To encourage participants to speak candidly, the proceedings at the Rutberg “Wireless Influencers” conferences are held away from the prying eyes of journalists. A few interesting ideas popped up during the discussions at the 2008 event over the last two days – but because of the confidentiality rules, I’m not able to name the people who raised these ideas (so I can’t give credit where credit is due).
The common theme of these ideas is the clash of openness and regulation – and (in some cases) the attempt to find creative solutions to this clash.
The first example arose during a talk by a representative from a major operator. The talk described the runaway success one of their products was experiencing in a third world country. This product involves the use of mobile phones to transfer money. The speaker said that the main reason this product could not be deployed in more developed countries (to address use cases like simplifying the payment of money to a teenage baby sitter, or transfering cash to your children) is the deadhand of financial regulations: banks aren’t keen to allow operators to take over some of the functions that have traditionally been restricted to banks, so operators are legally barred from deploying these applications.
I found this ironic. Normally operators are the companies that are criticised for setting up regulatory systems that have the effect of maintaining their control over various important business processes (and thereby preserving their profits). But in this case, it was an operator who was criticising another part of industry for self-interestedly sheltering behind regulations.
Later in the day, one of the streams at the event discussed whether operators could ever allow users to install whatever applications they want, on their phones. The analogy was made with the world of the PC: the providers of network services for PCs generally have no veto over the applications which users choose to install. On the other hand, in some enterprise situations, a corporate IS department may well wish to impose that kind of control. In other words, for PCs, there is a range of different degrees of openness, depending on the environment. So, could a similar range of different degrees of openness be set up for mobile phones?
The idea here is that several different networks could form. In some, the network operator would impose restrictions on the applications that can be installed on the phones. In others, the network operators would be more permissive. In the second kind of network, users would be told that it was their own responsibility to deal with any unintended consequences from applications they installed.
Ideally, a kind of market would be formed, for networks that had different degrees of openness. Then we could let normal market dynamics determine which sort of network would flourish.
Could such a market actually be formed? Could closed networks and open networks co-exist? It seems worth thinking about.
And here’s one more twist – from a keynote discussion on the second day of the event. Rather than a network operator (or some other central certification authority) deciding which applications are suitable for installation on users’ phones, how about using the power of community ratings to push bad applications way down the list of available applications?
That’s an intriguing Web 2.0 kind of idea. On a network operating with this principle, most users would only see apps that had already received positive reviews. Apps that had bad consequences would instead receive bad reviews – and would therefore disappear off the bottom of the list of apps displayed in response to search queries. “Just like on YouTube”.
dw2 
Good to see operators are discovering Banks in developed countries have us under tight control! It seems that in Japan, Asia and Australia the environment is different. Mobile payments and financial services are set to become a major force.
It was reported in ‘Australian Financial Review, January 14th 2008:
‘a shift in venture financing in wireless, from carrier-based applications to consumer-direct applications, including financial services. Both major Australian pilots, Commonwealth Bank of Australia/MasterCard and Telstra/National Australia Bank/Visa use a near field communications (NFC) chip in a phone.’
Also I was surprise to read that volumes are high:
‘around 800 million people are already making mobile payments, with the key components of success including top-up or recharge for prepaid services like airtime, transit and electricity/gas, mobile banking and tolls’
Comment by H3 — 29 October 2008 @ 10:03 am
David,
Google is a great supporter of community-rating mobile applications. This is the very foundation of their Android, too. Symbian, on the other hand, applies a much stricter security model called Platform Security. It’s interesting for me to hear that the chief evangelist of Symbian OS supports Android’s model. Can we expect a shift from PlatSec to the world of “full freedom“? Please note that I’m fully aware of that full freedom is not necessarily the best solution …
Comment by Gábor Török — 29 October 2008 @ 3:58 pm
Hi Gabor,
>Can we expect a shift from PlatSec to the world of "full freedom"?
To be clear, I raised these questions with my Research hat rather than my Evangelist hat. I’m not announcing any policy change. I just think the idea deserves more discussion – before we reach a decision.
>Please note that I'm fully aware of that full freedom is not necessarily the best solution …
I’m keen to hear people’s views of on the pros and cons of full freedom. But the other idea I was trying to float in my posting (again, for discussion) is whether it could be possible to engineer a practical experiment, rather than just trying to resolve the question rationally. The experiment would seek to determine, empirically, whether “community-rating” is a better evaluation process, for apps, than a central certification process.
// dw2-0
Comment by David Wood — 29 October 2008 @ 5:56 pm
Hi David,
>I'm keen to hear people's views of on the pros and cons of full freedom. But the other idea I was trying to float in my posting (again, for discussion) is whether it could be possible to engineer a practical experiment, rather than just trying to resolve the question rationally. The experiment would seek to determine, empirically, whether "community-rating" is a better evaluation process, for apps, than a central certification process.
I don’t think you could test this empirically on a meaninful scale and make it fair because any service offering open application installation on smartphones would attract a certain type of customer – not representative of the average.
However, you could test whether network operator fears of going open are unfounded by convincing an MVNO to offer “unlocked” Symbian phones and a community rating service, preferably with a fairly competitive data plan.
Indeed, the major network operators could even set up their own “virtual” networks with different policies and terms and conditions (including customer service call restrictions). I just don’t think they have any real incentive to do this until they start losing customers to someone that does. That’s not likely to happen until mobile applications are a little more compelling (or there are more available for free) than is the situation now.
Personally I think it would be a shame to lose the security and identity aspects of Symbian’s PlatSec and signing model. IMHO though, community rating would be a much better way of policing application quality than the current testing critera – they should be dropped, or at the very least made optional.
Mark
Comment by m_p_wilcox — 30 October 2008 @ 8:43 am
Hi all,
Symbian could invest in their own trial to figure out if community-rating works well or not. But besides that, it’s more important to watch what Google/Android can achieve with the same approach – you know, it’s much better to learn from others’ mistakes.
I believe, though, that whom should learn the most from these experiments is neither Symbian nor Google – but the operators. You know, Symbian and Google can have their strong opinion of the goodness of this kind of freedom, but if network operators are afraid of letting the community decide then we don’t have anything to talk about. Openness to hear the new and willingness to change the old are required from operators for this new model to establish.
Finally, let’s not talk about the whole PlatSec here, because it’s a complex system that we don’t need to discuss fully over. What is relevant in this context is signing, particularly 1: if un-signed applications can be installed (thus no traceability of vendor), 2: if those signed applications can be installed that are not signed by the operator. My opinion on this is that I’d like to see less involvement of operators in point #2 so that developers wouldn’t need to sign their apps operator by operator. As for point #1, I think ISVs should take the responsibility for their products and signing provides a good solution for this – Symbian use this method to trace ISVs, not sure what Android is doing/will do.
Comment by Gábor Török — 30 October 2008 @ 10:26 am