Of OpenGardens, Walled Gardens, Tim Wu, Net Neutrality, Carterfone and IMS


Probably my longest article on this blog. I am preparing a pdf version as well!


As many of you know Columbia University Professor Tim Wu has written a fascinating article called: Wireless Net Neutrality: Cellular Carterfone on Mobile Networks . If you have not read it already, I very much recommend that you do!

With a blog called ‘OpenGardens’, this topic is clearly of deep interest to me.

However, after reading the article, I have mixed views. The paper does a great job of highlighting the issues we face today, but I also believe that it mistaken on many other issues – and I find my selves taking a stance which is much more neutral.

This article outlines my thoughts on this document.

All comments welcome to ajit.jaokar at futuretext.com.

Particularly, I welcome feedback from Operators – since many in Europe I know are thinking of Opening up the network in some way.

Synopsis of my thinking• Tim Wu’s document does a great job of highlighting the issues. It does an excellent job about raising consumer awareness and I truly hope it drives some thinking at Government and at Operator levels.

• The USA is clearly losing out due to the practises highlighted in the document. However, these practises are not a global trend. Even within the USA, there is a range of behaviour exhibited by Operators. Legislation may not be the only option here – and indeed may not be the best option.

• The whole document is primarily based on the application of Carterfone principle to the Mobile Operators. I believe that the Carterphone principle can be applied to voice but not to data unless certain other conditions are fulfilled first – which I explain below. Thus, it is not a simple case of applying the Carterphone principle to the Mobile Network Operator as is implied

• Oddly enough, the document does not talk of IMS . IMS is a huge buzzword in the industry and addresses the very same issues that are outlined here. I discuss some implications of IMS below.

• In fact, if IMS were to be considered, Skype would be one of the ‘walled gardens’ (since it does not use the SIP protocol – which means it is not Open).

• It is ironic that the Carterphone principle (opening up the Mobile network as envisaged in the document), would benefit the two closest walled gardens – Skype and iPhone. (My previous caveats about applying the Carterphone principle to voice apply in this case)

• The issues highlighted are not ONLY due to the Operator in all cases. For instance, for Mobile Video, the government is an important player(and cause) of the delay in deployment.

• The document does not address net neutrality as I understand it. Thus, net neutrality seems to be a keyword designed to gain some emotional currency

• I totally agree on the discussions about Disclosure. There may be a case for legislation in this sector only i.e. enforcing greater disclosure

• Cooperation and not competition is the key!! In that sense, unless America does some fundamental rethink, it may well lose out in the end because of it’s culture of competing vs. collaborating

My core belief (and potential solution to the issues outlined by Tim) is :

In a consumption driven world, competition is the driving force. In a creation driven world(User generated content, Web 2.0, Mobile Web 2.0 etc), cooperation is the driving force. For reasons I explain below, we have no choice but to accept a diverse ecosystem in the West and legislation is not a good option. Instead, we should look at cooperation and try to identify the dimensions of the stack where we can get critical mass.


Verizon: Firstly, let us acknowledge that Verizon is an exception. Much of what they do will be severely limiting in a Web 2.0/Social networking world – simply because Version customers will want to interact with other people (and not just other Verizon customers). Like AOL, Verizon will find out that it can’t maintain a walled garden for ever. Now, having got the issue of Verizon out of the way, let us look at a wider picture

Skype and iPhone: The irony here is – Opening up the network in this way, would benefit two of the closest gardens : i.e. Skype and iPhone

No IMS?: Surprisingly, Tim Wu’s document does not mention IMS at all. This is strange, since IMS is such a huge buzzword in the industry today. IMS also plays directly on the net neutrality issue in the sense either you can view it as a walled garden (at the packet level) or you can consider the disruptive potential of naked SIP (SIP sans IMS). Or, like I do, think that IMS has some unique and powerful features and depending on how IMS is implemented – this can be truly beneficial to the industry.

In any case, what you can’t do is – ignore IMS!

SIP and Skype : If IMS were to be included, Skype does not look that good – because it does not follow an open protocol(SIP).

The Carterfone principle: The whole document is primarily based on the application of Carterfone principle to the Mobile Operators. In my view, The Carterphone principle can be applied to voice but not to data unless the Operator creates APIs (Application Programming Interfaces) first. Carterfone without APIs is practically useless.

