[Update: The January release incorporates those changes, all is well again.]

Update: Seems like Nokia isn’t ditching Symbian any time soon, or so they claim. Just wanted to add that for completeness’ sake. The remainder of this post doesn’t concern Symbian at all, so all points remain as little or as much valid as they were before.

The article goes on to quote Nokia manager Quim Gil who states that Nokia is in the process of turning Maemo into a mainstream platform for mobile handsets.

Let us disregard for a moment that “in the process of turning into” makes for a rather elastic timeframe and focus on the question at hand: What benefit does Maemo offer?

Now, I recently acquired a Nokia N810 which is currently selling at bargain prices, so I can draw on some first-hand experience, and while I personally like the device (especially its 800×480 screen and acceptable QWERTZ hardware keyboard) I can’t quite see Maemo  as the Android killer that it inadvertedly must be to stand any chance in the marketplace.

Let’s look at some of the facts:

• The only incarnation of Maemo in the wild is as the “internet tablet OS 2008″ found running on Nokias N810 and its predecessor, the N800. Both devices are not phones in any sense, i.e. apart from their un-phonelike weight and bulk factor, they lack any GSM/3G functionality. They rely on a tethered phone and/or WiFi availability. The devices are actually the precursors (or perhaps pioneers) of what is now being dubbed a MID.
• Nokia went through some effort to have the devices not cannibalize the market for their top-of-the line smart phones: Contact management is limited to storing phone numbers, emails, home pages and a few of VoIP/IM nicknames, with no street address, birth dates or even a mundane notes field in sight. There is no calendaring application and no way to read/edit documents on the go, save the (very useable) PDF viewer.
• The UI is consistent with what the device really is: A handheld computer. There is full multitasking, a task list, a start menu, pull-down/pop-up menus and even double-clicks. While the double clicks are atrocious (and their use inconsistent), the overall UI works rather well (for me).
• Being more a computer than anything else, it is in fact perfectly possible to bog down the device with tasks and have it grind to almost a standstill.
• The first mobile “phone” to sport Maemo is most likely to be the as of yet still unanounced successor to the N810 internet tablet, which by all indication ist “just” going to be a beefed-up tablet with GSM/3G support built-in.
• Maemo does not have any Java capabilities built-in.

What this means is:

• While the guts of Maemo may or may not be easily adaptable to a wide range of devices, the UI as it is now will simply not work on a device with a screen resolution closer to that of traditional smart phones (i.e. around 480×320).  Even the relatively high resolution of the N97 (640×360) is too low to comfortably work with the Maemo UI. A re-design towards more phone-like operation is needed. Given Nokia’s less than brilliant track record in the UI design department, this is going to take a while.
• Nokia needs to invest some serious effort in the PIM and mobile office department.
• CPU demand and/or amount of running applications needs to be limited in order to keep the phone responsive.
• The target demographic for any non-table Maemo device is most likely very different from the current Maemo user base. This includes the developer community.

My point being: Nokia needs to get to a lot of places Android has already been just to get even. There’s still a distincitive lack of features that would Maemo to distance itself from Android.

And one further point: Development. Let’s make a little comparison:

• Primary development language for Maemo is C/C++, using the Gtk+ and/or (starting with the upcoming release) Qt toolkits. Primary development language for Android is Java. There’s nothing wrong with C/C++ but among those developers most likely to develop for a mobile platform, which language proficiency will be the predominant one?
• The Maemo SDK is available solely for Linux, requiring the setup of a VM for Windows/Mac-based developers. Inside of Linux a fairly complex (though well-prepackaged) cross compilation environment needs to be set up. Exception: If you develop exclusively in Python, there is Pluthon which allows you develop natively on the three major operating systems, while running and debugging the code straight on your tablet.
• The Android SDK, on the other hand, is available out-of-the-box for the three major OSes and takes the cross-compilation issues away by being targeted at a virtual machine.
• Application visibility. As loathsome as the Apple AppStore may be to me, it benefits Apple in so many ways that any competitor is virtually forced to emulate it. If Nokia decides to go by that route, actually breathes some life into Ovi and chains its phones to it, this will most likely result in the loss of the installed developer base for Maemo.

So, what’s my point?

• Maemo is a competent OS for the type of device that it currently runs on. It has come quite a way, but it took its sweet time doing so. It has its merits and I’d like to see it around for longer and evolve further.
• I’m totally in favour of Maemo-driven phones, because this would add a significant amount of choice and openness to the handset market.
• I’m sceptical that Nokia will be able to make this work against the stiff competition from Apple, RIM and Android.
• I am a little afraid of the implications that a phone-Maemo might have on the love and care the  tablet edition of Maemo receives.
• I’m warning against the assumption that what appears to be a successful niche market product (and there must be some degree of success, otherwise we wouldn’t be expecting the fourth device in the internet tablet product line) has everything it takes to make a successful mass market product (which Maemo must become if it is to succeed Symbian on a significant amount of handsets).