I think there is a lot of truth to this.
If we play the numbers game, what was it something likeI hope your right
Why not? I'm referencing what could be, not what's currently done. No reason you can't install an update in chunks, retry when necessary, and report faults back to the mothership & techs for manual diagnosis. Unless someone needs to pull out a wrench and flash a module on the bench, I can't see a good reason they can't make this process more stable. However, plenty of reasons an org wouldn't - missing competencies, poor or inconsistent decisions, risk vs reward, time and/or money budget, etc. I find this problem fascinating hence my speculation.Can't be done.
As per Chris of Batterylife regarding 2.4 - the mandatory step on the way to 3.0 in Europe - the update has several parts, which need to be installed one by one, verifying every step and should something go wrong, the whole process needs to be re-started. So a tech needs to baby-sit the whole update. And there is a non-zero risk of bricking a module when doing the update.
Speculation can be fun if it's not wrapped up in feelings or pushing opinion as fact!Everything else is baseless speculation and hand-wringing.
I spent decades developing large, complex software systems. My guess is that their upgrade process is not idiot-proof [...]
