There is certainly a lot of interest in mobile Linux at Orange Partner Camp and I am told the ACCESS briefings have been well attended. I attended "an insight into Linux and our plans" run by Orange and PalmSource/ACCESS. While it was well attended a good portion of the audience was from the other platform providers present.

The session involved Didier Diaz, John Cook, Lary Berkin, and Bill Lee from ACCESS and one representative from Orange's device program. It started with John Cook telling the audience that there are significant differences between the PC environment and a smartphone. Straight away this highlighted an issue with mobile Linux. PC vs. mobile differentiation is something Symbian was talking about 3 or maybe 4 years ago and has long moved on. The fact that ACCESS needed to talk about this problem is perhaps a measure of how far behind Linux is in the mobile space.

Next, ACCESS discussed three reasons why they believe mobile Linux has not taken off: the lack of commercialization in Open Source, the need to enhancing Open Source for the smartphone, and the need to create new components for mobile orientated use cases. The commercialization argument is an interesting one. The Open Source community likes to portray itself as individuals or groups of developers who create software simply for the pleasure of it. The reality is very different. Most significant open source projects are already commercial, they are funded by major IT companies and the developers who cut the code do it as a full time salaried job. Open Source is essentially a very convenient mechanism through which to run a joint or multi-party venture, without the usual contractual overhead. So really commercialization should not be a problem.

Then ACCESS talked about the work they have done. Taking a monolithic PC kernel and paring it down to something that can match the requirements of a mobile device. Yet at the same time John Cook stated that ACCESS had chosen an Open Source kernel as ACCESS are not experts in kernel development. In addition, the ALP kernel is not a mandatory part of the platform and can be replaced by licensees.

Then there is the UI framework. ACCESS uses GDK, but it threw away a large portion of the 150 "widgets" within GDK and developed a set of new ones specifically for the mobile environment, such as telephony specific controls.

The discussion then moved onto development language support: native Linux, MIDP 2.0, and legacy Palm, supported through an emulator. One of the attendees asked what the advantages of working in native Linux will be, the answer was very telling, "Lots and we are developing some resources to explain this, which will be posted on our Web site". Not much of an answer, particularly given ACCESS hopes developers will start working with the platform next year.

Finally the toolchain was discussed. At this stage the toolchain is entirely Linux based. One attendee pointed out that most mobile developers work in a Windows environment (after all the two main mobile platforms use Windows based tools). This is something ACCESS acknowledged as an issue and is figuring out how to address.

The question and answer session threw up another possible issue. ALP is to be licensed in such a way that components within the middleware or application layers can be replaced by licensees, yet there will be a compliance program to ensure ALP devices have commonality. Perhaps not entirely contradictory, but not exactly suggesting a uniform platform which minimizes the work required to address multiple devices from multiple vendors.

Overall the presentation left me less, rather than more, convinced that Linux offers a viable alternative smartphone platform. There were too many unanswered questions given ACCESS expect to roll out to developers and se handset shipments next year. However, mobile Linux remains an interesting phenomenon to watch.

I mentioned that there were a number of people for competing platform providers in the ACCESS presentation. While this is possible at almost any venue, Partner Camp does provide a unique environment for competitors to interact. Edward Kay, Developer Relations Manager for UIQ's Developer Community, noted that "there is no other event quite like [Partner Camp]. The atmosphere is relaxed, informative, and creative; it provides a great way to interact with developers and competitors. On the coach in from the airport I was able to chat with the ACCESS guys. While in some respects we are competitor, we are all working to grow smartphone penetration and Partner Camp is a great place to exchange ideas."

I asked Edward if he felt UIQ was suffering from having only one active licensee in the form of Sony Ericsson. "Being dependant on a single handset manufacturer is a tough position, even more so for developers," said David. "New Licensees will be an important validation of the platform and I'm looking forward to some public announcements."

"The way Sony Ericsson has launched three different devices, in a relatively short timeframe, has demonstrated the flexibility of the platform," said David. "This has been very good for us, illustrating in particular the ability of UIQ to target lifestyle devices; I think this is going to be a big part of the future of UIQ."

Edward acknowledge that the delays in shipping the Sony Ericsson P990 has also cause some problems, but pointed to the Developer Advisory Board (DAB) as one practical way Symbian was looking to address similar issues. "DAB brings together operators, handset manufacturers, aggregators, and tools vendor to look at the things we as a community do well and need to do more of, and things we have done badly and need to improve," explained Edward. "I feel this is an important approach as individual companies probably won't get the cross industry support developers need."

Even with the delays initial indications are that user demand for add-on application for Sony Ericsson's UIQ phones is following the trend set by earlier P series devices. "We are already seeing the P990 climbing Handango's list of top devices adding software," noted Edward. "At the same time we have seen the list of available applications growing rapidly in the last two months."

Thierry Clerc, Partner Engagement Manager at Symbian who has responsibility for working with Orange and was with the S60 contingent noted that "in previous years the community was asking 'why should we work on Symbian?' but is now asking 'how do we work with Symbian?'"

Amongst developers Thierry noted that "small companies, who have decided to work on Symbian as their first or primary mobile platform, are very happy and just love it. Developers who started with Windows mobile complain that working with Symbian is hard, particularly because they have to work with Symbian, the UI supplier and operators in this ecosystem. I will be taking this back with me to see what we can do within Symbian to help. We are also actively looking at other ways to improve, for example when we launched the Platinum Partner program it was about helping developers get into the community. Now so many developers are in the community we realize that we need to offer more and we are thinking about how to do this."

Thierry described the Orange approach as "very exciting." Thierry was also pleased with the attendance at the Symbian related sessions, with about double the attendees seen at previous camps.

Amongst the day's sessions was an introduction to developing for UIQ. It was interesting to note the level of commitment from UIQ to exploring alternative ways for developers to take advantage of the platform's capabilities. More specifically support from UIQ for both Python and Ruby.

In another session Lin Litchfield presented some of the findings from Nokia's Smartphone360 research. This research involved users loading a small application on their S60 device that anonymously reported on device usage. Run in the UK, France, Germany, Finland, USA and Finland Smartphone360 collected data over 3 to 8 months periods from more than 500 users. Some of the key highlights included:

• 64% of users add one new application, while 24% added 5 or more.
• Add on application make up 12% of all application launches during the studies but accounted for 25% of packet data traffic.
• Games accounted for 26% of downloads, but only 2% of phone use.
• Add-on applications were opened more in the afternoon and evening.
• Applications were either removed shortly after installation or remain on the device almost permanently.
• Over 50% of users move items on the menus or between folders.
• 71% of users changed their phone's theme several time, only 13% never changed their phone's theme.
• 40% of S60 device use was for activities other than for voice and messaging.

The business end of my day concluded with "an insight into unlocking the multimedia potential of UIQ3 and Symbian OS." Presented by Adi Rome from Symbian and Magnus Woxblom from UIQ, this was a very good introduction session, providing attendees with just enough technical detail on how multimedia is handled in Symbian OS without risking confusing those unfamiliar with Symbian with masses of detail. For myself, I feel Partner Camp could have done with more sessions like this one. One attendee was however terribly concerned about the multi-threaded approach as "when I do this on Microsoft it keep crashing" because of resource conflicts. The poor chap had difficult believing Symbian has built its multimedia handling to work in this way and was unconvinced that Symbian had built a framework that actually worked!

And then it was time to party.

Tomorrow is the final day of Partner Camp. Return flights and Internet access permitting, watch out for the final camp report around the same time tomorrow.

Remember to check out our images from Partner Camp.