Paul Golding: Operators

Wed, 09 Feb 2005 11:55:34 GMT

@ If you're still reading this....

...you shouldn't be. Your browser should have been redirected to my new blog

http://wirelesswonders.blogspot.com/

Sat, 31 Jul 2004 12:33:09 GMT

@ Location privacy...

When teaching or discussing the principles and uses of location-based services, I am often asked about privacy and legal implications. I usually give anecdotal answers based on what I perceive to be operator approaches, which I assume to be best practise and legally valid. However, I did manage to track down the actual legal position in terms of rights to privacy enshrined within European Law.

Article 9 of the EU Privacy and Electronic Communications Directive requires that location data may only be used with the consent of the subscriber. Moreover, it should remain possible for subscribers and users even if they have subscribed to a location based service, to temporarily block the tracing facility.

Thu, 29 Apr 2004 00:42:09 GMT

:: So much for interference in the cabin ::

"Image you fly in an aircraft - your secretary calls you on your mobile to tell you that your PowerPoint presentation was changed by your boss. You switch on your laptop, connect wirelessly to your company intranet, and download the file.

An illusion? Not in two years time.."

So says the Wireless Cabin project, a European consortium, which means GSM gets priority - quite rightly too, but not without WiFi and Bluetooth also included. Interesting stuff. No community WiFi though - hard nosed profit-making stuff licensed from operator partners who own the spectrum, even inside airplanes apparently.

Sun, 03 Aug 2003 13:23:19 GMT

:: Mobile browsing security threats ::

Mobile users should be careful about responding to messages (i.e. WAP push) that take them to wap sites. They may just be fake.

In my forthcoming book, I've just started writing about securing HTTP and WAP connections. When I started thinking about the vulnerabilities of authentication, it occurred to me that mobile sites are perhaps a lot less safe then their desktop counterparts.

One way to gain a password from a user is to spoof a server. Pretend to be a particular website and then ask your users to log in. Voila! If they bite, then you have their username and password. This has been done on numerous occasions with famous websites and is an ongoing threat. In fact, regular users of sites like Ebay should learn to become vigilant against these types of spook attacks. With mobile sites it appears to be even easier.

Firstly, on most mobile browsers, to save real estate, the URL display box is not displayed, or there simply isn't one. This means a user typically has NO idea what website they are actually on in terms of its web address - surfing on mobile sites is sometimes an eerie experience, like walking in the dark. If a user is directed to a spoof website, they would have little or no idea.

Secondly, due to sparse interfaces, it takes little effort to mimic a mobile website, perhaps just by copying a logo at the top of the screen.

WAP push will soon start to become more widespread. It has been slow to catch on, but with the increasing number of picture-messaging phones, the necessary inclusion of the WAP browser means that more and more mobile users can access mobile sites, whether they know it or not. However, they don't need to know it to respond to a WAP push message - the phone itself will take care of accessing the embedded URL in the message if the user chooses to respond.

The other weakness is with the WAP push mechanism itself. It uses text messaging as the transport mechanism. With text messaging it is easy to change the sender's address in the message. In fact, many text-messaging bureaus offer this service to their bulk-messaging customers. This has some interesting and legitimate uses, but can also be malignant. It is easy to spoof the sender in order to make the message look legitimate, thus adding to the bait used to lure an unsuspecting user onto a spoof site.

Mobile users should be educated in the dangers of responding to WAP push messages.

Sun, 22 Jun 2003 22:27:29 GMT

:: Usability ? ::

Using the Vodafone Live service, I was sent a message that told me to switch my phone off and on again, as the menus needed to be updated. I did that. I then looked to see what had changed, but couldn't see anything obvious. There should have been an indication as to the new features, as the prospect of new things to do on the phone obviously gives one a sense of anticipation.

Anyhow, I had a surf around the Vodafone Live site and entered a link called "Messages" as I was curious what messages were being stored there. One further option was called "Vodafone Messenger", so I thought to try it, only to be delivered the message shown in the screen shot here. Interesting that I'm surfing the site as a registered Vodafone customer, able to interface with the site via WAP, but I'm required to go elsewhere (i.e. a PC) to register?? Seems a tiny bit odd to me. I'm sure there's a perfectly good explanation in the mind of someone at Voda.

