so it was made the flaurm bot if you don't use the phpflarum part it saves the mods from telling you I think you should have an option that lets you decide weather you want the bot to appear on youSupport post
p.s @clarkwinkelmann I did it is this what you meant when you told me that
There were some internal discussions prior to the bot introduction, I don't know if there was any plan to make an official announcement, I'll let the other staff members see if there was anything to publish from this.
Basically the bot asks for information that the moderators almost always have to ask themselves, wasting time for both moderators and users, as moderators cannot help further without this information and users end up waiting for a human, just to request obvious additional information and then waiting again for the next time a moderator visits the forum.
The bot is programmed to only post if the php flarum info output was not included. If it's included, then the bot will not reply.
For users who don't have SSH access, this is also something they should disclose in their first post, as we otherwise waste time discovering that fact, or even discover they are not self-hosting the forum, in which case we are often unable to provide any support at all. It's really unfortunate that it often takes 2-3 back-and-forth replies only to discover a user is on FreeFlarum and cannot be helped. We could maybe add a paragraph about the information we need from users without SSH in the bot reply.
We already discussed for both 2.0 and 1.x to add the information to the admin area as well. That would alleviate the issue of not having SSH access.
Asking for the hosting party or environment might be a good addition for the bot.
As for the general idea that the bot is replying too often: yes I agree. There are several ways out of there; one would be to hook AI into the equation to make it identify whether the information should be asked or b) a lot of finetuning would be needed to understand the OP and whether the output is actually necessary.
As this implementation is rather new, I'm okay with leaving it as is. We can probably iterate on this in a few weeks again.
