My thoughts on the Discord bot api changes overall are that while clearly Discord could have done a much better job at developer relations as well as the actual rollout and implementation of the...
My thoughts on the Discord bot api changes overall are that while clearly Discord could have done a much better job at developer relations as well as the actual rollout and implementation of the changes, it was a long time coming. Discord was never going to be happy with all these bots having full read access to all messages. And that going from the wild west with everything allowed to restrictions was going to be painful for bot developers.
Yes, user bots still exists, and that can never really be fixed. But that's fine - because user bots fall under the same threat model that hostile, spying users fall into, and people know that.
The hidden danger of these bots is that they're implicitly trusted. It's like in sci-fi dystopian movies where people haphazardly let the robot revolution happen under their noses because they don't treat their android servants as even a threat. Most servers have complex role systems that hide channels from users that just join - because forget bots spying on you, humans spy on you too! So that threat model is built into how people handle server administration.
Server owners will just add all kinds of bots without a second thought - implicitly agreeing on behalf of everyone on the server to allow that bot to read everything on that server. Doubly so for the Discord approved bots allowed to be in hundreds if not thousands of servers - of course I can trust it! This was just another cambridge analytica situation waiting to happen.
So normal bots will not be able to read messages, only slash commands, and bots that go through the process of getting the special intent (and Discord an entity to hold liable if it becomes known that said bot is malicious) can still get permission to read all messages. It's not hard to see what and why Discord is doing this for - although the execution could be much better, apparently.
Yeah I can totally understand the frustration from the OP links perspective but I can also easily imagine the issues discord was facing that were not disclosed publicly and which caused these changes.
Yeah I can totally understand the frustration from the OP links perspective but I can also easily imagine the issues discord was facing that were not disclosed publicly and which caused these changes.
My thoughts on the Discord bot api changes overall are that while clearly Discord could have done a much better job at developer relations as well as the actual rollout and implementation of the changes, it was a long time coming. Discord was never going to be happy with all these bots having full read access to all messages. And that going from the wild west with everything allowed to restrictions was going to be painful for bot developers.
Yes, user bots still exists, and that can never really be fixed. But that's fine - because user bots fall under the same threat model that hostile, spying users fall into, and people know that.
The hidden danger of these bots is that they're implicitly trusted. It's like in sci-fi dystopian movies where people haphazardly let the robot revolution happen under their noses because they don't treat their android servants as even a threat. Most servers have complex role systems that hide channels from users that just join - because forget bots spying on you, humans spy on you too! So that threat model is built into how people handle server administration.
Server owners will just add all kinds of bots without a second thought - implicitly agreeing on behalf of everyone on the server to allow that bot to read everything on that server. Doubly so for the Discord approved bots allowed to be in hundreds if not thousands of servers - of course I can trust it! This was just another cambridge analytica situation waiting to happen.
So normal bots will not be able to read messages, only slash commands, and bots that go through the process of getting the special intent (and Discord an entity to hold liable if it becomes known that said bot is malicious) can still get permission to read all messages. It's not hard to see what and why Discord is doing this for - although the execution could be much better, apparently.
Yeah I can totally understand the frustration from the OP links perspective but I can also easily imagine the issues discord was facing that were not disclosed publicly and which caused these changes.