Tue, 13 May 2003 13:41:49 GMT

:: WiFi -- 3G ++ ?::

At last, someone talks sense about the commercial realities of the WiFi v 3G "debate"; in other words, someone has to pay for it.

http://www.theregister.co.uk/content/69/30651.html

Sun, 11 May 2003 17:12:01 GMT

:: Service creation platforms ::

The problem looming is that with all these different potential services - foneblogs, picture galleries, email etc, it would be a shame to have lots of different sites to access. Operator portals are an obvious place to coalesce these services and operators need to think about integration of 3rd party services at the portal level. In my mind, there is a product opportunity for someone to supply operators with a decent service-creation portal that can enable any combination of useful applications to be brought together on a targetted user-group level. Just like community sites like Ecademy site are built using Drupal (or something like Plone - cool if you like Python), wireless portals should be possible using a similar concept, but the difference would be a charging concept that would enable the 3rd party to share revenue.

This idea is what we (Magic E) had originally intended with Zingo when we created the world's first wireless portal as a demonstrator for Lucent's 3G capabilities (a long time back!). We should have stuck to our guns as I think the time is right for this! Any takers??

Sat, 10 May 2003 16:29:54 GMT

:: Accessing Outlook via wireless (from O2)::

A subject of ongoing trial and error, wireless email is still taking its time to become useful. Having spent a lot of effort myself designing different systems for enterprise usage, this latest product from O2 is somewhat familiar in concept. It is based on the idea of leaving your desktop running so that a proxy system - in this case from Seven - can access your actual email account using Outlook as the programming interface so that the wireless interaction is with the actual Outlook data and not a copy of it. O2's licensed version is called xmail and it has been optimised for the XDA. However, it works with any browser (not that many have been tested though apparently) and is available to non-O2 customers (though there is a note that this may one day become O2-users only).

The service costs 9.99 (GBP) per month (excluding airtime), but there is a 10-day trial to see if you like it before paying. I downloaded the trial and it installed smoothly.

I accessed my Outlook information from Internet Explorer first of all and that seemed to work well. This in itself is a very handy feature - I now have webmail for my actual account (including all legacy data already sitting in my Outlook PIM database).

The big disappointment however was accessing the system via WAP from my T68i, which does not seem to be supported as I can't get past the log-in screen despite correctly entering my credentials. Having designed WAP email systems before, I have seen this problem many times and it is indicative of poor testing, but then this is a problem that plagues WAP designers - the shear number of devices to test against, which is a necessity even if the browser is licensed from a 3rd party and supposed to behave consistently across all phones (which they don't).

It did work via Eudora Web browser running on my Palm (with Bluetooth back to the T68i). The problem I am experiencing though is poor performance, seemingly because the SSL processing on the Palm is slow, at least in the Eudora implementation, so I am going to try another browser.

A nice feature is that a documents folder can be chosen to be accessible via the browser interface so that attachments can be added. This could be something like "My Documents" for example. However, as with all programs that interface with Outlook, folders and multiple accounts is always a problem, so neither are supported here. That means that I only get to see messages in my Inbox and is problematic for my VIP messages that are routed to folders. So what do I do? Should I turn off my folder routing in order to benefit from this product, or should I make do with the compromise? The only other alternative is to switch to a more sophisticated system like Blackberry, but I'm not interested in a proprietary device as I already have a Palm with Bluetooth cradle. If anything, I would move to something like a P800 if I was going to "trade up".

I shall see if it is worth spending the tenner a month to stick with this service. A 10-day trial is a bit limited and nothing like the standard 20-30 days used by most software vendors, but then operators are not that clued up on such things. Currently I check email from my T68i using IMAP, but I frequently leave Outlook running on my desktop anyway where it will suck down the mail every 5 minutes thus resulting in nothing being left on the server for me to view remotely. This is why I wanted to try out the xMail system. A similar product is available from Orange, who has also licensed the Seven solution. Products like this have already existed for sometime in the US market, such as from companies like Infowave.

Wed, 16 Apr 2003 10:27:00 GMT

:: Zingo Taxi - The reality of location services today ::

The BBC website has an article with their own review and reader feedback about the new Zingo taxi-hailing service now running in London. It suffers from various usability and reliability problems (I'm going to post an update on location technology soon). For this first time, I decided to "draw" my own cartoon...a picture is worth a thousand words (text messages?)

