The teams that did Maemo stuff etc. weren't allowed to ship any features before the Symbian side (which had more political and monetary power) had achieved feature parity...
Hardware was great and software was great, but needed a bit more polish. MeeGo was more advanced than iOS in many ways at the time of this memo, and Maemo had been released way before the iPhone. The 770 was released in late 2005 and it was totally futuristic.
However, Elop had no interest in going this route. When MeeGo was released they dumped the project publicly shortly afterwards and then N9 was sold with no advertising. Despite this, it had a phenomenal demand.
I totally understood it as soon as it was leaked. MicroSoft exec joined a phone company as a CEO back when MS was trying to convince manufacturers to adopt Windows Phone. Double bonus: they could kill and incipient Linux competitor in one stroke.
If you've read the Halloween Documents or docs from the Comes case, it's quite clear: classic "love them to death."
One does not just "join" the biggest company in a foreign country as CEO. The board decided to hire him after evaluating all options.
The board's chairman was this guy who was CEO of Nokia for 14 years. https://en.wikipedia.org/wiki/Jorma_Ollila
Why would he allow Elop to be hired unless he thought it was in the best interest of Nokia to align with Microsoft?
Not to mention the board could've fired Elop anytime if they thought he was prioritizing MS's interests over Nokia's.
Even if Nokia market share would have dropped to less then 10%, that would still have been profitable business for them, and would have made them a major player.
Going from being the dominate player in a major growing market to getting a few billion for it, is not really a success.
Only reason why N9 ever was released is because they simply brutally cut down the scope of the project, e.g. dropping Qualcomm-based variant for US/CDMA markets (that one was repurposed to be their first Windows Phone).
I used a N9 as my daily driver for 7 years. The card-based UI was incredibly elegant and easy to use with one hand. I prefer it to current iOS or Android. Besides, gestures blended really well with the curved screen edges. It had a great dark mode which also blended well with the AMOLED screen.
Offline GPS navigation was a pleasure to use, and unmatched till date. It had a terminal, which let me SSH anywhere to do quick jobs. I handled lots of tasks this way. For example, I used a remote Mutt instance running on my workstation to read email. The terminal was a real terminal running on the N9. For example, ifconfig could work on all network devices, including the mobile radio.
Messaging was highly integrated. Different services (e.g. XMPP, Skype, etc) were just addons. All contact management, chat and calls were performed from the same application irrespective of the protocol used to handle transport.
And lastly, it was an open device. You could install anything you wanted. It was truly Linux on your pocket.
Elop really robbed Nokia of what could have been an incredible product.
Intel gave up on x86 in phones right before Microsoft gave their Windows Mobile 10 Continuum demo, which was almost an eternity after the N9.
If N9 had been widely released and marketted and successful (it did well in small release, but who knows), that may have changed Intel's course. If Nokia hadn't abandoned it before release, it would have gotten application support; I joined WhatsApp around all this, and there were plans to support it as Symbian users were a large part of user base, so the next thing from Nokia was highly likely to be used, but not when Nokia drops it. Third party WA client development for the N9 was somewhat successful and became the start for a lot of other third party clients.
N9 was released in 2011, by that time Nokia was already dead.
Trolltech was acquired in 2008. Nokia could have have a modern alternative to iOS and Android from day 1 but they made the terrible mistake of reinventing the wheel and trusting Gnome.
And what the problems were with gnome? In 2009, the N900 was based on GTK and was a solid base to build on.
The N900 had pretty good reviews and 2 more years of work could likely have made it even better. And Nokia had the developers and control to do what they wanted however arogant Gnome developers were.
I have developed for neither, so I can't really say.
That said, I have been using Gnome based desktop for 10 years now, so I'm bias towards it.
Harmattan (Maemo 6 branded as MeeGo that shipped with N9) was Qt based from the start. Gtk wasn't officially supported at all there, although some support has been brought over by the community.
Also, I guess you have never actually used Qtopia? I did. It wasn't really a great base to build a smartphone platform on. It made a great feature phone though, and was a pretty nice demonstration platform for Qt Embedded. Nokia didn't have Qtopia until they brought Trolltech, which happened after several versions of Maemo were already shipped in working products.
The burning platform memo was the worst of all: lots and lots of mobile app developers intended to develop for Maemo as the third platform because Nokia was still a huge leader and they trusted Nokia. The moment Elop announced Nokia was shifting to Windows Phone, with phones months away, and no possibility of reusing any code from another platform (only .NET allowed, no way to reuse any C/C++ base they were sharing on Android and iOS), that was the moment everybody dropped Nokia. Suicide.
You mean, what Om2008 tried to do? ;]
Blast from the past!
I remember when Maemo first arrived wanting one, and they never made it mainstream.
They arrived at about the same time. They used the same CPU. They both had (approximately) 800x480 screens, which at the time was flagship class. They both used slide-out keyboards, removable batteries, a 5MP camera, 256MB of RAM, and some on-board storage plus a microSD slot. They weighed within 20g of each other.
The N900 had a resistive touchscreen, the Droid was capacitative. Both had an official price of $600-650, but would actually be sold for $200-250 when bought with a 2 year carrier plan (that was the norm back then).
Maemo had a small developer community. Android had a medium-sized dev community. Android had marketing -- for about 2 years, people were calling all Android phones "Droids". Maemo had none.
N900 touchscreen was great. Not only was it perfectly readable in direct sunlight, it was also much more accurate than capacitive screens and not anywhere close to cheap resistive screens' clunkiness people usually imagine when hearing the term. The only issue I had with it was that they tend to develop issues after years of use, so my Nokia N900 does some very annoying ghosting these days.
I still however miss Windows Phone, there were some major mis-steps by Microsoft (the intial side load / developer story) forcing a software break with the move to Windows Phone 8. However as a user provided you could live with the pre-installed apps it was a great phone.
Interesting, I have used N900 for almost a decade and I found its stylus completely unnecessary, never used it. As mentioned earlier, the touchscreen started to be annoying to use in its final years once it started to misbehave, but before that I absolutely preferred it to capacitive screens as it was harder to trigger accidental touches with it.
People ran IRC clients directly on the console on it. A few friends of mine ran them as IRC hosts on their home network well into the 2010s - it was basically a Raspberry Pi with a keyboard and display =)
They had everything they needed to run Linux on (some of) their phones quite early on. Battery life might not have been great, but they had plenty of Linux-running mobile hardware around, kept in bags outside of the building.