Bookwatch: Loosely Coupled--The Missing Pieces of Web Services is still new (the official publication date is August 2003) and is therefore somewhat hard to find. Here are some tips as of today:

• NerdBooks: US$27.59 and in stock. Best price for U.S. orders.
• Amazon.com: US$27.99, 2-3 days.
• BarnesAndNoble.com: $31.99, 2-3 days (some days "unavailable").
• Amazon.co.uk: Not in stock yet. :-(
• Amazon.ca: CDN$38.69 (30% discount), 1-3 months! (temporarily)
• RDS Press: US$39.99, direct from us, ships within 24 hours.
  (Not the cheapest, but the best source for fast or non-US delivery.)

Amazon.com Review of the Week:

Does IT Matter? Is information technology about to achieve commodity status? The discussion began with an article by editor-at-large Nicholas Carr in the May 2003 issue of Harvard Business Review in which he claimed "IT no longer offers strategic value to anyone." Carr has stirred up a real hornets' nest in the press and elsewhere.

Not one to let a good controversy go by without expressing an opinion, here's my take on the matter. Although I disagree with Carr's conclusions, I think many of his points are valid and important. Carr's argument is that like railways and electric power, IT is about to become a commodity and therefore no longer the source of sustainable competitive advantage. I addressed this issue relative to web-services technology in my new book. The following figure taken from Chapter 18, The Timing of Complex Projects, illustrates the point. (The dates are for example only.)

In the general case, Carr is right. All technologies that at one time offer competitive advantage eventually turn to commodities or disappear entirely. His mistake, IMHO, is that he lumps all of IT into a single category that will become a commodity all at once. Certainly, there's much more innovation yet to come from IT, and well-run organizations will continue to use technology to create competitive advantages, albeit temporarily.

Carr writes, "The trap that executives often fall into, however, is assuming that opportunities for advantage will be available indefinitely. In actuality [as I've illustrated above], the window for gaining advantage from an infrastructural technology is open only briefly." True, but all technologies that at one time offer competitive advantage eventually become commodities. This is nothing new. Competitive advantages created by technology have always been temporary and always will be.

At one time, FAX and email offered early adopters competitive advantages. But do your FAX machines or email give you an advantage over your competition today? Probably not, since your competitors likely have FAX and email, too. New technologies such as web services will offer a competitive edge for a brief period, as long as they're new to your industry. But once they're commonplace, these technologies aren't business differentiators any more than email or a FAX machine. Web services will be required in virtually every industry. Even local Mom and Pop storefronts will eventually use web services to communicate with their suppliers and manage their inventories.

Today, you may be able to achieve some short-term advantage from web services if your competitors don't offer or use them. But at some point, the tides will turn. Once web services become common in your industry, you'll be at a disadvantage if you don't use them. As Carr suggests, the period of competitive advantage will come to an end. "What makes a resource truly strategic--what gives it the capacity to be the basis for a sustained competitive advantage--is not ubiquity but scarcity."

But Carr apparently doesn't recognize that within the huge umbrella we call "IT" there continues to be new technologies that offer competitive advantage even while the older technologies drift towards commoditization. For this reason, IT isn't at all like the railroads or the power grid.

Carr claims that as a commodity, IT has become part of the infrastructure. "Infrastructural technologies...offer far more value when shared than when used in isolation...The characteristics and economics of infrastructural technologies, whether railroads or telegraph or power generators, make it inevitable that they will be broadly shared--that they will become part of the general business infrastructure." In the Computerworld interview, he's quoted as saying, "we're already starting to see that the capabilities of the IT infrastructure are greater than the needs that businesses have."

That's true for long-haul bandwidth and colocation space, but it's hardly the case for the balance of the infrastructure that will be required to support the large-scale deployment of web services.

Carr further suggests that "IT is, first of all, a transport mechanism--it carries digital information just as railroads carry goods and power grids carry electricity." Here I also disagree. If all we needed to do was push bits from place to place, he'd be right, and IT would only consist of routers, switches, and the people to keep them running. But the medium-term future of IT is to share business processes, not just the underlying information. To achieve this end, a great deal more work is needed. This isn't something that will be commonplace for at least another decade.

"IT is also highly replicable," Carr continues. "Indeed, it is hard to imagine a more perfect commodity than a byte of data--endlessly and perfectly reproducible at virtually no cost." But bytes have implicit value that far exceeds their commodity value. The contents of a container of orange juice are only worth the price of the orange juice--unless you're really thirsty. The data that's stored on our disk drives is far more valuable than the cost to reproduce or transmit it. Consider what you might pay to recover the data from a failed disk drive or a lost laptop. Those bytes are hardly a commodity except in the most abstract sense.

One of his recommendations is, "Follow, don't lead." But it's not that simple, for if you wait too long, you'll find yourself in the period of competitive disadvantage, playing catch-up. Being late is at least as dangerous as being too early to adopt new technologies. A successful IT strategist needs a methodology for adoption timing: to get it just right.

In summary, while I agree with Carr that the components of information technology trend (and always will trend) towards commoditization, such has yet to occur for web services, which have barely made it out of the starting gate. Furthermore, there's no reason to believe that web services will be the last IT development of the 21st century. So while some elements of IT have already become commodities, and others are sure to follow, more technologies offering temporary competitive advantage will continue to surface.