Fri, 04 Apr 2003 01:29:32 GMT

:: "Death of 3G ?" ::

I've been holding off making any comments about Clay Shirky's "death of 3G" siren, as a Techdirt posting (below) comments on. Maybe I'm too scared that it's true. It could be, but I wanted (and need) more time to compose a lengthy reply, which I may never get time to do. For now I wanted to dump some thoughts whilst fresh in my mind after pizza with a few guys in the wireless industry whose views I value.

Whilst Clay's analogy seems attractive, I am not so convinced. One could argue the merits of the "permanet" versus "nearlynet" model and the examples chosen, but I think the real problem here is that we need to work the numbers on this, just like we need to work the engineering through on the whole open spectrum debate that is perhaps struggling to climb out of its lofty political origins.

Why do I mention these two issues side by side? Well, there is the matter of quality of service that seems to get overlooked when comparing a controlled and planned network versus an ad-hoc one. Professor Lawrence Lessig’s road versus railroad analogy may not be the most apt if you happen to live where traffic jams are the rule of the day! This needs a lengthy discussion in order to appreciate its significance, and that discussion needs to take place between experts better than myself, though my hope is that the realm of analogies gets replaced by serious proposals.

Hand-waving myself for a moment the issue with WiFi hotspots is that they are just not going to do the job. Before long, users of hotspots will want hot-zones and before long hot-regions. Wait, isn't that 3G?

If the sociological shifts predicted (observed!) by Howard Rheingold are what the near-future holds, then users will become part of an information world that is pyschologically very time-sensitive, be it the need to know news about terror threats or, more likely, the need to satisfy a whole plethora of social appetites that we have yet to understand.

This time-sensitive info-social networking means constant (wide-area) data coverage is essential. In terms of economics I have my own views on why the value perception for such access will be markedly different to what we are seeing right now for mobile voice (hence why such comparisons are of limited value). The clue is in the idea that social status, well-being and wealth generation will be stronyl related to conducting our "business" online and that mobile access with continuity will become a valued essential and very much demanded. Quite possibly, the distinction between "business use" and "consumer use" will blur, further upping the ante for constant connectivity.

In short, there will be a viable economic case for 3G. What it can offer is so incredibly valuable and won't be on sale anywhere else - certainly not via a patchwork of WiFi hotspots. These definitely have value in the short-term and will always have a place in our wireless existence, but they will never replace the need for a wide-area mobile network.

Sometimes we seem to forget where we are coming from. The problem right now is that for wide area mobile access, all we have is GPRS, which is a voice network with data grafted on its back. Hence why peformance is poor and not likely to improve. So we need something better and that is 3G.

Now the current economic factors that are weighing 3G down are not going to stop 3G from going ahead. It seems inconceivable that for the next 10 years we are going to sit back and do nothing whilst a revolution in our work and social habits takes place that demands time-sensitive ("always on") access. Hotspots are not going to make it - they will be too sparse, too poorly maintained and never be able to offer the quality of service required.

I caution myself that these issues are not to be lightly tossed in the air - as I have done. It is not a matter of clever argument and guesswork, resorting to analogies and irrelevant history to make a case. We need to contrive various models of usage for future mobile-community scenarios and take a deep look at the engineering issues. Perhaps if we did this, as I know some people are doing, then we can focus our efforts where it makes sense to do so, for all our sakes. Afterall, for all our sakes, it actually makes sense that we get 3G - and WiFi - to work effectively well .

But for now, I am a "Field of Dreams" believer for 3G - "if we build it, they will come". Actually most operators can afford to bankroll this for a long time yet. Probably the best thing we can do, 3G pundit or not, is to talk up the confidence in the wireless future and attract more investment into 3G, not more pessimism. All those investors who have shyed away from technology stocks should be piling their money into wireless - it definitely has a future!

p.s. My associate at IPWireless tells me that 3G has a future "only if it's TDD" (but he might be biased :-).

