HTML5 the new desktop operating system?

The Defence CTO is reported to be interested in thin client computers linked to database applications hosted in data centres, but able to be used offline for mobile users. This would allow expensive, difficult to manage, power hungry desktop computers to be replaced with low cost units. The data could be in remote data centres under careful management, which appeals to Defence's needs for security.

Building thin client applications is easy: it is what we used to call "client server". I was part of a defence-industry project to build such applications in the early 1009s (I even published a scholarly letter about it: "Remote Presentation Client-Server More than Just Screen Scraper", Worthington T., Australian Computer Journal, Volume: 27 Issue: 1 Pages: 16-16 Published: FEB 1995). However, just as we got the graphical user interface to work for the remote client, along came the web and made such an interface a commodity item.

Using the web as the user interface for a remote database application is now an everyday application of technology. However, allowing the client to function while disconnected from the central database is very much more difficult. This requires having part of the application running in the client, along with some form of database.

Having an offline application would have required a complex bespoke software setup in a desktop computer and would have not been feasible in something which could be called a "thin client". However, in a review of the new HTC Desire Android mobile phone, I noticed that it used the Android 2.0 operating system, which has improved support for HTML5, including:

• Database API support, for client-side databases using SQL.
• Application cache support, for offline applications.

This is what is needed for building offline applications. It should be noted that this is in a mobile device, so building the same support into low cost desktop devices should be feasible.

The Defence CTO's desire for offline thin client applications now does not look that difficult: Use HTML5 for the user interface and provide local database and cache support. This will require as much hardware as a $400 netbook computer, running some sort of cut down Linux operating system, such as Google Android.

However, there are still some difficult issues for designers of such applications. Not all data can be made available "offline" as this would require each thin client computer to have as much storage as Defene's largest data centre. Security considerations would also limit access to offline data. One of the benifits of a thin client computer without offline storage is that when it is offline it contains little or no sensitive data. Thin clients with offline storage will have to be protected the same way as other computers holding sensitive data. Defence issues very detailed instructions on how to protect data, including how to destroy the devices (including what type of hammer to use when the enemy is at the gates).

Also with offline data there is the issue of synchronising with the central database. There is no foolproof what to solve this problem and each application will have to be designed accordingly.

Lastly a thin client computer is "thin" because it is not as powerful as a "thicker" one. There will therefore be some applications not suited to these devices.

Whole of Australian Government Mobile Tender

The Department of Finance and Deregulation has issued a Request for Tender for mobile phones, smartphones, wireless broadband modems and services for all of the Australian Government. There is a 1.4 Mbyte document available with the details.

The RFT covers:

• Mobile Carriage;
• Mobile Devices: including mobile handsets, smartphones and mobile broadband modems;
• Mobile Accessories; and
• Associated Services.

There is a requirement for one or more service providers to provide Telecommunication Commodities, Carriage and Associated Services to the Commonwealth. It is expected that the outcome of this RFT will be a panel accessible by all Agencies.

The Commonwealth aims to establish an arrangement that is flexible, efficient and responsive to changing technology and business requirements. ...
Timeframe for Delivery 3 years plus 2 separate options to extend for 12 months on each occasion. ...

From: Request for Tender for a Whole of Australian Government Panel of Telecommunications Commodities, Carriage and Associated Services, FIN10/AGI002, Department of Finance and Deregulation, 25-Mar-2010

Sliding into the Mobile Internet

Morgan Stanley have released "The Mobile Internet Report" (December 2009). They argue mobile Internet will happen more rapidly than previous Internet technology and have a more effect on companies involved. They argue this will be fuelled by 3G adoption, social networking, video, VoIP and improved mobile devices (they fail to mention mobile web standards as a leading part of the process). They argue the success of charging for mobile Internet in Japan provides a model for the rest of the world. This has proved not to work, with implementations of the Japanese mobile web (iMode) failing in markets outside Asia, as an example, Tesltra's abandoning their licencing of the service in Australia. MS also argue that developing nations will provide a model, which is hardly news for anyone who has visited India in the last few years.

Unfortunately MS undermine the credibility of their own report by providing it in a mobile unfriendly format: as hundreds of slides in very large PDF documents. In somewhat of an understatement they warn: "Note that the report is 50MB and may take some time to download.". The web page about the report scored only 3 out of 100 with the W3C mobileOK Checker.

Had MS taken the trouble to read their own report and perhaps done a few web searches, they might see that producing documents this way does not make them easily accessible on mobile devices. There are well developed mobile web technologies now available which they could have used (I have been teaching Mobile Web Design to ANU undergraduates for years).

Our global technology and telecom analysts set out to do a deep dive into the rapidly changing mobile Internet market. We wanted to create a data-rich, theme-based framework for thinking about how the market may develop. We intend to expand and edit the framework as the market evolves. A lot has changed since we published “The Internet Report” in 1995 on the web.

We decided to create The Mobile Internet Report largely in PowerPoint and publish it on the web, expecting that bits and pieces of it will be cut / pasted / redistributed and debated / dismissed / lauded. Our goal is to get our thoughts and data into the conversation about what may be the biggest technology trend ever, one that may help make us all more informed in ways that are unique to the web circa 2009, and beyond. ...

Our key takeaways are:

Material wealth creation / destruction should surpass earlier computing cycles. The mobile Internet cycle, the 5th cycle in 50 years, is just starting. Winners in each cycle often create more market capitalization than in the last. New winners emerge, some incumbents survive – or thrive – while many past winners falter.

The mobile Internet is ramping faster than desktop Internet did, and we believe more users may connect to the Internet via mobile devices than desktop PCs within 5 years.

