When you consider relative install sizes of various Fediverse software, it becomes clear why #Mastodon often needs to move in ways that ensure we don't break things and do right by all to ensure interoperability and trust & safety in the new things we put out — move fast and break things isn't an option.
Being responsible and careful is a necessity when you carry 70% of installed software by user count.
If we roughly implement things or release stuff without properly testing it and finding rough edges, we risk breaking or making worse, the Fediverse experience for 70% of all its users — that's a huge responsibility.
Trust and safety *must* be considered, protocol design and interoperability with other software is a must, not a nice to have.
We need to lead which sometimes results in needing to slow down feature development to ensure all aspects are considered.
Yes, sometimes big regressions make it in, but they're usually something to be avoided at all costs, and we prioritise fixing a bunch of bugs & performance issues.
And yes, there's our own internal issues with how we work that slows us down, but there's a lot of work going into improving that situation too.
That does also mean we can't immediately work on $HotNewThing because we've people busy working to solve the things we couldn't the last time.
Like Quote Boosts, you know that thing everyone was yelling for last year before clients hacked it in without any of the trust & safety controls? Yeah, that's what's currently being worked on — I've seen blts of early work and y'all would be seriously impressed.
We've also folks working on improving discovery and connectedness for smaller instances.
The harsh reality of 3 core developers and a handful of regular contributors is that stuff does take a long time to build. If you want stuff faster, help fund the people & project.
The only way to move faster is to be able to pay more people to do the work to support 70% of the Fediverse.
Emelia 👸🏻
in reply to Emelia 👸🏻 • • •Being responsible and careful is a necessity when you carry 70% of installed software by user count.
If we roughly implement things or release stuff without properly testing it and finding rough edges, we risk breaking or making worse, the Fediverse experience for 70% of all its users — that's a huge responsibility.
Trust and safety *must* be considered, protocol design and interoperability with other software is a must, not a nice to have.
#Mastodon
Emelia 👸🏻
in reply to Emelia 👸🏻 • • •We need to lead which sometimes results in needing to slow down feature development to ensure all aspects are considered.
Yes, sometimes big regressions make it in, but they're usually something to be avoided at all costs, and we prioritise fixing a bunch of bugs & performance issues.
And yes, there's our own internal issues with how we work that slows us down, but there's a lot of work going into improving that situation too.
#Mastodon
Emelia 👸🏻
in reply to Emelia 👸🏻 • • •That does also mean we can't immediately work on $HotNewThing because we've people busy working to solve the things we couldn't the last time.
Like Quote Boosts, you know that thing everyone was yelling for last year before clients hacked it in without any of the trust & safety controls? Yeah, that's what's currently being worked on — I've seen blts of early work and y'all would be seriously impressed.
We've also folks working on improving discovery and connectedness for smaller instances.
Emelia 👸🏻
in reply to Emelia 👸🏻 • • •The harsh reality of 3 core developers and a handful of regular contributors is that stuff does take a long time to build. If you want stuff faster, help fund the people & project.
The only way to move faster is to be able to pay more people to do the work to support 70% of the Fediverse.