-----------------------

Permanet, Nearlynet, And Wireless Data. As per usual, an insightful article from Clay Shirky looking at "permanets" vs. "nearlynets". His argument is that "permanets", networks built by a few large entities tend to fail when competing against "nearlynets", networks that are cobbled together randomly by individuals. His argument is that "permanets" are usually high quality - but expensive, and nearlynets are usually lousy but cheap. However, over time, nearlynets improve much faster than permanets get cheap. This makes sense. Any nearlynet has incentive to get better. Permanets, though, have to recover their high initial capital costs, and thus, have less incentive to get cheap quickly. As an example of failed permanets, he talks about airplane phones and Iridium, both of which make his case perfectly. In both cases, lower quality, but cheaper mobile phones took away the ability for the permanet solution to make money. Now, he says, that 3G is the next permanet, and WiFi is its competing nearlynet. Of course, when you think about it, it's a little odd that mobile phones were the "nearlynets" in his initial examples, but suddenly become the "permanet" in his prediction. There are, though, differences between regular 2G mobile phones and the 3G data plans that carriers are betting on. We've said in the past that the 3G providers are likely to price their offerings terribly (toll booth style, rather than flat rate) and Shirky points out that this is likely to push people towards the "nearlynet" of WiFi. Definitely worth reading.

Source: [Techdirt Corporate Intelligence: Techdirt Wireless]

Fri, 04 Apr 2003 00:12:37 GMT

:: Operators "become handheld OS companies" ::

How timely that I should read this article today. I was just writing (in my new book) about the nascent shift in mindset that the operators are going through, some would say from telecoms mentality to IT mentality. The investment mentioned below by Vodafone and Orange would seem like a bid to get directly involved in the IT action to claw back some of the revenues that they would say they bring to handset OS licensors. On the one hand I'm glad they have gone to SaveJE (pronounced "savij"). I was an early adopter of SavaJE on my iPAQ, wiping Pocket PC clean from its memory only to find that this particular version of the SavaJE OS wasn't to be supported any longer by SavaJE - very savage indeed!

I just hope that this is not a bid by these operators to control the end-point of the network and tie-down all the value added services

Vodafone, Orange Invest in Wireless Java OS. Wireless carriers are increasingly getting into the software business as they look for more ways to differentiate their offerings and gain some leverage over Nokia, Microsoft and Symbian. As the market moves toward smartphones software becomes more important to carriers and handset makers. One way or the other we'll end up with a dominant OS that is used across carriers and handset makers. The question remains which one and when. Vodafone and Orange are doing their part by investing is Java start up SaveJe that is working on a full Java OS for handsets. A Java OS would allow Vodafone and Orange to commission completely custom handsets with their own unique interfaces (not necessarily a good thing) while still allowing the devices to run common java apps.

Source: [Techdirt Corporate Intelligence: Techdirt Wireless]

Mon, 24 Feb 2003 22:47:36 GMT

NTT DoCoMo To Add Flash To Phones

Japanese mobile phone operator NTT DoCoMo Monday announced that it will embed Macromedia's Flash technology into its i-mode handsets. (Company)

[ Full Story - Thinkmobile]

Sun, 23 Feb 2003 15:15:07 GMT

"A Web Site for Every Mobile Phone"

NewBay Software launched a "Wireless Weblog" product at the recent 3GSM show in Cannes.

To weblog pundits, this should have some appeal, but to what extent it has wide appeal to mobile users is another matter. Of course, one can imagine that the functionality is limited, especially submitting blog reports via a text-message. But then it seems that NewBay have their eyes more on the multimedia messaging market. It is possible to post pictures and text to a blog.

It is not really a product aimed at "serious" (existing) bloggers. More likely it is an attempt at offering something for mass-consumption that would be white-labelled by the operator as part of their standard portal offering, or bundled into a multimedia services package, like Vodafone Live. Existing bloggers may opt to use interfaces like email, such as is already built into UserLand's Radio, albeit in a crude form and not able to handle images, but that is not much of an challenge to incorporate. Multimedia messaging by default is able to send MIME packaged messages to any email address, so a small script to extract the text and the image would suffice to add the message contents to the current blog (but don't ask me for the details as I don't write scripts for Radio).

