Here is the user and agent flow (both sides, live)
- user starts a corona test
- a button sets a toggle - user is ready for video = true
— using the same Glide Table, the agent on the other side will then see the inline list have someone popup for a test (takes 10-20 seconds)
- agent clicks into the user session…
… value column set toggle-isassigning = true
and behind the scenes a webhook is sent to integromat to pass the agent email, plus the verify ID for the session… then integromat writes this back to the Agent GS (so we can handle the ‘race condition’ - as there is up to 15 seconds where the inline list is waiting to pick up the value of ‘toggle-isassigning’ as being true!
next the agent is in a tab - filter = they can see the user row WHEN they are the race winner (integromat sends data in about 3 seconds… Glide takes 10 - 25 seconds to wake up)
- the winning agent gets a button to continue the session - and starts a video
(the Jitsi video is becoming more stable - quite often one side cannot see the other, but if you ‘restart’ then it works- by going back into the video session URL… weird)
Then the agent is doing their work every click increments the agent increment by 1. At the end, agent inc = 7 and the video finishes (and leave the user hanging for 10-20 seconds).
And immediately we should see (and it used to work) the timer for how much time to wait // but not Glide decides to go into space for about a minute before it decides to talk back to the user’s device (on this case… my new super fast mac, Chrome, etc).
Things may have even gotten worse too (on the timer not showing).
Not a great thing!