We can illustrate this by considering the example of Visual voicemail in iPhones. As a consumer, I love the idea of Visual voicemail. That’s EXACTLY what I want .. I hate trawling through old voicemail(and worse still – remembering keys to go next voicemail etc etc!). So, it’s fantastic to have visual voicemail .. BUT guess what? Supporting visual voicemail implies that the two (device and Carrier) are intimately in bed! Else, it is not possible to provide such a service(because the voicemail is stored on the network and not the device) .

Thus, in a Voice scenario, Carterfone may work – but in a data scenario it will not because to make a useful service(like Visual voicemail), the device needs the network to be abstracted(i.e. API enabled).

Tim Wu’s document misses this critical point.

An API enabled Mobile Operator actually is a very powerful proposition – something which I have been speaking of for some time as in Mobile web 20: Re-engineering the digital ecosystem with converged digital processes in a Post IMS/Quad play world and in The Long tail and Mobile Web 2.0

Thus, it’s certainly not a case of directly applying the Carterfone principle (i.e. Plug and Play) to Mobile Data scenarios.

Carterfone, APIs and Pipes: Note that in the above API enabled scenario, the Operator is still the hub(and not the dumb pipe). The Operator still retains leverage in this situation by managing the APIs

The iPhone : Tim says: >>> Most importantly, to the surprise of many, the iPhone only works on the network of a single carrier, AT&T Wireless. The hundreds of millions of consumers who are not AT&T Wireless customers cannot make use of the iPhone unless they become AT&T customers. The question is, why? Why can’t you just buy a cell phone and use it on any network, like a normal phone? <<<

No No No .. I was not surprised. See my longish post saying ..

The iPhone is extraordinary not because of it’s UI but because it’s the tail wagging the dog ..( But the real question is: How many dogs can it wag?)

In fact, as I said before, the iPhone CANNOT work with multiple Operators with ease because it needs to be in bed with the Operator to give that unique user experience. The client alone cannot deliver that user experience and I use the case of Visual voicemail to indicate why

Locked phones? In light of the above, locked phones are not such a big issue.

Qualcomm/ BREW : One must also not ignore the role of Qualcomm/BREW in this – i.e. Qualcomm by definition leads to a certain ecosystem biased towards a walled garden at all levels(network, apps and so on). Outside that scenario, the Mobile world is a lot more open place.

Mobile TV and Video: Mobile TV and Video is hobbled not by the Operators – but due to a range of other factors(including broadcast spectrum allocations in Europe). Thus, it’s a much more complex issue – with standards wars at technology/broadcast level – not a pure Operator scenario

Developers: When the article talks of developers, the implicit assumption in the document is : you need to be on the carrier portal/deck. However, Operators were never good at marketing mobile applications. In other words, even if you did end up on the deck, it may not translate to sales(excluding simple content such as games etc).

The Operator Portal also degrades the brand of the content/application owner. They are not the destination site. If they were strong or innovative enough, they would not want to be eclipsed by the Operator’s brand and instead would want to promote their own brand. Clearly, the off portal market in the USA is not mature (and this is the real problem i.e. short codes need to be viable and interoperable). In the UK, companies like Yell Mobile , which have good content, are primarily taking the off portal route rather than the Operator route.

Thus, if the service is compelling enough, I would not recommend the Portal route. As a corollary, I think the off portal market in the US needs to improve. That’s a real issue

WAP vs. Full browsers : Yes, WAP was a crippled version of the Web. That’s true. But, in the early days (low bandwidth, low CPU etc), was there an option? i.e. could we have really been able to run full web browsers on mobile devices? Today, as we increasingly see the uptake of full web browsers on Mobile devices from companies such as Nokia and Opera, these vendors are redefining the landscape(and I include Widgets amongst browser technology). This will lead to applications that span the Web and the Mobile Web using technologies like Mobile Ajax, Mobile Widgets and WICD () – making long tail applications possible.

Openmoko : Openmoko has been on my radar. It is interesting. Time will tell. It still needs a network though and I am not sure how that works.

Net neutrality: Defining net neutrality primarily in terms of terms of the Carterphone principle sounds limiting to me?

The mobile network is different and the Mobile network operator is not anonymous : Mobile network operator is not anonymous. In fact they are a (longish!) phone call away. But accessible none the less. Which means – they can be sued. I don’t think this fact can be ignored. For instance, if you get Spam on the Web, there is not a lot you can do. However, if you get Spam on the Mobile web – who do you call? Your Operator. In practise, this fact cant be ignored.