At the heart of blogging is the passing of links to readers so they may delve into the subject matter more deeply. On standard MMS phones it is not possible to insert URLs into the message. In any case, it is difficult to imagine what kind of blogging would be done on the move from a phone, which probably suggests that such activities will be of limited interest to mobile users in general, although specialist "on the move" blogging activities can probably be envisaged.

No doubt this product, or something like it, will appear in some operator portal offerings, if only as an additional sales hook for picture messaging phones. Moreover, at the moment everything needs to be tried out to see where user interests lie in using next generation wireless services.

As NewBay confess in their own bumf:

"How will People Use FoneBlog? - We don’t know for certain - but we’ve got some great ideas."

Tue, 04 Feb 2003 09:39:50 GMT

SIM-Based Roaming 2/3G to WiFi Hotspots [InternetNews]

HP, with their SS7 gateway and location register (HRL) platform - now extended to support GSM - have teamed up with TOGEWAnet who provide WeRoam roaming infrastructure to enable GSM-WiFi roaming based on secure SIM-card authentication.

What this means is that a user can roam from cellular to WiFi LAN services and remain on one account, paying one bill. The roamed-to hotspots will be paid by transferred accounting between the user's service provider and the hotspot provider(s).

By offering a solution based on SIM-card authentication, this makes it attractive to operators. As TOGEWAnet claim on their website:

"In October 2002 WeRoam® has successfully been implemented in the core GSM network of various operators to enable these operators to have a swift and cost efficient entrance into the world of WLAN."

The WeRoam solution interworks with RADIUS to complete the authentication onto the hotspot network. There are numerous benefits to this approach.

Sun, 02 Feb 2003 15:52:34 GMT

Sun "takes back control" of J2ME [Sun's Java Website]

Those of us engaged in wireless services were beginning to wonder exactly where J2ME was headed, or more accurately MIDP, and why it didn't seem to be on the right track. Clearly MIDP 1.0 is woefully inadequate. It had many missing features just in terms of its basic capabilities, nevermind the higher system functions that a mobile environment must have - like a push mechanism, interactivity with the user's PIM functions and so on. Meanwhile, MMS and video phones have pushed the services capability of phones beyond what MIDP 1.0 is capable of. Hence we have ended up with a situation where Vodafone have released their own extensions (NOT open source) and the Java Community Process is hosting apparently unconnected efforts to usefully extend MIDP: such as the Wireless Messaging API (JSR 120) and Mobile Media API (JSR 135) that run on top of MIDP 1.0.

Clearly other extensions are needed, like location-based API capabilities, better authentication techniques, MMS handling and so on.

Thankfully, Sun has recently announced a new industry initiative and specification for the future of Java on the handheld devices. This is called JavaTM Technology for the Wireless Industry (JTWI) and is running within the Java Community Process, with good industry backing, as specification JSR-185.

In their roadmap, they seem spot on with considering the following questions:

Given the potential number of optional JSRs that might be associated with MIDP, how can we ensure that each is architecturally coherent with the other pieces in a handset?
How can content developers and operators know when new APIs will be available in real deployments, so they can take advantage of them?
When would a product manager be able to rely on those APIs to define them as features in new devices?

Any operators with a keen interest in generating revenue from MIDlets should probably be involved with this process.

Thu, 23 Jan 2003 23:51:05 GMT

CAUTION: WiFi is a product, NOT a service.....[Clay Shirky's website]

In my opinion, WiFi (wireless LAN) is definitely a threat to 3G (i.e. UMTS and other wideband wireless services). I plan to write on this topic in more detail in my forthcoming book (more details soon....).

But this article by Clay Shirky has some interesting observations about WiFi being a product and not a service, hence why it doesn't seem to really offer any ongoing revenue potential (certain limited hotspot opportunities notwithstanding).

Wed, 08 Jan 2003 14:19:13 GMT

Just like DoCoMo in Japan, Vodafone have decided to offer their own Java libraries for J2ME. Hopefully this will be made redundant by MIDP 2.0 and the Mobile Media extensions. I plan to look at this in more detail to see what Vodafone is offering that these new profiles and extensions will not.

From Vodafone VIA website - "With the all-new Sharp GX10 camera-phone, which is the Vodafone Live! flagship handset, Vodafone has for the first time launched its own state-of-the-art flavour of Java - VSCL - specifically designed to deliver and support breathtaking animated games with 16 chord polyphonic sound, vibration support, 2D & 3D graphics and other multimedia content....."