Five IP-based products / services are growing / converging and providing the underpinnings for dramatic growth in mobile Internet usage – 3G adoption + social networking + video + VoIP + impressive mobile devices.

Apple + Facebook platforms serving to raise the bar for how users connect / communicate – their respective ramps in user and developer engagement may be unprecedented.

Decade-plus Internet usage / monetization ramps for mobile Internet in Japan plus desktop Internet in developed markets provide roadmaps for global ramp and monetization.

Massive mobile data growth is driving transitions for carriers and equipment providers.

Emerging markets have material potential for mobile Internet user growth. Low penetration of fixed-line telephone and already vibrant mobile value-added services mean that for many EM users and SMEs, the Internet will be mobile. ...

From: "The Mobile Internet Report", Morgan Stanley, December 2009 (also available in Simplified Chinese - 移动互联网研究报告摘要).

Google Chrome Operating System

Google announced the Google Chrome OS on 7 July 2009. Few details have been released, but this appears to be a version of Linux using a web interface conceptually similar to webOS used on the Palm Pre. It is not clear if Chrome OS will be a scaled up version of the Android operating system
used on the HTC Dream and HTC Magic mobile phones.

The idea seems to create a minimal operating system to support a web browser and then have most of the application provided remotely by web servers. This will not be the first such minimalist version of Linux. The Lenovo IdeaPad S10e netbooks selected for NSW secondary school students come with "Splashtop", an instant on version of Linux with simplified applications, in addition to the Windows operating system.

Google may help popularise such Linux based operating systems. This could make low cost, low power computers more feasible. It should cost less than $50 to add such a computer to an LCD video monitor or TV for example.

Hugh Fisher will give a free talk about Programming the iPhone, 4pm, 25 June 2009 at the ANU in Canberra:

DCS SEMINAR SERIES

Programming the iPhone

Hugh Fisher (Computer Science, ANU)

DATE: 2009-06-25
TIME: 16:00:00 - 17:00:00
LOCATION: CSIT Seminar Room, N101

ABSTRACT:
This seminar will give an introduction to the Apple iPhone (and iPod Touch) as a programming platform. Topics covered will include the hardware and OS capabilities, programming environment, the new styles of user interaction, and the restrictions on development and distribution. Don't expect any impressive demos or in-depth benchmarking: the target audience is people who've never programmed an iPhone but are curious about what is involved.

BIO:
Hugh Fisher is a long time owner and programmer of Apple computers, and also interested in human-computer interaction. (He is also a long time member of the Compter Science Dept.)

Twitter-like applications for humanitarian relief operations

In "Humanitarian Twittermifan" (May 26th 2009) the Sahana Disaster Management System blog discusses the use of Twitter-like microblogs for humanitarian relief. The particular application mentioned is for the World Food Program internal communication. This is implemented using open source microblogging package Laconica for the server and Twhirl for the client.

ps: In "National Bushfire Warning System: Micro-blogging for emergencies" I suggested the use of Twitter-like systems for use in bushfires.

Masterclass on National Bushfire Warning System

Apart from "Green ITSM" I have offered draft material on how to build a National Bushfire Warning System for Australia as part of a masterclass. on technical writing at ANU. So I need to write something on the topic. As a starting point, I have the abstract in the seminar announcement , a conference talk on "Community Warning Systems" and an "Australian Community Warning System Proposal" submitted to the Council of Australian Governments. Since then I have looked at "Obligation for Australian broadcasters to provide emergency warnings ", "Fault in Pacific Tsunami Warning System", "Mobile beep for emergency Cell Broadcast" and "Australian Emergency Alert System". My conclusion from this is:

1. SMS is not suitable for large scale use in a community warning system. SMS is too slow (taking around an hour to send 1M messages) and addressing information is lacking.
   
2. Cell Broadcast (SMS-CB) is technically suitable (able to send millions of messages in a few seconds to all mobile phones in a specific location), but not currently sufficiently supported by the mobile phone industry to be usable for warnings to the general public.
3. SMS and Cell Broadcast would be suitable as part of a system for relaying emergency messages from emergency services to the public via the broadcast media. This would improve on current methods using faxes and phone calls.
   
4. Short text messages could reference detailed web based information.
5. Standardised, efficient formats are required for web based emergency information. Accessibility and mobile guidelines can be used for designing an efficient readable format, as well as specialised guidelines for emergency information.
   

Given I now have my conclusion, I need to find some references to support it. ;-)

More seriously, the problem is to define the topic sufficiently to be able to find relevant work. Emergency management is a very wide topic, and even communication for emergencies has a large literature. One good place to start is the recent work on XML based formats for emergency messages, being pioneered by the OASIS Emergency Management Adoption Technical Committee and members such as Renato Iannella at at NICTA.

CSS 3 on the way for mobile and gaming web

John Allsopp gave an impassioned talk about CSS 3 at the March Canberra WSG meeting at the National Library of Australia. This talk, the approach John suggests and the technology it was about , could change the way the Australian Government does web pages and save lives.

This was impromptu talk due to the unavailability of the planned speaker. It had never occurred to me that anyone could be that passionate about Cascading Style Sheets (CSS), a highly complex and frustrating piece of web technology (the CSS standards are much bigger than the HTML one). CSS 3 is the newest and least supported of the CSS standards. Until this talk I had dismissed CSS 3 as something which would be good to have, but so little supported by web browsers that it was not worth looking at, and may never be supported.

John argued that new versions of browsers, particularly Firefox, Safari and Opera, had some useful CSS 3 features. The not so subtle subtext of his presentation was that web designers should give up trying to make web pages look identical on every web browser. Even if this were possible on desktop versions of browsers, it will make less and less sense as the web is accessed on hand held devices (such as the iPhone and Google Andriod smart phones) and games machines (such as the Wii). New devices will have very different size screens and user interfaces, so the desktop metaphor will not suit these.

