1.0 Release
alexjg opened this issue · comments
For anyone following along PVH's comments above are a list of things we talked about synchronously as being important for a release. I've created issues for most of them and added them to the tracking issue. With the exception of
mergeText
- change/patch confusiong
- BroadcastChannel breaks ephemeral data
We decided that since these can all be fixed in a backwards compatible manner and are low enough priority that we shouldn't block a release on them.
One question is whether we should block this release on #87 . I think @acurrieclark presents a good argument in that issue that we should redesign our message types. We already have interoperable implementations in various states in Rust, Swift, and Kotlin. It would be nice if we could standardise our wire protocol before even more implementations pop up.
On the other hand, I want to get this release out fairly soon and this might be the kind of design task that expands surprisigngly. Maybe we should just add a version negotiation step so that we can solve this later? The tradeoff obviously being that then we might end up supporting the current protocol as a legacy protocol for a while? @pvh @HerbCaudill thoughts?
So, sorry it took a bit to get back to you. I think this list is roughly correct. I would like to land @neftaly 's work on useAwareness and useBootstrap in here as well, but agree we should not block. I wanted to deal with 87 pre-1.0 but I feel at this point that it's not far enough along to commit to and just barely shy of being worth blocking a release (mostly due to uncertainty about what to do).
Since @neftaly asked, the goal is to finish this list for a release in two weeks. That means ideally clearing the list out this week and then spending the following week fixing bugs / polishing / etc.
I suggest fixing #129 for 1.0
@alexjg i think we could probably close this out?