Access held a "Developer Day" at LinuxWorld yesterday to demonstrate Access Linux Platform (ALP) and it's associated toolset. Access is best known for the NetFront full-web mobile browser and for purchasing PalmSource, including the Palm Garnet OS three years ago. ALP is Access' Linux based mobile OS which is meant to succeed Garnet as a general purpose operating system for smartphones and MIDs. I attended several of the Developer Day sessions to see what I could find out about ALP.
While I have some reservations about ALP's commercial viability, I came away impressed by the OS and its development environment. ALP is a complete mobile platform with a robust PIM and great support for running legacy Palm applications. Also, unlike Garnet (and Windows Mobile or Android), a complete and seemingly robust Java ME run time is an integral part of ALP. Native Linux, Garnet and Java applications all appear together in the Launcher - as far as the user is concerned they are all equal which is as it should be. Support for Garnet and Java ME is a smart move on Access' part as it means ALP will launch with tens of thousands of applications. The ALP Development Suite is a free download form the Access Developer Network site (www.accessdevnet.com).
The Development Suite consists of an Eclipse based native code SDK, a simulator and various utilities. All the development tools and documentation are free and run natively on Ubuntu Linux, which should please FOSS fans. A VMware image is also provided along a link to the free VMware player for developers who prefer to work in the Windows environment. The development environment is full featured with a JIT debugger which supports on-device and simulator based debugging. It seems stable too, during the day long event were Access provided laptops running Ubuntu with the Development Suite preinstalled for use by the attendees, the toolset performed flawlessly. Unlike Symbian and Android there don't seem to be many new paradigms to be mastered in ALP, it's mainly straight forward Linux, GTK and SQLite. An imbedable browser object and support for Access' W3C compliant NetFront Browser Widgets is also provided. Templates and wizards automate the development process and should make it relatively easy for C and C++ developers to get up to speed with ALP native development.
The version of ALP running in the simulator looks quite nice with some UI elements taken from Palm OS particularly in the Launcher (check out that HotSync icon and the Category dropdown at the top right) but with a more modern look. The iPhone influence can also be seen particularly in the Address Book application. I took some screenshots of ALP running in the simulator, including the Launcher (top), the Garnet Memo application (below) and the native version of Memo (bottom) Both Memo apps share the same underling database. The Office application is compatible with Microsoft Office (Word, Excel and PowerPoint) documents.
I honestly hope for ALP to succeed, in terms of features and the development environment it seems to be the nicest of the new mobile Linux mobile platforms. I'm concerned with the lack of OEMs and carriers committed to the platform though. To date only NEC and Panasonic have announced ALP handsets and only for Japan's DoCoMo. France Telecom's Orange has named ALP as an Approved Platform - but the Samsung i800, which was supposted to be Orange's first ALP handset, was just canceled. At Developer Days we were told to look for a significant announcement at the Orange Partner Camp to be held in Florida in December so perhaps we will see an ALP device in the West next year. However with Android, the iPhone, open source Symbian, Palm's own new Linux platform and OpenMoko all competting for developer and vendor mindshare it's hard to see how ALP will take off outside of Japan where it's a logical progression from the Linux based MOAP(L) platform.
Access seems to be doing the right thing by offering a robust development environment and engaging developers - but the other half of launching an new platform; signing up manufacurers and carriers seems to have stalled. Perhaps Access needs to take a different approach with ALP, at least initially, and make it more of a community project like OpenMoko, targeting Linux enthusiasts and providing support for hardware hackers and small OEMs to adapt the OS to existing and homebrew hardware. Maybe even a joint effort with the OpenMoko folks. Once ALP has some grassroots support and market share, it might be an easier sell to the big manufaturers and carriers. What do you think? Does ALP have a chance? What can Access do to make it a success?