John argued for designing web sites which will use advanced features on advanced browsers but still be compatible with old browsers. He argued that with new CSS features this can be done without lots of Javascript and CSS hacks, as used in the past.

All of that would have made an interesting academic argument, with a little more passion than usual, but hardly significant for day to day web design. However, John then went on to demonstrate some CSS features available now on newer browsers which can be used to create very elegant Apple Mac and iPhone type effects. The point he emphasised was these techniques were available now, did not require a large amount of code and were backward compatible.

What particularly struck me was how these features could be used for efficient hand held appliciaitons. CSS can be used to create elegantly shaded and curved buttons, without the use of any images. Such buttons will suit smart phones, without the need for large image downloads. Also animation techniques where the item selected changes size and colour can make maximum use of a small screen. At the same time the application will still be compatible with a desktop computer.

A few days ago I criticised the ABC for launching their "ABC Mobile" service with an interface which did not meet accessibility standards. Chris Winter, from ABC Innovation, responded to say they had fixed that problem. The ABC might like to direct their innovation at using some of these new CSS features. It should be kept in mind that the ABC provides information, not just entertainment and in particular the ABC provides emergency information during natural disasters. The techniques John was showing with CSS3 promise a way to provide visually appealing interfaces which are efficient and so would suit web pages designed for emergencies. I will be talking about this at a seminar on a "National Bushfire Warning System" in Canberra, 16 April 2009.

National Bushfire Warning System, Seminar, Canberra, 16 April 2009

I will be presenting a seminar on a "National Bushfire Warning System" at The Australian National University in Canberra on 16 April 2009. In this I will expand on my talk at the APCO 2009 Conference on "Community Warning Systems" and an "Australian Community Warning System Proposal" submitted to the Council of Australian Governments. All welcome, no need to book, just turn up.

Seminar Announcement
School of Computer Science, CECS
The Australian National University

Date: Thursday, 16 April 2009
Time: 4:00 pm to 5:00 pm
Venue: Room N101, CSIT Building [108], North Road, Canberra

Speaker: Tom Worthington
Title: National Bushfire Warning System

Abstract:

Recent bushfires in Victoria and floods in Queensland have brought the issue of warning systems for the public to prominence. Modern digital communications, the Internet and web, have a useful role in emergency communications. However,some technologies such as VoIP may make Australia more vulnerable. An alternative national system using Cell Broadcast technology via mobile phones is proposed. The potential for Social Networking to be used for emergencies will also be discussed.

Biography:

Tom Worthington is an IT consultant and Adjunct Senior Lecturer in Computer Science at the Australian National University, where he teaches the design on Internet, web and mobile phone systems, including for emergency management.

He is a former IT adviser at Headquarters Australian Defence Force. Tom is a member of the Project Management Committee of Sahana open source disaster management system, used for the 2004 Indian Ocean tsunami. He was elected a Fellow of the ACS for his work on Internet policy for Australia.

URL: http://cs.anu.edu.au/lib/seminars/seminars09/dept20090416
Seminars homepage: http://cs.anu.edu.au/seminars/

Australian Emergency Alert System

The proposal that Australian governments build a national community warning system to be ready for the 2009/2010 bush-fire season appears to have been reasonably well received. Several people pointed out similar proposals and the difficulties they faced with inter-state rivalry. These problems can be addressed by quickly building a small simple and cheap system, which states can interface to. Those states with existing sophisticated systems can interface to the national system simply. Those states which lack an effective system of their own can use the national system directly, with no need to buy any additional equipment.

The system could be named the Australian Emergency Alert System (AEAS) and modelled on the USA's Emergency Alert System (EAS). The AEAS would be conceptually similar to the EAS but take advantage of advances in digital technology to deliver a more functional system at a lower cost.

US Emergency Alert System

The US Emergency Alert System (EAS) is a national warning system introduced in 1994, as an upgrade of a cold war era emergency broadcast system. The system was designed to enable the US President to address the nation with 10 minutes warning, but that function has never been used. Commercial broadcast radio and TV services, as well pay cable and satellite services are required to interface to the EAS. While interfacing to TV, the EAS is designed to only deliver a spoken message, with no graphics. Newer devices, such as weather radios can detect a more advanced digital signal using Specific Area Message Encoding (SAME), which allows for filtering messages for specific regions and types. The system is used for tsunami and warnings.

Common Alerting Protocol

The EAS is essentially analog technology to which some digital enhancements have been added (such as SAME). The USA is considering upgrading parts of the system with the Common Alerting Protocol (CAP), an XML-based data format for exchanging public warnings and emergency information. CAP is a web based technology which is therefore much simpler to implement on digital devices.

Australian Implementation of CAP for a National Warning System

Australia has no EAS infrastructure to upgrade and no stock of weather radios to maintain compatibility with, so it can implement CAP more simply and directly than the using a network of purely digital devices. In addition, as Australia has a national, government owned broadcaster, with a national digital infrastructure, CAP can be interfaced to national radio and TV networks simply and at low cost. In addition the availability of the Internet and the web allows for much more detailed information to be provided to supplement the short emergency messages.

Levels of Detail of Information

One flaw in the design of proposed Australian emergency warning systems is a lack of depth in the information provided. Emergency messages for the community are brief. However, there needs to be more detailed information made available through other channels, typically standard information prepared in advance, not specific to the current emergency. It can be difficult for the public to obtain this information and also for the media to find it to provide to the public. The result can be that while the public may get the brief emergency message they have little idea what to do in response.

