The Garnet Source Code Deal: Part 2

In last week’s editorial, The Garnet Source Code Deal: Part 1 , I started a discussion of some of the options that could be facing Palm as they begin to move forward with their new perpetual license for the Palm OS.   This week, I wrap up that discussion with my ideas on Palm and Windows Mobile.

A quick recap

In last week’s editorial, I outlined three possible options for the Palm OS:

• Continue with Palm OS Garnet 5.4.x and add new applications as needed

• Extend Palm OS to version 5.5 and address customer requested features

• Layer the UI and key applications from Palm OS on top of a Linux kernel

In my opinion, the third option is the best way for Palm to get most bang for their licensing buck.   By developing a new OS on top of a Linux kernel, Palm would have some degree of leverage over Microsoft and an alternative to the Windows Mobile operating system.

Option 4: Using Windows Mobile as a foundation for the Palm OS

Using Windows Mobile as a foundation for the Palm OS seems like a radical suggestion from someone who has been such an advocate for the Palm OS platform.   However, there are some subtle reasons for doing this.

Palm already has a business relationship with Microsoft.   What Palm might do is to go back to Microsoft and obtain the rights to use the Pocket PC 2003 OS and build a StyleTap-like Garnet emulator.   By using Pocket PC 2003 Palm would have immediate access to a multitasking, Wi-Fi enabled operating system.   And the use of the Pocket PC 2003 OS should be a cost effective solution now that it is already several years old and Windows Mobile Crossbow, presumably Windows Mobile 6.0, is right around the corner.

For this option to work, Palm would have to work at ensuring that the emulation layer for Palm OS applications would be able to instantiate multiple threads in the processor.   Think of it as being like Microsoft’s Windows-on-Windows (WoW) subsystem.   If Palm were to leave the emulation layer a single-threaded application process, Palm OS applications would work no differently than they do today.   This means that even though the OS would be capable of multi-tasking you would not be able to multi-task Palm OS applications.   Building an emulation layer this way seems to me, counter productive and a waste of Palm’s limited development dollars.

The only way a “Palm OS on Pocket PC 2003” project would work is if Palm leveraged some of the applications already built-in to the Pocket PC 2003 OS (phone, email, web, and Pocket Office) in addition to emulating Palm OS applications.   Palm OS applications would still be single-threaded as they are today while Pocket PC 2003 applications would be able to multi-task next to Palm OS applications.   (This option is very similar to the Palm OS II approach I talked about last week.)

Option 5: Apply Palm’s know-how to Windows Mobile

One of the reasons that many people enjoy using the Treo 700w|wx so much is because of the usability improvements that Palm added Windows Mobile.   (The awesome Treo form factor helps a little bit too, I’m sure.)   Almost a year after Palm started shipping the Treo 700w, I am not aware of any other Windows Mobile licensee who has obtains the rights to make such modifications to Windows.

According to Palm’s press release on the subject, “The new agreement also provides Palm flexibility to use Palm OS Garnet in whole or in part in any Palm product, and together with any other system technologies.”   “In addition, Palm has secured an expansion of its existing patent license from ACCESS to cover all current and future Palm products, regardless of the underlying operating system.”

The Palm press release implies that Palm would be able to port their popular personal information management (PIM) applications to the Windows Mobile platform.   Whether or not Microsoft would allow this to happen is another story.   However running Palm OS applications on Windows Mobile has been validated by the popularity of StyleTap, the Palm OS emulator for Windows Mobile devices.

I believe that Palm will continue to add their “special sauces” to Windows Mobile to differentiate their Windows-based products from competitors.   I don’t believe that porting Palm OS applications to Windows Mobile is a terribly good OS strategy though.   Palm’s President and CEO, Ed Colligan, has repeated on numerous occasions since September 2005 that they are committed to the Palm OS platform.   Because of these reassurances by Palm’s management, I am confident that Palm will continue their current strategy of using multiple operating systems to give their customers the choice of picking the best Palm products for their needs.

The bottom line

The new perpetual license agreement really completes the Palm picture.   Palm is now in total control of their hardware and software solutions now.   As we move into 2007, I see Palm has having three key parts to their operating system strategy:

• Palm OS Garnet will be used as the low cost OS for entry level devices

• Palm will continue to work toward a Linux kernel-based Palm OS

• Palm will work with Microsoft to enhance Windows Mobile for the Treo

I know that these aren’t earth-shattering predictions that some people might have been expecting.   What these predictions, I believe, are realistic and obtainable goals for Palm over the next two years.   If there is one thing that we’ve learned about Palm over the last ten years is that they are a methodical company.

Palm has been executing on the same vision since the beginning.   Palm is a mobile computing company first and foremost.   The single most important application in the mobile computing field is communication, specifically voice and email.   In 2007 and beyond we will see Palm continue to execute on those ideas by using software developed in house by Palm and licensed from Microsoft.