mlOPMtoFediverse@lemmy.mlIFTAS is in a Funding Crisis
    link
    fedilink
    English
    arrow-up
    3
    arrow-down
    1
    ·
    2 days ago

    Look, open tools are great. I assume you’re referring to FediSeer, or efforts like it.

    For IFTAS’ purposes, they found themselves in a weird situation. Their CCS system for fighting CSAM had to be developed independently, by contractors that were paid. This is because they needed a service that:

    • Could integrate with a national database of CSAM hashes
    • That could be plugged into a federated, open source network
    • That could report on hash matches detected on the public network to the requisite authorities (a legal requirement for instance admins)
    • That would be willing to work with them and take on risk.

    There are off-the-shelf products for this. But, they’re prohibitively expensive, typically geared towards large corporations, and generally unwilling to take on a network of thousands of instances. As a consequence of going their own route of development, their work is beholden to a number of constraints. For example, access to the hash database for the National Center of Missing and Endangered Children (NCMEC) more than likely has legal constraints on implementations not releasing source code.

    TL;DR - they built some things that were designed to solve very specific problems. That development depended on grants and donations. Some things, like FediCheck, may actually be open source and simply exist in parallel with FediSeer as using a different scope. They probably have more plans in the pipeline for stuff that generally doesn’t exist for a big part of the network to use today. They’re running out of money.


  • So, we’ve actually been covering IFTAS for a while: https://wedistribute.org/tag/iftas/

    The org was initially founded in 2023, and they started as a high-level community effort to try and tackle the following issues:

    • Fighting CSAM in the Fediverse (massive undertaking, requires collaboration with NCMEC)
    • Giving admins tooling for coordination against known troll instances and curation capabilities
    • Providing documentation and guidelines for how each platform is distinctly different
    • Providing mental health resources and digital privacy protections to moderators
    • Surveying admins across the network regarding needs their organization could provide.
    • Policy recommendations for instance admins, such as how to handle EU’s Digital Services Act

    I’m probably missing some additional things here. My point is, they weren’t some rinky-dink organization that just emerged uninvited out of nowhere, they developed out of common needs instance admins and moderators in the community have.

    The two systems they offer (as listed in the article) Fedicheck and CCS, as far as I am aware, already have open source alternatives in db0’s Fediseer and whatever his anti-CSAM tool is called.

    This may come as a surprise to you, but overlap between efforts can and does exist, and does not lessen the value of the things overlapping. FediSeer is a perfectly legitimate tool and effort, but these other things were being done at an institutional level, so a different approach was taken. Developing tooling to fight CSAM is complicated, regulation-heavy, and in this case depended on the org having to develop their own tooling after spending a long time talking to existing services that did not want to take on that risk.

    Anything this group is doing should be open source, should be well advertised, and should be well discussed Fediverse-wide.

    While I fundamentally agree, I believe there are reasons their software contractually cannot be open sourced. Presumably because of the integration and reliance on NCMEC and their CSAM hash database. As for being discussed Fediverse-wide…I mean, a decentralized network has no center? There’s a pretty big part of the network that knows about them and has worked with them, but your perception of reach is relative to your vantage point.

    Just because your Scout Troop and the AA meetings use the same building, that doesn’t mean that AA members have any interest in supporting the scouts, or in having the scouts tell them how they should run AA meetings.

    This analogy doesn’t really make sense in regards to the Fediverse. This isn’t “two different groups in a building”, this is a community-developed Non-Profit organization that mostly emerged out of a desire to help make life easier for instance operators. Nobody has to use anything they produce, but a lot of people have benefited from what they’ve provided.












  • Sean Tilley@lemmy.mlMtoFediverse@lemmy.mlIs Funkwhale dead?
    link
    fedilink
    English
    arrow-up
    3
    ·
    1 month ago

    To my knowledge, the project isn’t dead…but, it has been moving at a horribly slow pace for a very long time.

    Funkwhale is a pretty cool project, but it’s one of those things where the ActivityPub implementation really was bolted on well after the core experience was defined and developed. It was meant to be a Grooveshark clone, while a lot of people were hoping to use it in a more social way, like SoundCloud.