The telegram test spec says: https://github.com/ooni/spec/blob/master/nettests/ts-020-telegram.md#telegram-access-points-check
If at least an HTTP request returns back a response, we consider Telegram to be working as intended
However the fastpath uses a more sophisticated heuristic: https://github.com/ooni/backend/blob/master/fastpath/fastpath/core.py#L654
I think the later is better, since it allows us to capture cases of blocking that are more subtle, yet currently what this leads to is views being out of sync in explorer MAT and measurement pages.
We should figure out how to address this in probe, fastpath and/or explorer.
The telegram test spec says: https://github.com/ooni/spec/blob/master/nettests/ts-020-telegram.md#telegram-access-points-check
However the fastpath uses a more sophisticated heuristic: https://github.com/ooni/backend/blob/master/fastpath/fastpath/core.py#L654
I think the later is better, since it allows us to capture cases of blocking that are more subtle, yet currently what this leads to is views being out of sync in explorer MAT and measurement pages.
We should figure out how to address this in probe, fastpath and/or explorer.