Welcome
Hello everyone,
I can tell there is a lot of interest in a tool to enhance the mod abilities of Lemmy.
I’ve been researching ways to do it. I have six options that I’m considering.
I’m trying to keep the topics high-level and avoid implementation discussions.
These are undeveloped high-level thoughts. I still have a lot of research into the Lemmy code base and ActivityPub.
It’s late for me, so these notes are from a stream of thought. There may be changes as I think more about this or get community feedback.
Option 1 - SaaS
Summary
A service I host. Probably maintained by donations. I might have an additional tier for enhanced moderation tools.
Pros:
- Simple setup/registration.
- No additional hosting cost for Lemmy admins.
- Ability to use community-shared moderation.
- Community data analysis to find bad actors across the federation.
- Quicker to develop. No need to worry about tech support, setup, configuration wizards, etc.
- Could be set up with any ActivityPub instance. (Mastodon, Lemmy, etc.).
- If successful, it could be my full-time gig.
- I would be more engaged
Cons:
- I have to store personally identifiable information & manage other compliance regulations.
- Feature releases are limited to a single pipeline.
- You don’t get your data.
- No direct DB access to do advanced queries and changes.
- Limited by ActivityPub capabilities.
Option 2 - Self-hosted
Summary
Admins would have to install, set up, and maintain the software.
Pros:
- No cost for me to run anything.
- Admins store their data.
- Direct access to databases to do advanced querying.
- Open source community project requiring/allowing outside development
Cons:
- Additional network traffic if shared moderation is used.
- Link flooding and spam tracking across instances aren’t as likely to exist.
- Would mostly be purpose-built for Lemmy. No additional federation support.
- I would be less engaged and rely on community support
Option 3 - SaaS that requires a self-hosted service
Summary
This is sort of a mix of the above. The interface and interaction are central; however, an agent must be installed to interface with the Lemmy instance.
Pros:
- Best of both worlds.
- A custom API could be built to work with the adapter service. Not limited by ActivityPub.
Cons:
- Worst of both worlds.
- Purpose-built for Lemmy to start. (Adapters for other systems could be built).
- More complex to build.
- Additional server cost and maintenance. More complicated setup for admins.
Option 4 - SaaS & Self-Hosted
Summary
You could use the SaaS solution or install it locally.
Pros:
- I could charge for SaaS to finance the project.
- All the pros of SaaS and Self-hosted
- The people have choice
Cons:
- More complex to build
- All the cons of SaaS and Self-hosted
Option 5 - Ask the devs to do it
Summary
We tell the devs of Lemmy to include better tools and how we want them.
Pros:
- Less work for me/us
- Moderation built into Lemmy and not a separate app
Cons:
- We have to wait for them to do it all
- They might say no
Option 6 - Contribute to the dev repo
Summary
Build enhancements to the core of Lemmy’s moderation tools.
Pros:
- Community support to add features
- Users only have to update to get new features, no configuration
- No additional server costs
Cons:
- Only works on Lemmy
- Lemmy becomes a Monolithic application - scaling challenges
- Pro: A flag could be added to only enable moderation-specific scheduled tasks on a single instance.
Isn’t it possible for anyone to code it directly into lemmy? With it being open source i mean.
In that case you wouldn’t need to build a separate application, just write it into the back end.Great point. I added this as option 6.
Before exploring options, I think we need to define on what we need in our moderation tools.
RE: the options, SaaS’ing this mod tool seems like it goes against the spirit and philosophy of the fediverse. The devs are buried in core work and we need mod tools now.
I’ve been asking people to post in this community what their needs are.
For example:
- Fediverse moderation in 4 simple principles
- Beehaw’s mod tool needs
- My philosophy on moderation in the fediverse
There is an immediate need for some base features, which will be the focus of the first release.
Could you fork the source start adding your features and then merge later?
Yes, that would be the plan if it’s contributed to the core.