Fri, 13 Dec 2002 22:27:50 GMT

Three (hutchison 3G) in the UK have announced their prices for 3G services.

Tue, 26 Nov 2002 22:33:48 GMT

GPRS is unable to pay for lunch [ZDNet]

Snippet:

......"He noted that the revenue brought in by GPRS among the four Denmark network operators in the past year amounted to only about 23,000 euros (roughly £14,000), which would not be enough to pay for the lunches of the engineers that installed the GPRS equipment."

O dear! But the value of the article is in the suggestions for content-sensitive billing. A simple idea that has so far - amazingly - escaped the operators attention. I thought these ideas were well known, apparently not. Can't say that the billing systems don't exist for content-based billing - isn't that what the Internet is supposed to be all about? The same applies for location-sensitive content. I wrote extensively on this topic in a report done for MetroWalker location service provider in Hong Kong.

Tue, 26 Nov 2002 22:26:20 GMT

Wi-Fi hot spots in UK get mapped [ZDNet]

Not quite extensive coverage, but early days I guess!

Tue, 26 Nov 2002 07:43:05 GMT

3G and Wi-Fi: Will they Ever Get Together? [Think Mobile]

One has distance, the other has speed. Together, they could make a pretty sweet pair -- if they can make the relationship work not just technically, but financially.

Wed, 13 Nov 2002 00:23:09 GMT

3G is dead???

Marty Cooper, "inventor" of cellular, says that 3G makes him yawn and it's virtually no better than GPRS. Not sure what he is advocating otherwise (needs looking into), but he reduces our attention solely to the data rate and leaves out all other features of UMTS that could be noted. This also tends to happen a lot with advocates of WiFi dismiss 3G. I'm not saying this is where Marty is coming from, but dismissals are usually along the same lines - 3G is too slow and who needs it now anyway post-WiFi. Well, I may have misunderstood something about the US wireless market I have to admit, but I frankly don't understand the comparison.

Mass adoption of wireless applications of one form or another is, in my opinion (as someone who spends a lot of time talking to operators and training people in the technology) inevitable. And I mean mass adoption, let's say at least 50% of users in today's saturated mobile markets will be using wireless apps (richer than text messaging) within the next 5 years. Some of the exciting stuff we have yet to see - well, let's be frank, we have yet to see most of it and maybe that's why Marty is yawning. Here I agree with Marty that the operators seem to be the slowest link in the value chain, but through their rapidly maturing developer programs they seem to be heading in the right direction. Of course I am talking about the UK here and European operators are probably in similar positions.

The point is that most people won't need huge data rates, they just need reliable truly always-on service, which means being available wherever the user might be. That is what 3G networks are being rolled out to provide, as well as better voice capacity and service. UMTS is WAN not LAN, and WiFi is definitely not WAN, even taking into consideration the possible proliferation of public access points (and that has many issues that put it back in the dark ages compared with the mobility, billing and security management mastered in GSM some 10 years ago, nevermind that I have to make sure I'm in Starbucks when I want to check my email or update my blog).

Now are we expecting that these mass-adoption users are going to be accessing mobile applications on PDA's over WiFi? I don't think so. Will most of them ever need the promised high data rates of 3G. I don't think so either. But collectively they will need a reliable and sufficiently beefy mobile data network that only 3G can provide, certainly not GPRS which is really a GSM afterthought, so it is wrong to take the stance that we have GPRS, so why bother with 3G. This could be discussed in a separate article.

On a related point, I am still slightly baffled by what the "open spectrum" advocates are really driving at when they point fingers at the "monopolistic" hoarding of spectrum by operators. I'm not saying Marty is on this wavelength, but I am reminded of some of their false claims (not including the technical ones to do with the real capacity of specturm being greater than we give it credit for - not sure where this new bit of information theory has come from). The problem with some writings I have seen is that there is an assumption that an open free-for-all market will provide the best possible connectivity and service for end users? Says who? Where can we find this theory being practised? Comparisons with the Internet are false ones. Ideas such as standards and competition through open standards seems to have been missed, like the success of GSM thanks to the huge commitment of operators and massive investment by the network infrastucture vendors. The technology needed in digital cellular has always been on the very bleeding edge of the cost-effective capabilities of silicon. Just the air-interface processing alone was such that a GSM handset when they came out was hugely more powerful than the early Pentium PC's. The economies of scale for GSM were by virtue of the pre-agreed commitment to the standard, but I am prepared to be corrected. The affordable end-user economies were by virtue of the competing operators in the same markets - a level of competition that remains highly aggressive in most GSM markets. I think these considerations are real ones when debating further opening of the spectrum, which incidentally I agree with in principle.

