I’m concerned about the New API, and its impact on the functionality available to developers. The Channels endpoints, for me and the app I’ve been working on, were easily the most important and useful available. As of right now, the app I’ve been developing for the last 2 years will need a massive overhaul to be compatible with the New API, and will lose a significant amount of functionality, even if all of the potential functionality in the Trello “Exploring” section is implemented. It seems like rather than merging Users & Channels together into something better, Channels was almost entirely eliminated from the New API.
There was some to be desired by v5, but in general, the functionality and plethora of information available to developers made up for any issues it had.
It’s also not clear from the Trello and New API page whether chat functionality will remain available via IRC, or if we’ll still be able to get events via IRC after v5 is replaced by the New API.
I’m sure people are sick of me saying this, but these types of changes are much more impactful to independent developers like myself. I don’t know that announcing the changes to the API along with a date when they’ll take place before the entirety of the New API is available was the right choice. From what I can tell by the Trello, if the “Exploring” section items aren’t definitively going into the API at some point, the final functionality hasn’t even been determined yet.
I agree with your last paragraph. Where I do my professional development, we work on software that provides (at this point) APIs that are 20 versions old, going back to 2005, along with solid documentation. Why? Because you can’t expect clients to retool everything that they have done every few months. When we get Preview Releases, they are completed and fully documented as well.
I didn’t even switch over to v5 when I noticed the various bug reports and issues coming out until the middle of this year, thinking that February was the cutoff, now I am wondering if the time investment in v5 was even worth it. While I am looking forward to Helix, I wonder how much research and planning is really done versus seat-of-the-pants development decisions.
Chat doesn’t belong to kraken or v5 it’s documentation is just under there, coz Twitch is bad at laying out documentation (and most of the docs are just a import of ye olde githuby)
Curious if you have specific feedback about data that is missing for channels in the new Twitch API. We looked through to remove things that the website relied on that didn’t seem to add a ton of value for third party devs. We likely missed a use case, so specific feedback will be helpful.
Re: roadmap. Please be aware this isn’t the entirety of the roadmap. We’re only adding ~3 months out to the roadmap.
Barry is right about chat. It spans versions similar to authentication. Just hasn’t been moved into the right section for docs.
I’m developing an all-in-one solution for iOS. The app I’m developing enables gamers who aren’t actively at their computers to have full control over their stream. Functionality ranges broadly and will expand if the functionality in v5 carries over to the New API (is that the only name it has atm? It feels odd to keep calling it that…).
I’m very paranoid about someone with a team and more resources pulling the rug out from under me before I can finish my app, but, some of the functionality includes…
–removed due to concerns of someone deciding to take the idea for their own–
Had to dig through a lot of forum posts to finally find this. There’s a notice on top of the docs at Chat & Chatbots | Twitch Developers suggesting that it will be discontinued, so people were saying twitch is killing the irc interface with no replacement at all in the new api, which seemed too stupid to be true.
You really should update that misleading notice, it’s creating FUD out of nothing, and as far as I can see this post in this forum is the only thing stating otherwise.
I am also concerned about some of information that is missing from the users endpoint in the new api. For example In my chat bot I frequently uses a users account creation date as part of a heuristic for detecting and subsequently banning spam bots. I noticed that account creation date is missing from the new user endpoint.