Closed kensanata closed 1 year ago
User agent is one of those things where the client should send it (https://www.rfc-editor.org/rfc/rfc7231#section-5.5.3), but many neglect to. Since it's not actually required according to http, we should make this requirement configurable. We have plans to allow user-agent filtering, so will probably fold this into that. I'll make a separate issue.
[edit] So in short, not really a bug in either bitlbee or gts imo, but something we should make more flexible.
Cool! Thanks. Closing this one.
Bitlbee is a local IRC server where users connect with an IRC client and then Bitlbee uses plugins to connect to social media accounts, acting as a proxy. I'm the author of the Mastodon plugin for Bitlbee.
When I use bitlbee, authentication doesn't quite work. I can log in, authorize the app, and get a code, but Bitlbee gets an error when it tries to connect:
This is from
internal/middleware/useragent.go
. I'm not quite sure what to make of it and I don't understand the comment in the code.I guess my first preference would be to drop this test, since I don't understand the benefit.
Assuming there is a reason for this code, I will need to check where this happens. The actual requests later (after OAuth is done) are made with the UserAgent header
BitlBee 3.6
so I would have to change it upstream in BitlBee (link to the suspected C code in BitlBee) itself – which makes me think that requiring a UserAgent header is not actually a requirement.What do you think: Is this a GoToSocial bug, or a BitlBee bug?
What's your GoToSocial Version?
GoToSocial 0.6.0 git-f9e5ec9