I can see where you got the impression that I thought it was, and I’m sorry to have frustrated you to think that I still believe it even after having been corrected. I repeated that those commands will continue to work on the Twitch website. I did not say how they will continue to work. I did say that they will not use “common” APIs.
It requires an update for that functionality to continue working. Without updating, the functionality will be broken.
You are entitled to your opinion, but it’s disingenuous to suggest it is a settled matter. I am of the completely opposite opinion.
(Emphasis added.) Here it seems you are mistaken. Twitch uses different APIs from third-parties. If Twitch decides at some point to stop using its exclusive, undocumented APIs, that will be a big improvement. Tmi recognizing chat commands has not been preventing Twitch from making that change.
The original announcement above has been updated to include a phased shutdown timeline leading up to February 24, 2023 as the final date for the change to take effect.
“The functionality segmentation is unfortunate… so we’re going to segment the functionality even harder.”
What they actually think is unfortunate is that a standard interoperable protocol is still allowed to access the platform now that they’ve achieved market dominance. They also think it’s unfortunate that people might object to its removal.
This “functionality segmentation” nonsensical low-effort gaslighting of a justification (also noted by @DevMcGee as such) is adequate to prevent the sort of Streisand-effect Twitter had when trying to squelch Mastadon/Nostr crosslinking or that Slack had when they shut off their IRC bridge, because it sounds plausible to non-developers who don’t understand the details.
They don’t think the developers will coordinate, either.
Twitch has done something much smarter than Twitter or Slack in that they’ve frog-boiled developers for years with a progressive series of tiny incremental breakages and arbitrary incompatibilities (breaking +o in favor of nonstandard IRCv3 tags, PRIVMSG in favor of .w, etc.) that aren’t worth the coordination effort to fight in part. The sanest developers at any given time see this stuff and leave, and, by a process of evaporative cooling, the remaning developer pool becomes more and more credulous and masochistic. This has been allowed to happen for so long that there will be no compatibility left to lose and no real objection made in the final (as @prod observes) forced migration from streaming user message events over a socket to streaming them over a websocket.
Welcome to the Bezos land of Embrace, Extend, and Extinguish. The same place they let you use x509 authentication for AWS services so long as you don’t try to use an external certificate authority. Your infrastructure can check in but it can’t check out.
I have a game bot that I run just on my stream to keep chat kinda clean I allow users to whisper the bot. Found out yesterday whispers no longer work through strait IRC commands so now have to write a bunch of new code for whispers in my bot despite them working just fine for the last 6 years.
I’m fully expecting that one day the entire chat system on twitch will no longer use irc. They have very clearly been extremely slowly working there way towards that. Instead of using a well documented and easily accessible system they’d rather make you jump through hoops to communicate with your users.
Now we have no mod commands for our channel to stop the bots coming in, I can see right now this situation is going to allow more bots on twitch for people who was using IRC to stop the bad bots with there good ones using the API now i have to learn maybe a new programming language Python or Java which i have nothing against other than now i need a server to use the API mod commands? so i need to sign up with google or amazon give all your details including house address, credit card number, run web sockets and web hooks and need vast library for something that was a very simple process now this has become alot more complex and im not really sure if i want to be part of this unsafer twitch i was fine with using my own bot to moderate my channel and others i can no longer offer protection to my friends many of us don’t want to use the sponsored bots that everyone know like stream**** or stream**** always money involved it seems
Or you could simply make API calls from the same machine that you had your bot running on, using whatever programming language you like that can make HTTP requests (you’re not limited to just Python or Java).
You don’t need to sign up with Google or Amazon, or give your personal details to a 3rd party, or use webhooks or vast libraries or anything like that.
Thanks for the reply, ive been looking into it i used to write one line of code to do all my functions like ban timeout etc now i have to make requests to the API using HTTP or HTTPS ? to ban users with my bot ? i welcome any positive input
covers the URL to POST to and the payload to send.
Normally this would be done in the same way as you handle token processing in order for your bot to connect to chat. Since you’ll need to also maintaince a token for use. But you can generally use the same token as the one used to connect to chat with. So you should already have the ability to make HTTPS calls