But what people seem to misunderstand totally is that mobile telephony and WiFi are totally different things - completely different.

One particular point to note is that by and large mobile telephony is exceptionally high quality and robust with mature 2G systems. The mobile telephony service of most GSM operators is incredibly predictable. It works nearly all of the time and it works well. The phones work incredibly well in most places you try to use them these days. And how many times do you have to "reboot" your mobile phone? How often do you need to resend a text message? A lot of people seem to forget that it is within this very stable and so conservative environment that operators are trying to launch unknown, largely untested services. I for one was shocked with the relatively fast adoption of J2ME, or more accurately MIDP, or even more accurately MIDP 1.0!! In fact, MIDP 1.0 is missing all of the useful features that would ordinarily make it acceptable for deployment in such an environment - all the features that you will find in the MIDP 2.0 spec (see my earlier posting). Technical support for this stuff is going to prove challenging. The very real possibility of hitherto unknown weaknesses, say for hacking and viruses, could cause real problems in mobile markets. Many mobile phone users are not hardened PC users who accept (or do they?) the rough edges of modern software practises.

This is perhaps where the misunderstanding lies in the false comparisons between WiFi and 3G. WiFi is just a technology. 3G is technology, yes, but it is intimately linked with service. It is accessed and paid for as a service by end-users, not as a piece of technology. The open-spectrum advocates who look with excitement at WiFi and what else might be possible, are possibly confusing these ideas. Open spectrum may lead to some interesting technologies emerging and yes, we may benefit from such innovations. But a useful service does not necessarily follow. And here again, the comparisons with the Internet do not apply. The main service paid for in the Internet world is ISP access and being crude, the levels of skill, expertise and investment to become and ISP versus a mobile operator are incomparable. But i stand to be corrected. Open spectrum may not be the answer to effective mobile communications, but there arguments are probably more complex than I have made them out to be here. (I usually write postings late at night!)

I am frustrated by the sedate pace of 3G and wireless data services getting to the market, especially as what is needed is innovative applications, but their have been very little incentives for innovators to bring their wares to the operators' doors. Pre-GPRS, operators were just glorified utility companies with huge marketing budgets and clueless about data. They are still struggling to make the transition. But for my money, 3G is definitely needed and will provide a whole raft of useful services to end-user in a mass-market situation. It is certainly not dead, unless we are talking about the dead before becoming alive.

Tue, 29 Oct 2002 20:24:02 GMT

DoCoMo soars by tapping power of emotions.

''In Japan, nearly one-third of the population already works, plays, and shops with wireless, continuously connected to a universe of data, services and other people. The force responsible is a young, visionary company, with a name that means `anywhere' in Japanese: DoCoMo.'' This quote shows just how pervasive DoCoMo is in Japan. In spite of its short life of a mere three years, it's already a major force in the country, with 30 million users. What is DoCoMo and why is this company so notable that a book should be devoted to it? The subtitle of John C. Beck and Mitchell E. Wade's book dubs DoCoMo a ''wireless tsunami,'' a moniker that is indicative of the influence of the company in Japan's moribund marketplace. [Mobilog]

Wed, 23 Oct 2002 18:46:20 GMT

Viruses on Mobile Phones

There has been some concern about viruses attacking mobile phones, or wreaking havoc from the phone, such as making unwanted calls, firing off errant text messages or sending lots of useless packets over the GPRS network. With current phones this is difficult to envisage, and most stories that circulate seem to have no substance to them. There have been attempts to deny service by bombarding network users with text messages - one guy was fined for this in the US, but that was not a phone-based virus.