The security risks are also higher due to the personal nature of the device. For example, a child may be using a PC(which may be also be used by an adult). But in case of the child’s phone, it is always being used by the child – hence more risk.

The point of these arguments is: there are genuine reasons to be cautious.

Why change and why now?

In the last few years since I have been tracking this space, why is the idea of opening up Telecoms networks suddenly of interest?

Here is the reason in my view.

Telecoms is a mature industry. Voice revenues still drive a lot of the business. Data revenues are small, indeed non existent. However, Voice revenues are under threat from VOIP – and that’s a universal phenomenon – on both fixed and mobile networks. (albeit on mobile networks today, it is still mainly voice over WiFi).

Also, in most places in the West, the markets are saturated. Thus, growth prospects for voice alone are limited. In contrast to the Web companies, especially with the uptake of Web 2.0, Telecoms is not an attractive investment proposition in light of the (lack of) future growth.

Hence, witness the interest in Fixed to Mobile convergence (fixed operators trying to poach customers from mobile operators and vice versa) and acquisitions in fast growing markets like India. And also an interest in Mobile data for the same reason.

Hence, market forces will drive the change in many part of the world. Other factors also help. Better devices, mature networks, better browser technology etc.

Case for Government intervention?

Tim Wu’s article concludes: “At some point, I think Americans are going to put their foot down and say, ‘We won’t tolerate this anymore.’”

That’s true

The real question is: What could be done and should the government intervene?

In fact, a government mandated environment leading to fantastic rates of growth and innovation does exist today. It’s in South Korea where the Ministry of Information and Communication (South Korea) or MIC Korea plays the overseer role.

But is that the best way? Is that the optimal path in the West?

Think about this, Korea and Japan have made great strides internally but have struggled (and will continue to struggle in my view) to export their Mobile/communications services globally. That’s the dark side of government mandated standards – high internal growth – low global growth.

Thus, excessive government intervention is the wrong thing in my view.

See my blog Should you be thinking of Vegas on your next flight to Tokyo or Seoul? .

Speaking of Governments, ours(UK) did intervene in the dot com boom and reaped a windfall through the 3G spectrum auctions . By saddling the Operators with debt, I believe the British government squashed a golden opportunity for British companies to take a lead in the Wireless space. We definitely don’t want more of that – Thank You!

The Web itself is not exactly free of monopolies. Take the case of Microsoft. However, with my Randian / free market view, I would oppose any regulation on a company like Microsoft. After all, if customers truly hated it, they could change(nothing prevents them from making that choice – much as the same with leaving a specific Operator). With Google applications , that may well happen – but the changes will be market driven and not regulatory.

The search for disruptive elements

The problem with the Mobile industry is: It’s still very arrogant. We talk of concepts like Mobile Youth – but the IPTV industry does not talk of ‘IPTV Youth’ or the fixed line industry of ‘Fixed line youth’.

Once we accept that there are only ‘people’ and they want to communicate irrespective of transport mechanisms – then we have to ask ourselves the question : Where can we get critical mass?

In my view, no single Operator can gain critical mass because the Operator’s subscriber base is fragmented along many dimensions, for instance devices, Pre pay-Post Pay etc.

So, disruptive elements can arise if we unify the stack across Operators along some dimension

At the application level, I believe that Web technologies will do this (especially full web browsers, widgets etc). At a network level, the combination of devices and WiFi/WiMax is the key.

These will happen organically – and we will always have to get used to working in a diverse ecosystem.

Co-operation as a solution

To recap, I believe that :

In a consumption driven world, competition is the driving force. In a creation driven world(User generated content, Web 2.0, Mobile Web 2.0 ), cooperation is the driving force. For reasons I explain previously, we have no choice but to accept a diverse ecosystem in the West and sweeping legislation is not a good option.

In that sense, unless America does some fundamental rethink, it may well lose out in the end because of it’s culture of competing vs. collaborating. In contrast, the European / GSM approach is more collaborative and suited for future growth

To conclude

1) We will continue to live in a diverse ecosystem and that is good

2) Operators will end up with Open APIs and that’s not a Pipe.

3) As we go from a consumption driven ecosystem to a creation driven ecosystem, cooperation will be a driving force and not competition. The need to communicate will overcome outdated business models.

4) Market forces are the main drivers to opening up in Europe(the need to show growth to the investment community for example)

5) In my view, government regulation is the wrong step(except in cases like Disclosure). A combination of Web enabled devices, devices supporting WiFi etc will drive disruptive applications.

