MS has 3 new mobile devices in the works, HP is currently working on one new phone. These will get the "new" W10M
Only Andromeda and the HP device I could consider plausibly "confirmed" and even then they haven't been announced let alone given a release date. At best we are looking at 2018 for any Microsoft handsets IMO, and an HP refresh even if this year, won't be even close to the number of users of other handsets. Andromeda also runs its own separate OS. So while that might get new features, we can pretty much leave that out of any win10m discussion, leaving just the HP, if we were to count our chickens before they hatched.
I think you are missing my point.
I'll break to 3 simple ones
1) Handover.
Even if all these three, currently unconfirmed, potentially imaginary phones were released this year, along with HP's likely but potentially imaginary phone - what kind of downright stupidity would it be commercially to immediate drop all support for your only existing customers, before you even know how the new devices will perform in the market? MSFT has a long history of generous support windows, and I am positive it understands the notion of transition and handover given its known track history - it supports phones long after apple and google would have given them the shaft, Win10 potentially has an installed userbase of around 2.5% of the US market alone. While they are not buying new phones and thus creating 'marketshare', shafting them would be a commercial mistake. MSFT needs to be nice to its existing fans, as it has been with prior support windows, even through all the reboots.
2) The actual code itself. Cshell was clearly _written_ for a 32 bit phone. It RUNS on a 32 bit phone. No new snapdragon cpus are 32 bit. Any new phone would be 64 bit, because all snapdragon processors out recently are 64 bit. It would run a different codebase for the OS itself. If msft planned on dumping all current 32 bit phones, there would be no sense at all in coding such features for the 32 bit version of the operation system at all, and it would not run on such phones. Why would MSFT write 32 bit code if it wasn't going to any devices? Surely they would write it in 64 bit code and it would fail to run on any 32 but device if what you claimed was true. There's just no way that makes sense. It must by sheer logic, come to at least one 32 bit handset and by virtue of that logic, most likely extend to the popular, still high performing models in that archecture (baring performance requirements if those end up being a limitation for some models).
3) No phones/Timing. All of those supposed phones could be prototypes that never get released, like the surface mini, or they could get released in 2019, or they could be figments of our imagination. I could be grey by the time they release a surface phone. Andromeda, yeah it'll come but its not exactly a phone in any traditional sense - its untested in the market - and the HP - well that's a pretty elite phone for some business people. Windows on arm and windows s needs space to breath in order to aid the UWP platform before any mobile release would even be that attractive....and windows on arm on tablets and laptops isn't even out yet.
The late summer updates are nearly here, the fall creators update for PC, with timeline etc, that was promised to win10m is less than a year away. At least some of these updates will hit before windows on arm is even available. What your proposing is that four windows 10 phones, new ones are released between now and they end of the year, all four popular existing phones are dumped like a hotcake, because who cares about customers, and MS will just count on the success of their new phones, because commerce is always a certainty for a company that has more repeated failures in mobile space than blackberry
I think many people doubt there will be even one phone released this year. I suspect HP may do their refresh this year, but I doubt we will see anything from MS. Maybe Andromeda too (but running its own unique OS for its dual screens, not win10m). One new phone that is very expensive, enterprise focused, as the only supported win10m phone would be silly.
The timescale makes no sense. MS has been slowly unfolding this plan for years. Why would it take this delicate operation and potentially mangle it, just to get more phones in peoples hands, if the UWP platform isn't more mature to deal with it, or present customers might feel so jilted, as the only near certain consumers of the new devices, the whole thing just fails?
I can't imagine what you are seeing here. Help me out, tell me, what you see happening between say now and this time next year? I literally can't picture how you see this going down.
I think if you spelled it out, I might understand where we differ, or understand your POV in some way.