The current practice in Australia is to ask the public to tune to local ABC Radio. This can now be supplemented by asking them, where Internet access is available, to read a specific web page. Templates of information can be prepared in advance, so only the details of the current event needed be added. The web pages can be carefully designed using well established principles for efficient and rapid communication of information. The web pages can be made compatible with assistive technology for the disabled, with mobile phones and with TV screens, so the media can relay the information.

The AEAS can provide brief messages, suitable for transmission to mobile phones as a text message, displayed on a TV screen and read out on radio. Those messages can then refer to more detailed information available via the web. Broadcasters can read the web pages and convey the information to the public via audio and visual means.

Web Interface for the AEAS

The assumption in the US and current Australian warning systems is that after the brief emergency warning is issued, some other system will be used to send detailed information. No coordinated system exists in Australia to relay such information to the public. During the Canberra 2003 and Victorian 2008 brush fires, web based systems of the state governments were unable to provide effective information due to poor design of the web sites and a lack of planning.

It is proposed that the AEAS include a national emergency web site, to provide coordinated information to the public. Links to specific national and state systems can be made from here. The web site will also provide an interface for officials at the national, state and local levels to issue AEAS information, where those officials do not have access to specialised CAP interfaces. The officials will be able to upload a bulletin about the emergency as well as the brief text message.

Emergency agencies which have preformed templates for emergency bulletins will be able to use these with the AEAS. Agencies which have automated systems to fill in the templates will be able to use these to speed the process. While messages would normally be originated from a national and state emergency headquaters, the system would allow an authorised fire-fighter, using a smart phone, to issue a warning for their local area from the field.

Australian Community Warning System Proposal

My talk on Community Warning Systems at the APCO Communications Conference was reasonably well received. So I have proposed Australian Governments build such a national CSW system before the next fire season:

To: Council of Australian Governments (c/o COAG Secretariat, COAG Unit, Department of the Prime Minister and Cabinet)

This is to propose that Australian governments build a national community warning system. Such a system can be in place for the 2009/2010 bush-fire season at a cost of less than $10M. The system would be used for fires, tsunami, terrorist incidents and other emergencies. It would be capable of delivering a warning message in less than a minute to most of the Australian population, or a specific geographic area.

At the APCOA emergency communications conference today, I gave a brief outline of problems with the current proposals for the use of telephone based warnings and SMS messages. The alternative proposed is the use of "Cell Broadcast" text messages for alerts. The Internet, with specially designed web pages, can be used to provide preparatory information to the community before an emergency and information to help with recovery afterwards.

The technology for issuing text warnings is built into the mobile telephone network in Australia. A small amount of work will be required to make this system available for emergency personnel to issue warnings. The more difficult task is to educate the emergency personnel and the public as to what the system is for and how to use it.

See: Community Warning Systems - Balancing Technology and Reliability, Tom Worthington FACS HLM, for the APCO Australasia Annual Conference, Australian Technology Park, Sydney, 10am, 3 March 2009.

La Fonera 2.0 Linux Wireless Broadband in Australia?

Fon offer a WiFi router which includes a community service in some countries. The idea is that you allow others to share some of your broadband connect via WiFi and in return you can share some of theirs when travelling. The latest version of the router has a USB port and there is an Wiki where applications for the Linux software are discussed. The idea is you can plug in Scanners, Webcams and Printers. But the applicaiton which interests me is to plug in my 3G USB modem, to make a small wireless router.

The Fon hardware is $US49.95, which is less than most 3G routers. 3G modems get a brief mention in the Wiki and there is mention elsewhere of some Huawei H200 and some Sierra modems working.However, Fon do not appear to be shipping to Australia. Also the software is current beta.

ABC Mobile Web design

ABC's Media Watch program ("ABC Designs For The Future",15 September 2008), had a detailed discussion of the possibility of advertising on new ABC mobile web sites. Contrary to what the program said, the ABC could simply and cheaply provide a mobile option for its existing site. Examples were show of prototype ABC Weather, ABC Grandstand Sport and news pages. Most of the analysis was about the appropriateness of having advertisements on an ABC web site. What didn't seem to get discussed was why have separate mobile sites at all. The ABC's web site, including media watch already work on mobile devices and just need some minor improvements to make all the ABC web pages available on mobiles.

Presenter Jonathan Holmes said: "Try to view ordinary websites on one of the flash new 3G mobile devices like Blackberry and iPhone, and they're very clunky. And they'll also quickly bust your download cap. So everyone's racing to develop new, mobile friendly, websites - the ABC included...". That is certainly the case for Media Watch's own web site, which when viewed on a mobile device (I used Opera's "Small Screen" open to emulate a mobile device) has too much content and too much animation. However, this can be fixed by simplifying the standard web site and then using the options built into web standards to have it adapt when viewed on Mobile devices.

By making the main web site mobile compatible, this will also improve the quality of the web site for all users. Web designers have a tendency to clutter web pages with too much content. Mobile devices have small screens and limited bandwidth, providing a useful curb on the designer's enthusiasm. Where high bandwidth is required, the web server can automatically adjust the content to suit the device.

There have been several past attempts by the mobile phone and web industries to make mobile web sites a viable business. These attempts have mostly failed. The ABC should make a small investment in making all its web content mobile compatible. This can be done with a small project and mostly using automated tools. It would be a minimal expense and effort compared to creating and maintaining multiple new web sites. After it has done that the ABC can see who actually wants to view what content on which mobile devices. The popular services can then be improved.

Billions of dollars have been lost over the last ten years by companies attempting to create a profitable business from mobile web sites. If more companies wish to risk more money in the attempt, they are free to do so, with the consent of their investors. However, the ABC is a government owned enterprise funded by the public and so not be investing heavily in speculative ventures with minimal chance of success. The ABC is required by law to make its services widely available. There are ways for the ABC to provide mobile web access at minimal expense with little risk using available technology. I suggest the ABC take that approach and leave risky DOT.COM style web speculation to others.