PS: In an ironic twist, the March 26 issue of Business week’s best performers – has Google in the top spot. But guess who is on No 7? Yes, Verizon communications!. It indicates to me, that if customers want to leave a service, they will – but they don’t at the moment because they are primarily consuming content.

However, I believe that the future will belong to those Operators who open up because the creation driven ecosystem will demand that (in contrast to the current consumption driven ecosystem)

Image source: Image shack


A note on comments:

I have been having a lot of problems with Spam. Hence, comments are disabled. Please email me at ajit.jaokar at futuretext.com and I shall be happy to post your comments

Inferred identity using a skype directory ..


A couple of weeks ago, I met Joel Selvadurai founder of messagr at the launch party of Techcrunch UK.

Joel is working on an interesting application.

I have blogged about inferred identity using a hypothetical mechanism called PeopleRank.

Thus, I have always been interested in Identity and Reputation Systems

Messagr is an interesting aspect of inferred identity (derieved identity).

Essentially, its a Skype ‘directory’.

What’s special about that I asked Joel?

The basic skype directory does not give much (and also only what you have enteretd)

However, if you have embedded Skype in your blog, then knowing your skype id, it can then infer more information about you based on the tags associated with your blog

This is indeed a fascinating idea! Very much in line with my own visions of Identity.

Speaking of which .. I am attending the Digital Identity Forum in London tomorrow

If you are there, lets meet

Positioning Skype as IM


A fascinating article from the BBC. Full text alongwith link below

The bit I found insightful is

Yet instead of enhanced telephony Skype is focusing on enhancements like video calling which just make it more of a competitor to Google Talk, AOL Instant Messenger and MSN Messenger, all of which offer voice chat as part of their portfolio.

Hanging on the internet telephone

Last month the mobile phone industry held a major conference and exhibition in Barcelona.

Handset manufacturer Nokia used the opportunity to launch the 6136, a mobile that switches seamlessly between GSM and wi-fi networks and lets you swap between your mobile operator and voice over the internet as you talk.

Around the same time Ofcom, the UK communications industries regulator, published a consultation document on Regulation of Voip Services.

This gave us until May to comment on proposals that would bring network telephony operators under the regulatory umbrella instead of treating Voip as a fledgling service that needs freedom to innovate.

It has been clear for some time that voice calls over the public internet will change the way that the telephone industry works, and these two announcements show that this awareness has filtered through to the highest levels of the industry and of government.

A few years ago it would have been unthinkable for a handset company to offer a new phone that let users make free calls without using their network provider.

Vodafone, T-Mobile and the other networks would have let them know that such a phone was not what they wanted, and the idea would have been quietly dropped.

Now there seems to be a widespread realisation that lots of voice minutes are going to be moving online anyway, so it makes sense for a network provider to keep their customers happy, take what revenue they can get from the calls made over their network and look for other revenue-generating services to offer.

Emergency issue

The Ofcom consultation highlights the growing use of Voip in the UK, and suggests that Voip providers who want to be serious providers of publicly available telephone services should have to accept the regulatory framework that applies elsewhere.

Since every incumbent telephone service provider sees Voip as part of its offering, this creates a problem for pure internet players like Skype and Vonage.

There is a lot more to voice than just making phone calls, and at the moment Skype is at the bottom end of the curve for these advanced uses

One issue that is particularly important to Ofcom is access to emergency services through a 999 number, something that is technically tricky for Voip providers who are not also conventional phone companies.

Back in May 2005 Skype cut its links to the Norwegian telephone network for a time after the regulator there insisted that Voip providers should offer standard emergency calls. Instead of complying, Skype now accepts that it is not a telephony replacement service and hopes to escape regulation for a while.

However, Skype faces a much bigger issue than how it deals with requests for a 999 service. It may well have proven that the market is ready for internet calling but its architecture and the way it integrates with other Voip services mean that it is likely to be bypassed when network telephony goes mainstream.

Skype is a peer-to-peer network, with no centralised server to run or pay for. Even its directory is distributed over the network.

It manages this through a technical architecture and set of protocols that it has developed itself, outside the standards bodies which have been working on voice over internet for many years.

The dominant standards in this area are SIP, or the session initiation protocol, and H.323, originally developed for doing multimedia over local area networks. Both are widely used by voice providers, but not by Skype, which instead has its own proprietary standards for its peer-to-peer offering.

