Out of curiosity, i just tried it and it is working. It does:
- Show dialog to approve upgrade step
- On approval deletes the current app (NB: the app stays running and receives back the channel# that will replace it)
- Installs the new app, eventually.
Now, "eventually" above means i did not see the new channel being installed immediately, so there was some moment when my Home screen had neither. I presume new app would have installed on next update check - in my case it did via forced update. (This might have been the straw that broke camel's back as in Co saying "oh, why fix it - let's just abandon it"). Here is what the format of what the calls return (color meaning is like in
this thread):
GetUpgrade() wrote:[{
code: "99999"
name: "channel name"
description: "short description here. Lorem ipsum dolor sit amet, consectetur adipiscing elit..."
cost: "$0.00"
HDPosterUrl: "http://channels.roku.com/images/88888...hd.jpg"
SDPosterUrl: "http://channels.roku.com/images/88888...sd.jpg"
freeTrialQuantity: 1761741830
id: "99999"
paymentSchedule: "None"
qty: 0
]}
doUpgrade() wrote:"99999"
I think doUpgrade() can still a valuable tool in a specific scenarios like the one i described above - moving users from old to new channels. Yes, "freemium" is a better approach if one has the luxury of doing it
de novo. But given RokuCo has spent all the time and effort to implement
doUpgrade() in the first place? - it should probably stay. It's not like it was a trivial work to boot.
Speaking of magical contortions, how many of you guys outside the Co know there is (was?) the concept of Roku app bundles, where buying one item leads to downloading multiple apps?