Software Programmers

Joel Spolsky writes:

The mediocre talent just never hits the high notes that the top talent hits all the time. The number of divas who can hit the f6 in Mozart’s Queen of the Night is vanishingly small, and you just can’t perform The Queen of the Night without that famous f6.

Is software really about artistic high notes? “Maybe some stuff is,” you say, “but I work on accounts receivable user interfaces for the medical waste industry.” Fair enough. This is a conversation about software companies, shrinkwrap software, where the company’s success or failure is directly a result of the quality of their code.

And we’ve seen plenty of examples of great software, the really high notes, in the past few years: stuff that mediocre software developers just could not have developed.

$100 PC

[via Niranjan] Jason Pontin writes:

Nicholas Negroponte, founder and chairman of MIT’s Media Lab, showed attendees the screen of the Hundred-Dollar Laptop, or HDL. Beginning in 2006, he said, he would build 100 million to 200 million HDLs every year–and distribute them to the children of the poor world. Many attendees had read about Negroponte’s idea and dismissed it as quixotic. Hearing how an HDL might be built, seeing a part of it, and realizing the scale of the project produced a rustle of delighted interest.

Negroponte recently wrote to me about what he hoped the HDL would do: “Education: one laptop per child. Whatever big problem you can imagine, from world peace to the environment to hunger to poverty, the solution always includes education. We need to depend more on peer-to-peer and self-driven learning. The laptop is one important means of doing that.”

Can a $100 computer be built? Maybe. Negroponte does not plan to use three expensive components of conventional laptops: Microsoft Windows, a traditional flat-panel screen, and a hard drive. Instead, the HDL will be loaded with Linux and other open-source software; its display will use either a rear-projection screen or a type of electronic ink invented at the MIT Media Lab; and it will store one gigabyte’s worth of files in flash memory.

Of Mobiles and PCs

Russell Buckley writes:

My vision (shared by many other mobile thinkers I respect, like Tim Oren and Rajesh Jain) is that the phone will take over from the PC, as storage on handsets get better and as more and more data is stored on the web anyway. This is already happening with Flickr for photos, gmail for email and Bloglines for RSS, so you don’t need huge amounts of storage on the device you carry around with you anyway.

When you need to do something that requires a larger screen and keyboard, you’ll just dock your mobile into one of the millions of stations dotted around the work, home and public environments and away you go. No logging in or re-configuring is necessary as your phone is both the key and holds the configurations that you’ve already decided on.

This will mean your phone’s operating system is the only relevant one – and that leaves Windows at an evolutionary dead-end.


The Economist writes about NTT Docomo’s plans to enable consumer payments via mobile phones:

Over 4m of its 51m subscribers now carry handsets equipped with FeliCa, a technology developed by Sony that DoCoMo incorporated into some of its handsets last year. FeliCa allows a small chip to be read wirelessly by a nearby scanner. FeliCa-based cards are already used as wireless train tickets by millions of commuters, and from next year it will be possible to use a FeliCa-enabled handset instead.

Why the sideways move into consumer finance? Having withdrawn from unprofitable forays overseas, DoCoMo has plenty of capital to invest and is looking for new sources of growth at home. It also wants to differentiate itself from KDDI, the increasingly competitive number-two operator. (Vodafone is a distant third.) Credit cards are far less popular in Japan than in other rich countries: the average card is used for less than $1,000 of purchases a year, compared with more than twice that in Britain, France, Germany and America (see chart). Building cards into the handsets of Japan’s mobile-phone users, who are ever eager to experiment with new tricks, seems a good way to get them to buy more on credit. DoCoMo would then benefit financially through its stake in SMCC.

What’s Next for Enterprise Software

Dan Farber writes about the Enterprise Ventures 2005 conference:

Software as a service is on the rise, but it tends to be applicable in areas of non-strategic core functionality, said Ted Schlein, a managing partner at Kleiner Perkins Caufield & Byers. I guess that means that CRM is core but not as strategic as financials. That may be true today, but over time more strategic applications will move to the Net, just as most of us don’t keep money under mattresses anymore and rely on banks. As the technology matures, and as software services can be evaluated and rated on the basis of factors such as reliability and trustworthiness, the model of outsourcing functions in a multi-tenant or single-tenant model will be prevalent.

Schlein also talked about Microsoft as being less relevant in the enterprise. Given his venture firms backing of Google that’s not surprising. The ‘Web operating system’ implies that a Windows operating systemor any other lower level core software is less important as users spend their time using Web-based communications and other kinds of applications. It’s not an either or situation. Over time, Microsoft, Apple, Novell and any other company with a client (desktop) interface will offer more rich media, Web-based versions of their software, beyond browsers. It’s not a religious issuethin or thick clientit’s more about getting serious bandwidth and quality of service in the network, and having hybrids that allow users to work both online and offline with the same interfaces and tools.

TECH TALK: Next-Generation Networks: Carriers and Vendors

by Ninad Mehta

IMS delivers enough benefits and value that most wireline and wireless service providers will adopt it, according to a recent report by Yankee Group. Going back to our discussion 2 days ago, SPs in developed countries will start deploying IMS architectures in their networks within the next 1-2 years. The wireline carriers in the developed countries will see a strong need to deploy IMS as they plan to introduce fixed/mobile convergence services using partner SPs. Most Tier 1 NAR SPs (Verizon, SBC, BS, AT&T, Sprint) are either doing trials using IMS or deploying IMS in their networks as this is written. Tier 1 SPs in Europe (BT, DT, FT, Telefonica-x) are all also busy in doing market trials or deployments. Wireless carriers in developed countries are taking IMS very seriously as well most of them have a lab or field trial going on and some have started deploying IMS within their networks.

Carriers in developing countries are still focusing on increasing teledensity and so they havent spent as much time and resources on IMS. This is changing as the government policy in several countries dictates that any infrastructure spending should be restricted to next generation infrastructure technologies.

On the supply side of IMS, the Telecom Practice arm of Venture Development Corporation classifies next generation infrastructure vendors in the following categories: (1) Enabling technology and components providers, (2) Network component providers, (3) System vendors, (4) 3rd Party application service providers and (5) System integrators.

This classification builds from bottom up and includes established as well some newer niche players.

According to a report by the Yankee Group, with respect to IMS components, most telecom equipment vendors have chosen to develop IMS core components (within the control layer) themselves and partner with niche providers for IMS components that belong to various gateways and application servers. Since IMS architecture is built upon well defined, standards based interfaces; theoretically, any vendors IMS component should work other vendors components. In reality, things would be quite different! I am sure system integrators and services providers are salivating at the revenue opportunity in putting things together.

Another important aspect of IMS supply side vendors is that non-traditional telecom vendors show up on the list of vendors providing various IMS components. These vendors have been historically associated as computing infrastructure providers. Examples include HP, IBM, Intel and others. There is a natural fit between blades based computing using ATCA (Advanced Telecom Computing Architecture) and distributed IMS components.

Over time, we will see IMS vendor consolidation since the market is already crowding up with numerous suppliers for each of the IMS components. We will also see consolidation within and between wireline and wireless service providers.

[Ninad Mehta works at Lucent in New Jersey. The views expressed in this column are his own.]

Continue reading