Internet Fridge 2.0

It appears that several makers have announced new Internet fridges, such as Whirlpool. These have a flat panel touch screen tablet computer in the fridge door. The same claims are made as for failed models five years ago: being able to check e-mail and surf the Internet in the kitchen and a fridge that will reorder what was needed online in the future. Unfortunately the idea of using a computer while standing up in the kitchen makes no more sense than it did last time. None of the fridges can actually automatically scan their contents.

A fridge door is not a good place to mount a computer screen. The fridge is in a high traffic area of the household and the user would be constantly having to get out of people's way as they went past, or wanted to get items out of the fridge. Operating a computer with a vertical touch panel is difficult and uncomfortable at the best of times. Different height people would have to bend over or stretch up to be able to operate the unit.

LG's Internet fridge was modified to be the LG Refrigerator TV: the LCD screen was retained but the controls were simplified and the computer replaced with a TV tuner. This has been more successful than their Internet fridge. There is less operator input needed for a TV and much of the time at breakfast, the TV is not being watched, just listened to. Electrolux have a similar "Screenfridge".

What I suggested last time, and some makers seem to have adopted, is to clip a wireless tablet computer to the door. Samsung attemtoped this with their RH269LBSH which had a 10.4 inch LCD touch pad computer on the front. This could be detached and operated using wireless networking. This makes a little more sense than a computer on a fridge, but not much. It would be much cheaper and more convenient to simply have a small computer, such as one of the new low cost sub-notebooks, on the bench-top, or even attached to the wall, elsewhere in the kitchen.

You could make your own Internet fridge by covering the back of a PDA, or an iPhone with magnetic plastic. This turns the device into a fridge magnet. You can operate it attached to the door, or pull it off and use it. Alternatively a charging cradle for an iPhone, could be attached to the fridge door. This could be similar to units for car dashboards. But keep in mind that an electronic device attached to a fridge door may be easily damaged.

Another problem with the Internet fridge is their high energy consumption. Makers only provide the Internet function on their largest, top of the range two door models. I spoofed this by proposing an Internet Bar Fridge.

Web Pages for Mobile Phones

The W3C is continuing to progress with how to get web pages into a suitable format for mobile phones and other handheld devices. They have released "Content Transformation Guidelines 1.0" (W3C Working Draft 14 April 2008). These were prepared by Jo Rabin, mTLD Top Level Domain (dotMobi) and describe how web pages can be automatically transformed to make them more suitable for mobile devices.This can be done by a proxy server in between the web server and the web browser, or it can be done by the web server itself or even by the hand held device.

Of course content which was prepared without any thought that it might be read on other than a desktop computer is not going to necessarily transofrm well. The Mobile Web Best Practices 1.0 - Basic Guidelines suggest what a mobile device will need and would help in doing the web pages well in the first place.

But a looming problem is that the mobile people are working on XHTML standards and meanwhile a different group is working on HTML 5. This problem becomes apparent in devices such as the Apple iPhone and Apple iPod Touch, which the designers believe can render desktop web pages themselves and so do not activate mobile CSS style sheets.

PocketSurfer2 Mobile Web Device

PocketSurfer2 is a hand held wireless web terminal, which is expected to be released for Christmas (Amazon.uk are taking orders). It uses the mobile phone network to provide access for web browsing and email access. The device only provides text and still image access, no phone capability and no video, resulting in a low cost device which does not require a 3G broadband phone network. This makes it less capable than devices such as the ASUS Eee PC, but cheaper and more self contained.

The device got a reasonable review in APC, but much will depend on how much the wireless service is going to cost.

Mobile web compatibility test suite released

The W3C have released their mobileOK Basic 1.0 Tests as a Candidate Recommendation (that is a draft standard). These are tests intended to see how compatible web pages are with mobile phones and other hand held gadgets. The aim is "Making Web access from a mobile device as simple, easy and convenient as Web access from a desktop device". As well as the test specification document, W3C have provided some very useful software implementing some of the tests. A press release has background on the project.

I have tried out the tests and found them very useful, even where the web pages were not specifically targeting mobile devices. My only quibble is that perhaps the W3C's Mobile Web Best Practices Working Group should run the tests on their own web pages. ;-)

RM Asus miniBook

The ASUS Eee PC sub-notebook diskless Linux computer will be sold in the UK as the "RM Asus Minibook", by Research Machines. There is a review of the miniBook from UK PcC Advisor Magazine.

RM is a UK company selling IT products to schools and universities and are positioning the miniBook as a student computer:

The RM Asus miniBook provides individual access to Learning Platforms for every learner at an affordable price.

It can be purchased outright or through flexible financing schemes, including warranty and insurance, spreading the cost to suit the available budget. ...

From: 'RM Asus miniBook -The genuine "anywhere, anytime access" pupil device', RM, 2007.

This approach should work well. The MiniBook is a bit quirky as a mainstream business product, but should work well with support from a school or university. The institution could provide wireless access and even recharging to suit the mini book and also web based education resources which could work within its limitations. The machine should work well with Moodle and similar web based education content management system.

DIY Educational Computer

An enterprising notebook seller could produce a similar product from a low cost notebook computer. That would cost more than the miniBook, but might produce a better computer. As an example start with the Twinhead Slimnote 12KF, which a 12 inch screen subnotebook computer with a AMD 1.8Ghz CPU for $AU1,549. I have the more expensive Twinhead 12D with an Intel processor (Twinheads are also sold as Avertech). Omit the internal DVD/RW Drive and replace the hard disk drive with a 8 GB Sd Flash Card, then install Linux in place of Microsoft Windows.