Skype certainly works, providing a service that is generally reliable, offers good voice quality and seems scalable to millions of users. But that may not be enough.

There is a lot more to voice than just making phone calls, and at the moment Skype is at the bottom end of the curve for these advanced uses.

Voice competitors

A lot of work is going on to create network-based analogues of switchboards which closely integrate voice services will all the other ways we like to use our computers, and attention has focused on the open source Asterix project.

As Cambridge computing entrepreneur Quentin Stafford-Fraser points out on his weblog, one such service, an open source application called Gizmo, lets him “have UK phone numbers which will forward to my Gizmo session here in California. For free. I can use Gizmo to call up my Asterisk server and listen to MP3 files and podcasts stored on my hard disk. For free. I can connect directly to Google Talk, or to dedicated Voip phones.”

Supermarket giant Tesco recently launched a Voip service

Closed services that are just replacements for conventional phones, whether they are provided by Skype or by the existing telephone companies, don’t do this.

Yet instead of enhanced telephony Skype is focusing on enhancements like video calling which just make it more of a competitor to Google Talk, AOL Instant Messenger and MSN Messenger, all of which offer voice chat as part of their portfolio.

That is not telephony and it is not a way to make money: there is a reason why Google, AOL and Microsoft appear so prominently in the names of their services and it is not that IM generates lots of cash.

And although Skype gets revenue from SkypeIn and SkypeOut, offering links to the phone network, it will be harder to persuade users to install a completely separate client and pay yet another intermediary for phone calls once these networks offer enhanced, standards-based voice services of their own.

Last September eBay paid $2.6bn for Skype in a move that many found hard to understand and which it may already be regretting.

Given the speed with which the existing phone companies have moved in on the Voip market there would seem to be two options for the new owners.

The first is to rebrand it as eBay Messaging, a voice-based instant messenger service with added phone integration.

The second would be to list this well-loved but proprietary Voip service on the world’s biggest second-hand marketplace.

I am sure someone at eBay knows how to do that.

Source: BBC

mobile wifi, dual mode handsets and voip – my analysis


With the success of companies such as skype, VOIP is seen to be a ‘hot’ technology. Following the hype surrounding VOIP, ‘mobile VOIP’ is also caught up in the hype cycle.

The idea is simple – cellular calls could be made over an IP network -thus saving cost.

Currently, mobile VOIP is synonymous with ‘voice over WiFi’ i.e. voice calls made over a WiFi network. Although the WiFi network is growing fast, the world is far from being a 100% WiFi enabled space. This means, mobile VOIP suffers from the physical limitations of being near a WiFi hotspot.

The requirement of being near a WiFi hotspot plus the high cost of Mobile WiFi handsets, means that mobile WiFi is currently a niche technology. It’s initial deployment is expected to be in the enterprise or within hotspots.

The real potential of mobile VOIP lies in the use of dual mode handsets. Dual mode handsets support the seamless handover between a cellular(in practise 3G and beyond) network and a WiFi network. The technologies used in this space are currently being defined for example – Unlicensed Mobile Access (UMA) and the Mobile Integrated Go-to-Market Network IP Telephony Experience (MobileIGNITE) alliance.

Predictably, the incumbents such as mobile operators, are reluctant to support mobile VOIP because it’s a threat to their existing business(mobile voice calls). However, companies from outside the

existing value chain are keen to promote mobile VOIP. Most notably fixed line operators and handset manufacturers

However, the biggest barriers to the uptake of mobile VOIP is the pricing for IP traffic. I believe that the technology will really take off only when cheap, ‘unlimited use’ IP traffic becomes possible.

Thus, as with so many services in the mobile data industry, the barriers are not technological but commercial. There is no doubt that mobile VOIP will have a part to play in the evolution of mobility in general. Its eventual success and role will depend on a range of technical and commercial factors some of which are outlined above. However, its real significance lies in the fact that it will put a downward pressure on voice call prices (which is still the mainstay of income for mobile operators).

If cheap unlimited use bandwidth becomes a possibility then the market may well take off in other directions. For example – it could be possible to make voice calls from a 3G network through an IP client on the phone without going through a dual mode handset etc. The success of such schemes depends on low costs for IP traffic. However, note that the technology exists even today to make this possible.

Ironically, the mobile network itself is shifting to an IP core with technologies like IMS. When that happens, it should be possible to make end to end VOIP calls!

What’s your view? Is this balanced enough or have I missed something?