dcsimg
Linux Today: Linux News On Internet Time.




More on LinuxToday


Nathan Cochrane -- Applications for the masses

May 07, 1999, 09:39 (66 Talkback[s])
(Other stories by Nathan Cochrane)

WEBINAR:
On-Demand

Desktop-as-a-Service Designed for Any Cloud ? Nutanix Frame


By NATHAN COCHRANE

``It's not the box, it's what you do with it, stupid.''

These words have resided with me ever since I first heard them as a sales assistant at 15 while working part-time at a local computer store. The customer was annoyed at my very naive notion that the computer under discussion, an early Commodore Amiga, should be his preferred choice because of its technical superiority -- 4096 colors, stereo sound, and swift 7.14MHz 68000 processor.

He eventually bought a boring, early model IBM compatible with a green screen that cost twice as much because it had Lotus 1-2-3 and Wordstar, the dominant spreadsheet and word processor of the day. My error was I had fundamentally misunderstood his purpose -- I thought he wanted a computer. Instead, he wanted to solve a problem. He wanted to run his financial accounting business and correspond with associates mostly. Compatibility with his office systems was also a consideration. The computer was the means, not the end.

CHANGING APPLICATIONS

Today, applications can be enabled either through traditional shrinkwrap software, downloaded from the Net and then run, executed remotely over the Web or using a combination of these methods. Tax returns, for instance, can be provided through a Web page instead of a CD-ROM. Most literally, an application is the act of applying these methods, it is not the method itself.

With this change in focus on delivery, new business opportunities and threats emerge. You can grant free access to a Web-based tax form complete with built-in calculators, but to submit that form to the Government you pay $29.95. Most users won't care that the service runs on a technically superior server or that their PC is the latest Intel Pentium III or PowerPC G4 -- they just don't want to go to jail.

The biggest criticism of the Linux development community is it doesn't care enough for the end-user -- all the focus is on tweaking the Kernel or server uses. Although this criticism isn't entirely fair, this widely held belief contains a lesson.

Usability is under the microscope and is making significant headway thanks to the efforts of GNOME and KDE developers. The recently announced tweaking database in reponse to the controversial Mindcraft benchmarking is another example, and perhaps a pointer for other initiatives.

But more needs to be done to address other user issues, and leverage Linux's specific strengths by providing software and services.

DESIGNING FOR NEW APPLICATIONS

As technologies converge and devices diverge into new form factors, arguably the only ``killer app'' is a universal way to access information, currently usually a Web browser. Everyone's needs are different, so tools will be applied differently. I may want to access a tax return form, another may want to play a game, someone else may want to check a library catalog, send e-mail or listen to music. One user's killer app may be a mystery to another, and even a user's applications will vary over time as their needs alter. It may be essential for me to file this report, but after that I may want to relax and enjoy some entertainment.

Linux's most obvious strength is that it isn't owned by anyone. It doesn't cost anything to get the Kernel and GNU support software, it poses no threat to anyone because there are no business alliance concerns, code is svelte, is infinitely scalable and customisable. It fits many form factors.

It belongs on top of a television.

In the last three years, Microsoft has pushed with a vengeance into the worldwide Pay-TV industry. It wants every set-top box to run a version of WindowsCE, and the back-ends to all be Microsoft server architecture. Last week it was reported in the Wall Street Journal that the Seattle company is willing to dip into its balance sheet for a war chest of $US5 billion to encourage AT&T to use its systems.

AT&T and other cable operators have been understandably skittish about these sorts of approaches. The Pay-TV industry does not want to be in the same position of serfdom that it has seen in the IT industry.

Could Linux be the ideal solution?

It provides a mature, robust, scalable and licence-free architecture that can not only run a client, but the server as well. There are plenty of skilled and talented developers who are good-to-go with their expertise. And any plan to position Linux as an end-to-end proposition could be fulfilled in record time.

But best of all, because no-one owns Linux, it is no threat to the media titans. And it is how they apply the same base technology, not the technology itself, that will dictate success and profits.

Maybe this is the reason for Sony's rumored embrace of the technology for its next generation PlayStation? I know if I was heading a digital television convergence development effort in a big media company, Linux is a platform technology I would investigate closely.