A cheaper computer, which might be even more desirable for students could be produced by starting with a lower cost computer, such as an Acer for under $AU1,000. The Dell Inspiron would be another option; some models are supported with Linux by Dell. and start at under $AU1,000.

W3C Mobile Web Best Practices checker

W3C's Mobile Web Best Practices Working Group have a set of tests for what makes a mobile web page: "mobileOK Basic". Some of this have been implemented in the W3C Mobile Web Best Practices checker. This goes a bit further than a syntax checker, and for example tells you if your web page is too big for the average phone, but not as far as accessibility tests.

Just for fun I ran the tests on the W3C Mobile Group's own web page:

Results of checking W3C Mobile Web Best Practices Working Group

This page failed on 5 tests

1. The page is not valid with regard to the XHTML Basic 1.1 DTD.
   This test is related to the following Best Practices: VALID_MARKUP (techniques)
2. The markup of the page is bigger (19138 bytes) than the upper limit of Default Delivery Context.
   This test is related to the following Best Practices: PAGE_SIZE_USABLE (techniques) PAGE_SIZE_LIMIT (techniques)
3. The added sizes of the markup and of the external resources embedded in the page are bigger (27608 bytes) than the upper limit of Default Delivery Context.
   This test is related to the following Best Practices: PAGE_SIZE_USABLE (techniques) PAGE_SIZE_LIMIT (techniques)
4. The image at line 23 column 12 has no width set in the markup.

   …

   This test is related to the following Best Practices: IMAGES_SPECIFY_SIZE (techniques) GRAPHICS_FOR_SPACING (techniques) IMAGES_RESIZING (techniques)
5. The image at line 23 column 12 has no height set in the markup.

   …

   This test is related to the following Best Practices: IMAGES_SPECIFY_SIZE (techniques) GRAPHICS_FOR_SPACING (techniques) IMAGES_RESIZING (techniques)

The page passed 11 tests

1. The page provides caching information and does cache validation.
2. No links are set to open in a separate window.
3. The page doesn't use objects nor scripts.
4. The page doesn't use tables.
5. The page doesn't use image maps.
6. The page doesn't use auto-refresh nor redirects.
7. None of the length attributes use absolute values.
8. The page has no item with a width set to more than the allowed size.
9. All the images have an alt attribute set; make sure it describes well the content of the image.
10. The page uses a logical order of headings.
11. The page has a title set to "W3C Mobile Web Best Practices Working Group"; check it is well descriptive of the content

This page raised 11 warnings

1. The page has text/html as a media type; that media type is not recommended for the Default Delivery Context.
   This test is related to the following Best Practices: CONTENT_FORMAT_SUPPORT (techniques) CONTENT_FORMAT_PREFERRED (techniques)
2. The link to http://ready.mobi at line 45 column 0 targets a resource whose media type None is not supported by Default Delivery Context; the text of the link should make that clear.

   …http://ready…

   This test is related to the following Best Practices: LINK_TARGET_FORMAT (techniques)
3. The link to http://dev.w3.org/cvsweb/2007/mobileok-ref/ at line 51 column 42 targets a resource whose media type None is not supported by Default Delivery Context; the text of the link should make that clear.

   …k to date in CVS at 

   This test is related to the following Best Practices: LINK_TARGET_FORMAT (techniques)
4. The link to mailto:public-bpwg-comments@w3.org at line 112 column 134 targets a resource with a scheme mailto not supported by Default Delivery Context; the text of the link should make that clear.

   …e send a message to 

   This test is related to the following Best Practices: LINK_TARGET_FORMAT (techniques)
5. The link to mailto:daniel.Appelquist@vodafone.com at line 143 column 32 targets a resource with a scheme mailto not supported by Default Delivery Context; the text of the link should make that clear.

   …ng Group Chairs are 

   This test is related to the following Best Practices: LINK_TARGET_FORMAT (techniques)
6. The link to mailto:jrabin@mtld.mobi at line 143 column 107 targets a resource with a scheme mailto not supported by Default Delivery Context; the text of the link should make that clear.

   …Appelquist and Jo Rab…

   This test is related to the following Best Practices: LINK_TARGET_FORMAT (techniques)
7. The link to mailto:mike@w3.org at line 144 column 17 targets a resource with a scheme mailto not supported by Default Delivery Context; the text of the link should make that clear.

   …Staff Contact is Michael(tm)…

   This test is related to the following Best Practices: LINK_TARGET_FORMAT (techniques)
8. The link to http://www.w3.org/blog/BPWG?tempskin=_rdf at line 190 column 5 targets a resource whose media type application/xml is not supported by Default Delivery Context; the text of the link should make that clear.

   …

   This test is related to the following Best Practices: LINK_TARGET_FORMAT (techniques)
9. The link to http://www.w3.org/blog/BPWG?tempskin=_rdf&disp=comments at line 191 column 5 targets a resource whose media type application/xml is not supported by Default Delivery Context; the text of the link should make that clear.

   …

   This test is related to the following Best Practices: LINK_TARGET_FORMAT (techniques)
10. The link to http://www.w3.org/blog/BPWG?tempskin=_atom at line 195 column 5 targets a resource whose media type application/atom+xml is not supported by Default Delivery Context; the text of the link should make that clear.

    …

    This test is related to the following Best Practices: LINK_TARGET_FORMAT (techniques)
11. The link to http://www.w3.org/blog/BPWG?tempskin=_atom&disp=comments at line 196 column 5 targets a resource whose media type application/atom+xml is not supported by Default Delivery Context; the text of the link should make that clear.

