Nokia's internal presentation after iPhone was launched (2007) [pdf]

Back in the days of the Bell System, the upper management at AT&T believed that it was going to be circuit-switched forever, even as Bell Labs was building packet-switched audio networks and it was becoming clear that packet-switching was a vastly more efficient solution to moving large amounts of mixed data around at a time. The development of efficient switching networks [0] was fundamentally resulting in continually building bigger networks that took up more space -- it was the Strowger step-by-step problem all over again. Moving to a packet-switched system meant that you could have an infinite number of "circuits" so long as you kept track of the paths taken.

But even as AT&T Long Lines implemented this, upper AT&T management was firm that the fundamental design of the network was not to shuffle packets around but instead to connect point A and point B with services on either end for the subscriber.

Even when they did eventually try to accept the packet-switched system, ISDN was too big and bulky, too slow for anything practical, and by the time it was useful, Ethernet/IP came along and ate its lunch.

[0] https://en.wikipedia.org/wiki/Nonblocking_minimal_spanning_s...

Nokia never really had a chance--the N-series was a mess of patched software, they had no real Linux alternative, and their supply chain was fragmented six ways from Sunday because they churned out dozens of SKUs.

Then everyone went into denial because they couldn't believe Apple would be successful by going outside established norms (like refusing to customize the homescreen or packaging for telcos, etc.).

A few telcos tried to respond by picking their own champion smartphone (Verizon did that in the US). I ended up having to talk my CMO out of going all out on promoting the Blackberry Storm (which was a dud of epic proportions).

I later became the product manager for the iPhone as well, and that was an amazing roller coaster I will eventually write about (it's been around 17 years, so I think I'll get to it sometime soon).

But I would recommend folk interested in the intervening years to read Operation Elop: https://asokan.org/operation-elop/

I also had a front row seat to that...

Windows Phone 7 had Silverlight as the app framework, which, to remind, was itself basically a rewrite of a subset of WPF in native code for perf (although the public API remained .NET).

And then after that it was WinRT / UWP, which was effectively further evolution of Silverlight in terms of how it looked to app devs.

WP7 was a really low point for the series because not only the new app dev story was completely and utterly incompatible with anything done before, it also had a very limited feature set in terms of what you could actually do inside the app - much worse than the iOS sandbox.

WP7 -> WinRT transition was easier because WinRT was so similar to Silverlight in terms of APIs (in some simple cases you literally just had to change the using-namespace declaration to compile). It also added enough functionality for more interesting apps to be viable. But by then, the reputation hit from both devs (who were being told to again rewrite everything they already rewrote for WP7) and users (who were being told again that their devices won't get the new OS, and the new apps are incompatible with the old OS) was too much for the platform, IMO.

And then on top of all that Google actively sabotaged it by refusing to make apps for its popular services - such as YouTube - and actively pursuing third-party apps that tried to fill that gap.