moparisthebest> Suppose the _real_ way to get things moving would be to develop an actual exploit and wave it around threateningly 😈️
Yes this for real
moparisthebestSam: you always display the call button regardless so when they log onto their client that supports calling they see the missed call
moparisthebestAlso when they don't answer you can "leave a voicemail"
SamOnly if they have ever had a client that supports calling.
moparisthebestWhy? Maybe they'll install one tommorow?
moparisthebestBesides, either way they'll get your voicemail
mdoschCalling is a bad example. If they install a client with calls tomorrow there is no point in calling today or did I miss the time travel XEP?
menelI think the example was, that the fallback is good enough, (http upload with voice) , so you _can_ use it anyways
jonas’replace tomorrow by "in 2 seconds when the iOS client you had not seen yet gets pushed because of your call"
jonas’I think the more realistic threat is, indeed, confusing PEP implementations
larmamdosch, You can still at least display in a client tomorrow that the person attempted to do a call before they sent the voice mail (because the Call Invite is in MAM). Also from caller side, there really is no difference between "no client that supports calls" and "no client picks up call", so you need to be prepared for this to happen anyways.
larma(not saying it might be a good idea to let the user know if it seems like the other side has no client with call support)
MattJOn a related note, I plan to automatically reject incoming call invites on the server if it's not immediately deliverable (via a live connection or push notification)
MattJIn the long term I see capabilities being attached to push registrations
larmaWhat's wrong with call invites in MAM?
larmaAh, with reject you mean, to actively reject (not just error)
MattJIt just improves the UX a little for the caller, and will hopefully weed out some issues I've been seeing people having
larmaMake sure to only do this when there is subscription...
larma<gone/> condition from https://larma.de/xeps/xep-0353.html#table-1 would be good for this, right? Or do you think there should be a new condition for this?
MattJI think <gone/> fits (if you do :) )
larmaIt was originally meant for when the caller goes offline while there is a pending call, so it has no real meaning when sent by the callee.
larmaSo now we have one 🙂
jonas’Did I just observe an exchange between XMPP community members, agreeing on a solution to an issue *immediately*?