    …

    This test is related to the following Best Practices: LINK_TARGET_FORMAT (techniques)

The following information can help assess the conformance of the page to the Best Practices

1. The page uses style sheets; check that it can be used in a device where style sheets aren't available.
2. The markup contains 10 percent of unneeded whitespaces summing up to 918 extraneous characters.

Non-Applicable

The following information shows which Best Practices were not applicable to the said page. ...

Ultra Mobile PCs in Canberra

At the ACS Canberra Branch meeting Tuesday night I noticed David Macke from Synergistic Network Solutions, taking notes on a Samsung Ultra Mobile PC. It turns out his company sells this and similar portable gadgets. I have been skeptical of the UMPC idea, but it may have its uses. To me the subnotebook computer has wider appeal.

What to Do with an Open Source Phone

Shayne Flint gave a presentation about the OpenMoko "Open Source" mobile phone in Canberra on Wednesday. I can't give an unbiased review as I organized the talk, but I thought it gave a good overview of the issues.

The Neo 1973 phone demonstrated shows a lot of potential as a platform for open source applications, but the units shipping now a re only for developers. The NEO is still at least six months away from being a consumer product.Shayne showed how the phone was compatible with commonly used Linux applications. He also demonstrated it displaying web pages, including ANU's new "ANU Mobile" web service.

But the question now is what should the researchers try to do with the phone? Just getting the standard phone applications to work (such as dialing a call, sending an SMS and using the calendar) will not rate as research.

Probably the least useful aspect is to see the phone as a phone. Instead it can be thought as a Linux PDA, or a low power touch screen portable wireless computer. As Shane says the significant aspect of the device is that it can run Linux applications. So I see the interesting bit as being how to adapt applications to it, rather than building new "phone" applications.

Some areas for research:

1. Adapting web pages: For some years I have been teaching ANU students how to modify web sites so they will run on mobile phone screens as well as desktops, using the same technology as used to provide access for the disabled. This is much easier to do with smartphones, such as the Apple iPhone, due to their larger screens. Modifications to commonly used open source web content management systems. Ideally, the CMSs would generate the code for the smartphone by default and automatically, with no user intervention needed. This would require no additional software on the phones and ideally the same web pages would work on desktop computers, the OpenMoko devices and Apple iPhones. As well as CMSs, some specialist applications with web interfaces could be adapted, such as OJS for academic publishing and Moodle for education.
2. Mobile Offline Web Applications: Web based applications which can run offline may provide a useful phone option. Toolkits such as Google Gears could be adapted to work better with the phone.
3. Mobile Web 2.0: AJAX and similar interactive "Web 2.0" applications would be ideal for OpenMoko. Application toolkits could be modified to adapt to a mobile phone screen (as above). The server software could be modified to run on the phone (while maintaining compatibility with remote servers.
4. Mobile Corporate Social Networking: Social networking applications could be adapted to the phone. As well as using this for social purposes it could be applied to business.
5. Advanced vehicle systems for developing nations: One use for the phones may be in transport.
   

Open Source Alternative to the iPhone

Recommended:

Australian Computer Society

Green IT Special Interest Group

In Conjunction with the ACS SQA SIG

August Canberra Meeting

Topic: Open Source Alternative to the iPhone

Speaker: Dr. Shayne Flint, Department of Computer Science, The Australian National University

Venue: Australian National University, Room N101, Computer Science Building, North Road, Canberra
Date: Wednesday 15 August 2007
Time: 5:30pm drinks/nibbles for presentation 6pm-7pm
Event Prices: Free.
Registration: Not required
Announcement: http://education.acs.org.au/mod/resource/view.php?id=3636

Summary

Dr. Flint will demonstrate the software development kit for the OpenMoko "Open Source" mobile phone. While the Apple iPhone has been getting media attention, another touch phone has been quietly under development by the Linux community and will be first to market in Australia. OpenMoko is set to revolutionize mobile communications by providing the power of Linux in a hand held touch screen device. Shayne will discuss some of the software engineering projects being formulated for use with the phones location sensitive and wireless communications features.

Shayne will use the OpenMoko device to demonstrate the ANU's new "ANU Mobile" web service, which has just been released.

Tom Worthington, Chair of the ACS Green IT Group, backs the move to mobiles, arguing that low power handheld devices with open access applications can be used to make a positive contribution to environmental sustainability. Power hungry desktop computers can in many cases be replaced with mobile devices. Some trips can be replaced with ad-hoc wireless meetings, using Web 2.0 "social networking" making further fossil fuel savings.

About the Speaker

Dr Shayne Flint is a Senior Lecturerat the Department of Computer Science, Australian National University, where he teaches Software Engineering. Dr. Flint is the originator of Aspect-Oriented Thinking, an approach that systematically develops, manages and integrates the knowledge and expertise of many disciplines to develop complex software systems.

About ACS Green IT

The ICT Environmental Sustainability Group ("Green IT") brings together professionals interested in balancing economic and environmental aspects of information technology and telecommunications. It is a special interest group of the Australian Computer Society. The group aims to hold joint meetings with other professional bodies interested in technology, the environment and sustainability. Sign up now to get updates on ICT and the environment.

$199 Mobile Internet Device

Asus have announced the $US199 "Eee PC". This is a diskless B5 size subnotebook PC. It is similar to the Palm Foleo mobile companion announced a few days ago. The Eee PC looks a more viable product, but it is not clear there is a market for any of these Mobile Internet Device web terminals.

