[hpk42]

This mechanism was requested/discussed in context of upcoming Android automatic testing

[wchen342]

LGTM, thanks!

[Hocuri]

@wchen342 is there already a PR in the Android repo that uses this new config flag? If not, can you create one, ideally before the PR here is merged?

The reason is that we generally try to merge new APIs only after someone created a companion PR in one of the UI repos. This avoids adding APIs that are not used (because the UI side work is forgotten or turns out to be harder to implement than expected), and it avoids APIs that don't actually fit.

Another question: (Why) Do you need the property that relays from invite links are excluded? The reason why I'm asking is I'm wondering whether this requirement comes from a misunderstanding about the init_transports() API. I mean, why would the UI tests pass an invite link with relays to init_transports(), but then not want these relays to be used?

[hpk42]

Another question: (Why) Do you need the property that relays from invite links are excluded? The reason why I'm asking is I'm wondering whether this requirement comes from a misunderstanding about the init_transports() API. I mean, why would the UI tests pass an invite link with relays to init_transports(), but then not want these relays to be used?

This came less from @wchen342 but from me, and it's indeed probably not neccessary. Actually it's better to allow taking relays from invite links. Usually in tests they will be just other profiles on the same test relays, and in principle such invite links could also be on relays that can interact with the test relays. Core does not need to prevent it. Tests can do it. I'll modify the PR and drop this invite link restriction.

[wchen342]

Yeah we don't really need the relays from invite link to be excluded.

can you create one, ideally before the PR here is merged?

I'm not sure how that works? I assume I will need to update bindings manually and compile against the PR head here? Is that how it usually works? Also @adbenitez

[hpk42]

@wchen342 -- so the branch here is now very minimal. Would still be good if you can ensure it suffices for Android testing purposes. I'll then merge. Thanks @Hocuri for the scrutiny.