I don't know who needs to hear this, but I'm getting tired just thinking about the testing path you would need to follow in order to fork mastodon and keep it as an in-place upgrade from baseline.
stand up a baseline mastodon instance
federate with some peers
fill it with realistic data
migrate to the fork
test the fork
test federation
repeat across several combinations of mastodon base versions and other fedi servers
If in-place upgrades aren't the goal, then just build something new. Or contribute to one of the various preexisting projects. Don't adopt mastodon's tech debt for no reason. It's not worth it.
I know I'm eager to have more help, if existing projects suddenly sound better.
I thought for a bit about what it would take to be able to maintain an upgrade path in the future and keep that tested and up to date and got exhausted just thinking about it. The amount of work is just so much and it is grinding work, not fun work or exciting work.
…but if that isn't the goal, then there are other projects that desperately need people, or you can roll your own if you want to scratch an itch or tease out some ideas.
@jenniferplusplus It's just like… If mastodon had a more reasonable amount of tech debt or there were better standards for ActivityPub or we had better test suites in general then it would be much less of an issue.
But those things aren't true, so the work is enormous and testing that continuously is enormous and maintaining that infrastructure is by itself time consuming and just…
@jenniferplusplus when I talk about it I call it "generic activitypub server". I started it as a "reference implementation" for the GoActivityPub library.
Add comment