EeePC 701 Specification

Display:7"
CPU & Chipset: Intel mobile CPU & chipset
OS: Linux/ Microsoft Windows XP compatible
Communication: 10/100 Mbps Ethernet; 56K modem
WLAN: WiFi 802.11b/g
Graphic: Intel UMA
Memory: 512MB, DDR2-400
Storage: 4/ 8/ 16GB Flash
Webcam: 300K pixel video camera
Audio: Hi-Definition Audio CODEC; Built-in stereo speaker; Built-in microphone
Battery Life: 3hrs (4 cells: 5200mAh, 2S2P)
Dimension & Weight: 22.5 x 16.5 x 2.1~3.5cm, 0.89kg

From: ASUS Introduces All-New Eee PC for Complete Mobile Internet Enjoyment, Asus, 2007

The ASUS unit is in price, size and concept very similar to the Sphere/Zeos Palm Top PC I traveled around Europe with in 1994 and the OLPC computer. But they are twice as large as the educational computers currently in widespread use. If priced low they could make a useful replacement for a desktop PC, laptop or PDA. But they may be too small to replace a laptop and too big to replace a PDA.

For those needing occasional, trouble free web access at home, this might provide a useful alternative to a PC. This could also be used for business travelers who have wireless broadband access and an organization with a web based suite of office applications, such as Google Apps.
The ASUS unit appears to have a case designed for a 9 or 10" wide screen, but with a smaller 3:4 screen installed (presumably to lower the cost). The TWINHEAD 10D/Averatec AV1020-ED1 Notebook PC has a similar case, but a 10.6" wide screen. The Twinhead system costs about $1,700, or about ten time the projected price of the ASUS, but as well as a larger wide screen, it also has a hard disk drive, DVD drive and (presumably) a faster processor.

Web Terminals: will anyone buy them?

On 30 May 30 2007 Palm announced the Foleo mobile companion is to be available in late 2007. This is a diskless subnotebook with a 10" screen, designed to work with web based applications. This is similar to Intel's Mobile Internet Device (MID) , but with a bigger screen.

These units can be thought of something between an enlarged PDA and a shrunken laptop. The idea is that instead of carrying around a notebook computer with applications and data on a disk, you can have something smaller and lighter. The data and applications reside on the web and are accessed via an Internet connection, via WiFi to a hot-spot, or Bluetooth to your phone.

The media are skeptical of the Foleo and I share that skepticism. The unit depends for its success on being more portable than a subnotebook computer, enough web based applications and enough bandwidth to be useful. Palm's unit is not a lot smaller and lighter than current subnotebook computers.

Palm seem to have gone with the smallest size unit which can have full size keys on the keyboard. The 10 inch wide screen is about the same width as a QWERTY keyboard (I use a 12 inch 3:4 notebook, which has the same width screen).

Intel's Mobile Internet Device (MID) , is planned to have a 7 inch screen and will therefore have a much more cramped keyboard. I have owned and used such Pocket Computers in the past and have found them very useful. But they have not caught on as a mainstream product. Perhaps if customers seem them as large and smart PDAs, rather than dumbed down laptops, they will be successful.

The availability of web based office applications may make the difference. If you are using something like Google Apps and you normally use a web interface to access documents and email, then the web terminals will seem a natural way to work.

Previously using a non-Windows subnotebook computer has meant learning to use a different set of software and working out how to transfer documents from your desktop computer to the subnotebook and back. The subnotebook has a very limited memory and there is the need to juggle applications and data capacity.

With wireless Internet access and web based applications, the situation changes. The applications will look the same on the portable unit, as these are the same applications running on the same remote web server. There is no need to transfer documents as they can be worked on directly online. The document storage capacity is unlimited, as they are stored on the remote server. A broken or lost unit will not result in any loss of documents, as they are stored on the remote server.

However, this assumes you have cheap, reliable and sufficiently high speed Internet access to to use web applications. At least 64 kbps seems to be needed. A WiFi hot-spot would be sufficient, but mobile phone access may be problematic. iBurst at 256/64kbps works well (I use it in Canberra/Sydney/Melbourne and Brisbane). Ordinary GSM access at 9.6 kbps would not be sufficient. 2.5G would be adequate and 3G more than enough. But in Australia at least, this would be prohibitively expensive for most users, at the rates carriers currently charge.

One possibility will be the use of local versions of web based applications. The idea would be that the user would be able to download a document and its application, edit the document and then upload it later. Essentially this is the way AJAX applications already work, but they depend on Internet access being available every few seconds.

Mobile Internet Devices to Take Away the PC Market?

Intel is now promoting two categories of hand held computers: Mobile Internet Devices (MIDs) and Ultra Mobile PCs (UMPCs). The MID is a low cost Internet terminal, intended to work with web based applications, while the UMPC is a more powerful device which can run standalone applications. These differ from Microsoft's failed Ultra-Mobile PC, by having a keyboard, a slightly smaller screen and Linux as the default operating system:

Mobile Internet Devices (MIDs)
MIDs personalize a new category of small, truly mobile consumer devices enabling a PC-like Internet experience, coupled with the capability to communicate with others, enjoy entertainment, and access information on-the-go.

Ultra Mobile PCs (UMPCs)
UMPCs combine the benefits of a mobile Internet experience, with the practicality of security and manageability capabilities expected for business or education use. While these devices can be used for communication and entertainment, they also have access to a variety of PC software applications for productivity on-the-go.

From: Intel Ultra Mobile Platform, Intel 2007

These devices will look very similar to pocket computers from a decade ago, but with much better screens and more powerful processors. If the price is low enough, a Mobile Internet Device with WiFi could provide a useful replacement for a home computer and the UMP for a business computer.

Computer manufacturers will face the dilemma, that if they make small cheap useful computers, why would anyone buy their full size, more expensive laptops and desktops? A Mobile Internet Device with a full size screen and keyboard plugged in would make an adequate desktop web terminal for using with web based applications. But if traditional computer companies do not offer these products, then companies which make calculators and consumer electronics will step in and take most of the laptop and desktop computer market from them.