[Carr's article was covered by Steve Lohr in the New York Times, then Computerworld featured an interview with Carr as well as a rebuttal of his article by Patricia Keefe. Information Week's editor-in-chief, Bob Evans, provided another counterpoint, Craig Barrett, Intel's CEO, responded at an analysts meeting, and Steve Lohr wrote another article in the Times. John Hagel and John Seely Brown also posted their preliminary counterarguments to be published in full in HBR's July issue.]

IDC: Web Services to Enable $4.3B Hardware Market. As IT moves to adopt web services, IS managers will end up spending more money on hardware than on software, an IDC analyst said. Guess they haven't read Nicholas Carr's article yet.
Posted Friday, May 23, 2003 1:25:33 PM

W3C or OASIS? In this short Computerworld interview, Tim Berners-Lee succinctly and accurately describes the differences between the two standards bodies. OASIS is young, and the standards it's created haven't survived the test of time. While not perfect (and certainly not fast) W3C has produced standards that are relatively stable. (I think IETF has been even more successful, but they're not too involved with this web-services stuff.) Without OASIS pushing things through quickly, it would take at least a few extra years to achieve interoperability in the WS-xxx areas such as security and orchestration. Wouldn't it be great if OASIS took technologies to the point of interoperability and then submitted them (without patent restrictions) to W3C? Hey...it could work!
Posted Thursday, May 22, 2003 11:41:34 PM

Speaking of OASIS...Thankfully, OASIS appears to be proactively addressing the security vulnerabilities of web services. This goes beyond developing YAP (Yet Another Protocol) and gets into the pragmatism of a web-services world. According to the press release, OASIS announced three activities surrounding a "vulnerability description language":

• a classification scheme for Web security vulnerabilities,
• a model to provide guidance for initial threat, impact and risk ratings, and
• an XML schema to describe Web security conditions that can be used by both assessment and protection tools.

Web Services Security--The Book. My on-the-road reading last week was a great book by Mark O'Neill, CTO of Vordell, and others. This is the book to date on the topic. I particularly like the blend of strategy and practice that Mark and the others have achieved. They've managed to get straight to the point: The best way to secure web services today is through XML Signature, XML Encryption, SAML, and WS-Security, and this book explains how those technologies work. I was able to learn a lot more about these topics than from the specifications or the online white papers.
Posted Saturday, May 31, 2003 3:36:41 PM

Reliable Messaging. I was offline for nearly a week (Blackberry access only), so I'm just now catching up. One highlight so far is Phil Wainewright's weblog posting in which he lists three currently available options for web-services reliable messaging platforms: proprietary, JMS, and ebXML Message Service (ebMS). Follow the link to the article by David Longworth, and read Phil's excerpts from his email exchange with Jean-Jacques Dubray of ebPML.org.

In fact, none of the above will cut the mustard long term. Microsoft will never accept JMS, and as Phil suggests, the web-services guys just don't want to embrace anything that comes from EDI. I think we'll see reliable messaging evolve during 2004, once the web-services community (primarily through OASIS) finishes security this year.
Posted Saturday, May 31, 2003 4:10:43 PM

Who Needs SOAP and BPEL? Phil Wainewright also noted that "analysts at Gartner have backtracked on their previous definition of a web service, and now concede that it can be software automation that uses at least one of the three so-called foundation standards -- SOAP, WSDL or UDDI -- rather than having to use all three....Whether you use any of the higher-level web services standards, including SOAP and WSDL, is irrelevant. All you need is HTTP and XML, along with HTML and JavaScript for the user interface."

Phil is right, of course. From Amazon.com to Salesforce.com, we're seeing early but important examples of "real" web services that require not much more than XML over HTTP. I don't agree with Phil, however, when he writes, "they're starting to realize the power of connecting best-of-breed resources in a loosely coupled architecture." I don't see these early implementations as being loosely coupled. But his overall point is well taken, and one that the REST folks have been trying to make for some time now.
Posted Thursday, May 22, 2003 7:54:15 AM

Web Services to Alter Consulting. The rise of web services will force computer-services companies to dramatically scale back their systems-integration businesses and focus on other tasks, according to Jason Bloomberg at research firm ZapThink. (It seemed like this story got picked up by every newspaper and business magazine in the country. Good PR, Jason!) [Source: News.com]
Posted Friday, May 23, 2003 1:27:30 PM

IBM and EDS Lead Web-Hosting Market. Research firm Meta Group concluded in a recent report..."The key differentiator in the market is professional services...Challengers AT&T, Verio, and Sprint have solid hosting offerings, but are struggling to expand into high-end services, Meta said. Meanwhile, Qwest Communications and Cable and Wireless suffer from overinvesting during the boom years, and are focused on regaining financial health." [Source: InfoWorld]
Posted Saturday, May 31, 2003 4:51:32 PM

More news on Loosely Coupled...

Free Books! Teros, a leading application-firewall vendor, will be giving away free copies to qualified attendees in their booth at the Gartner IT Security Summit 2003 in Washing D.C., June 2-4. [Washington Hilton, 1919 Connecticut Avenue, NW, Washington, DC 20009, Hours: 12:00-8:00 Monday, 8:00-1:00 Tuesday]
Posted Friday, May 30, 2003 10:55:23 PM

Last Week's Hottest Title. Loosely Coupled was #1 among the 3,940 books available for download from Books24x7. Awesome!
Posted Saturday, May 31, 2003 11:04:26 PM

Doug's Recent Appearances and Webcasts