More immediately, perhaps, there are other specific applications that are required. One is online gaming.

Interactive entertainment developer, Sierra Studios, plans to release an online collaborative environment based on JRR Tolkien's Lord of the Rings by March next year. Middle Earth will enable as many as 10,000 players at a time to take up residence in the mythical land. Users will be able to buy and sell property, own pets, join families, chat to each other via an IRC-like system and engage in the minutiae of everyday living.

But the scope of the project means the developers operating under a traditional closed proprietary model will not deploy for any client system other than Microsoft Windows.

``This decision is not based on any partisan opinion of the merits of a particular operating system, platform or company. It is simply a practical necessity; creating Middle-Earth for one platform is a Herculean task in itself,'' the Middle-Earth FAQ says.

``Cross-platform development of the constantly evolving client software would complicate this task to the extent of jeopardizing our efforts.''

As I read the company's business plan from what little information is available, Middle Earth is better suited to a service than a traditional shrink-wrap software model. It allows for geographic-specific hosting and franchising opportunities -- servers should be as close to users as the network allows.

The value of a network, in this case an online gaming community, increases with each node or person. Consider the growth of the telephone system, transportation networks, and Internet; they took a long time to develop leading to high tariffs, which discouraged users and sapped their innate value. As costs fell, more people were sucked into the mix until critical mass was reached and the industry exploded. A telephone isn't much good if you have no-one else to talk to; an e-mail bounces if there is no recipient and a road is no good if it doesn't go anywhere.

Selling the software is a barrier to adoption, so not allowing for profit maximisation from service opportunities. Instead, a better approach is to give away the binary executable AOL-style, bundle it on magazine covers, mass-mail it to anyone who wants it, and offer it for download. The priority should be to get as many people online as quick as possible with as few tethers as possible.

Additionally, giving away the source code under an appropriate licence reduces development costs, allows for new ideas to be rapidly implemented, and most importantly gives it the reach of a growing number of connected individuals running non-Microsoft systems. Combine it with a set-top box like the one mentioned earlier and the only question is how do you spend the millions you make from an initial public offering?

Sierra says: ``Once Middle-earth has launched and the client is stable a port becomes more practical, particularly if a credible developer were to make an approach.''

What more credible group than that which demands open peer review of the code? Here is an opportunity for the community to demonstrate its skills, ensure it has access to what looks like a great game, and for some people to make potentially a lot of money.

Better yet, why not borrow the concept and develop similar environments for other compelling mythical universes under licence? Fancy becoming a stormtrooper patrolling the arid wastes of Tatooine? Maybe you would prefer to be a Drazi trader on the Narn Homeworld in the Babylon-5 universe? Or would you like to drop by Anne McCaffrey's Pern, or even live through the eons in Asimov's Foundation series?

Another application I would like to see are those annoying Wizards. It's not that I want to use them -- I find Wizards and assistants get in my way most times -- but many users find them reassuring. An attempt to knock the HOW-TOs into assistant form would counter a lot of concerns about Linux's end-user appeal. These could even be deployed entirely remotely, leveraging ISPs use of Linux or a BSD onto the user's client.

Finally, Linux needs Microsoft Office.

Regardless of how many people feel about Microsoft products, the pragmatist understands the overwhelming majority of enterprises and users have standardised on Microsoft software for their word processing and other general office needs. This is a situation, borne of irrational religion and ideology in many cases, that is not likely to reverse any time soon. To make Linux a more attractive proposition for the mainstream, it needs this support.

Bill Gates has shown he is not inept when it comes to identifying new opportunities. That he has chosen to completely ignore the fastest growing niche market, Linux, with more than 20 million users speaks volumes.

Linux is a great innovation platform and allows natural synergies between nodes across a network. It should be used as a tool to reach into the lives of people who otherwise don't, and shouldn't have to care.


Nathan Cochrane Nathan Cochrane is a regular contributor to Linux Today and journalist for FairfaxIT's online and print publications. His weekly column, OpenLine (http://www.it.fairfax.com.au/columns/openline/index.html), looks at the open source and Free Unix communities, including Linux. He can be reached at ncochrane@theage.fairfax.com.au

More articles by Nathan Cochrane can be found at here.