The main concern comes from the new breed of phones that are able to run Java programs called MIDlets, named after their adherence to the MIDP (Mobile Internet Device Profile) Java specification. The MIDP 1.0 specification does not allow for MIDlets to access resources on the phone, apart from the screen of course displaying the user interface. Most MIDlets are games. However, handset manufacturers can release custom extensions (Java APIs) to MIDP that allow access to phone resources, such as text messaging, like the Nokia SMS API for the 3410. These could pose problems depending on their implementation.

The MIDP 2.0 specification, recently released, brings several key technologies to the aid of operators hoping to avoid rogue MIDlets:

Trusted MIDlet verification using code signing with X.509 certificates
Ability to for operator to set permissions policies for which low-level APIs can be access by a MIDlet
Default prompting of user to positively acknowledge that MIDlet is allowed to send a message or initiate a call

These measures seem to provide a very secure safety net to prevent the spreading of harmful MIDlets or other obnoxious effects from untrusted programmers.

The idea of operators controlling who gets to do what on a phone may seem irritating to some. It certainly seems the antithesis of an open system of the likes being promoted by Kevin Werbach with his views on open spectrum. But we have to consider that we are not introducing games and applications into an elastic computing environment - we are introducing them into a previously sterile and largely consumer-orientated environment where users are familiar with placing calls and sending text messages in a predictable and highly robust manner. In that context, it is understandable why operators are concerned about rocking the boat with anything unpredictable or hitherto unusual to the users - a virus attack would certainly cause a major shockwave across any operator's stable customer base. This context needs to be understood before we advocate throwing out control, licensing and sanitation as much as that might be upsetting to net "anarchists".

Fri, 18 Oct 2002 08:40:40 GMT

Seek + Find

So, with Barry and Dave, we tried to use Voda's Find+Seek service whilst sitting in Pizza Hut. Remarkably it managed to find us! Yes, apparently we were only 1.1km from Pizza Hut so we thought we had better move to that one instead :-)

Again, it gave a precise address as to our whereabouts - "2 Blabla Drive" - giving, as I noted earlier, the impression of accuracy only in need of some calibration. Cycling through the "Where am I now" option, it jumped around in where it thought we were, at one point actually getting quite close. Having noted that the reason for "pin-pointing" us at a precise address is to give a real and locatable reference point, at one point it was just giving "you are in Blabla Way" i.e. with no specific address. Thinking about it this morning, that was because Blabla Way has no residences or buildings actually on it - they all hang off of adjacent roads or plazas, including - one of the BIGGEST AND MOST OBVIOUS landmarks on Blabla Way - PIZZA HUT!! Had the geo-coding had a degree of "intelligence" built in, like suggesting obvious landmarks where there might be one (I mean everyone knows what Pizza Hut looks like) it might just have found us.

Another peculiar feature in the "find nearest...." option was a huge mismatch in how I versus Voda define nearest. When listing restaurants, it listed every single type in the directory (never mind that these were alphabetically listed as opposed to popularity ordered - to save paging around in WAP world - like Indian and Pizza should be near the top). Included in the listing was Greek, so I hit this option and it gave me a place a mere 56km away. Great! In my opinion it simply shouldn't list categories that have no restaurants nearby, but then what is near? [Full discussion of radius-of-interest versus radius-of-opportunity was included in my feasibility report for Metrowalker, and is also included in my course.]

The database was in any case devoid of any qualitative information, so had it told me that that particular Greek place had the finest smorgasbord of wares, then yes, I may have been tempted. At the end of it all, just when you find the thing you are looking for, the WAP page informs the unsuspecting user that he now has to pay 35p to get the information. Fair enough if that's their way of making money, but considering I just spent 110p on GPRS packets faffing around with a hopelessly designed system, I think that was rubbing salt in the wounds.

Not to be completely put-off, I was curious what the text message - which is what they send you for 35p - was going to tell me. Each time I got to one of these 35p walls, it mentioned "additional information" was part of the package as well as the address and phone number of the place of interest. So, out of curiosity, I located the cinema nearby and asked for the details. To my horror, nowhere, not in the SMS or the WAP page did it mention anywhere the films being played. Maybe - most likely - that's in another menu somewhere in la la land, which is where wireless data services is going to end up at this rate with such poor usability.