I’ll start by saying the motivation is clear and I think this is an elegant mechanism for new extensions. However, the new component redesign is threatening to remove 30% of our users and you should consider a better transition plan.
Here’s where we stand on this: we built Live Emotes, which lets viewers express themselves with customizable emotes on the stream. It was a launch extension and today we support millions of unique viewers every month interacting on thousands of streams.
Live Emotes supports video overlay as well as component formats, since both made sense as originally designed. There’s a major specification difference in the component redesign forcing us to stop supporting components. Specifically, being hidden by default.
-
As far as our data shows, around 30% of our installation are set up as component.
-
Traditionally with deprecation of big Twitch APIs developers are given time, often years. Why are extensions treated differently?
-
Developers have no effective way to communicate with broadcasters about breaking changes. Suppose we drop component support and want to let broadcasters know they should switch to overlay, the only way to do it is an error message showing up instead of the component… And, during that time, we have no choice but to still falsely show up as an extension that supports component.
-
We have considered redesigning the extension to support OBS Browser Source and render it on stream, then have the controls as the new component. We’d like that, but we are limited in implementing, testing, launching and then converting thousands of users on such short notice.
Looking at the proposed timeline, I urge you to reconsider the deprecation of legacy components in a short two months period. Additionally, consider better transition tools to developers. Stability is key in drawing and keeping developers.