'Web services' create hype, but what about applications? This article by Chris Oakes in the International Herald Tribune, is one of the best introductions to web services. I recommend forwarding the URL to any non-technical people who want to understand.
Posted Thursday, March 28, 2002 12:51:47 PM

Java Takes the Bait. Today at Java One saw a slew of announcements of support for web services. Particularly, consider web-service client announcements in mobile devices from Nokia, Vodafone, NTT DoCoMo, Verizon, Sprint and Motorola. Nokia, the top maker of phones, has said it expects to ship 50 million Java-enabled phones by this year and 100 million by 2003. But who's going to profit from web-service enabled phones? Not Nokia. It's revenues have collapsed now that most of us have cell phones. They're forever relegated to the replacement market, slugging it out for market share. Even if they manage to sell 100 million of 'em, there won't be much profit. And it won't be the carriers that win big either. They, too, have been reduced to selling a commodity.

As Clay Shirky suggested nearly a year ago, it's Microsoft that is so brilliantly positioning themselves to reap the benefits. Gates and Gang must have been watching the news from Java One this week, and saying, "Yeah! They went for it." Those phones represent 100 million web-service clients, and it's the web-service provider(s) that will benefit the most. For Microsoft, that's .NET My Services. Get your competitors to mass-produce a commodity client for which you can then supply high-profit services. Brilliant. And just think...Microsoft doesn't care one bit that the clients are written in Java.

Just one final question to ponder: Can AOL Time Warner pull together a web-services offering to capture the mobile loyalty of their 35 million steady customers before Microsoft takes them away? (Maybe they'll finally team up with Yahoo to get this done.) It's no longer just a battle over no-revenue instant messaging at the cell phone. Now it's shared calendars and everything else in Hailstorm/My Services. For a monthly fee, of course. [Dangling thought: Jabber, the open-source XML-based protocol for IM and more, starts to look interesting.]
Posted Tuesday, March 26, 2002 10:47:32 PM

Investment Opportunities in Web Services? Many people would like to make a few bucks off of web services. Surely, they believe, there must be some legitimate business models underneath all the hype. Well, maybe not. At least one major and well-respected venture capital firm is saying "No" to any business plan that touches web services. Their take is that it's all hype.

Who will make money? Not the J2EE players: IBM, BEA, HP and (finally) Sun. While they have no choice but to compete with Microsoft and one another, none of them can use web services to offer anything new to their customers. These vendors have the opportunity to lose market share to one another by failing to offer adequate tools, but almost nothing they can do will cause substantial dollars to switch in either direction across the Java/.NET boundary. Microsoft, on the other hand, has the ability to create entirely new business models, providing software-on-demand to those platforms they don't already control, such as mobile devices. (IBM's Global Services will, I believe, become the leader in professional services consulting for web services.)

One niche that gets some attention is web service networks (WSNs). The most notables are Grand Central Communications and Flamenco Networks. And just today a newcomer appeared when Blue Titan launched at PC Forum. This morning, I interviewed Dave Spicer, CEO of Flamenco Networks, and I'm meeting with Grand Central next week. As soon as I can reach Blue Titan, I'll report back with an overview of the sector. I've also yet to complete my interviews with the EAI vendors--another potentially profitable niche in web services.

Web services are an important development, and they'll forever change the way we link systems to one another, but other than perhaps the WSN and EAI niches, I have to agree with those VCs. No one other than Microsoft and perhaps IBM through professional services has a way to turn that significance into cash.
Posted Tuesday, March 26, 2002 8:26:23 PM

Business Process Management. As web-services protcols become the standards for enterprise-application integration (EAI), what role will BPM methodologies play? Some recent opinions from others (I'm still formulating my own):

• Hurwitz Group recommends that enterprises keep their eyes on Web Services as part of a BPM solution. Look for improved process support and management support for Web Services platforms. Point-to-point Web Services integration could lead to SOAP scum -- a residue of integration without BPM.
• Gartner: "We believe that workflow, which is managed by BPM (business process management) is the way that organizations are going to go."
• Jenz & Partner refer to web services and BPM as an unbeatable team. Dieter Jenz is tackling the ROI and TCO issues in a multipart column. (Part 1, Part 2)

XML in Financial Services. For $795 you can download the full report from ZapThink Research. Or you can read the summary of findings for free. While the report considers XML to be an important technology ("expenditures on XML technologies in the Financial Services sector will grow to over $8.3 Billion by 2005") it's doesn't see the same for web services ("toolset immaturity and the impact of B2B and web services standards will impact integration efforts"). [Source: Leon Benjamin, The Ecademy]
Posted Monday, March 25, 2002 10:33:09 AM

XML Databases. Anyone who's developed an object-oriented system with a complex relational backend will enthusiastically enumerate the advantages of a native object-oriented database. While legacy applications will long be based on the relational model, new web-services applications may benefit from native XML databases. Some interesting references:

• Ronald Bourret maintains an XML Database Products web site.
• Ron also has a page of XML/database links.
• Oracle Corp., IBM and Microsoft Corp. are taking steps to make unstructured XML data fit as comfortably as traditional structured data in their respective data management systems. [eWeek] I don't expect these gateways to offer much in the way of performance, but they may turn out to be useful for providing XML-based access to legacy relational data.

UDDI Myths. Read Julian Bond's comments on a Gartner Viewpoint article in ZDNet Tech Update, Four myths about UDDI, the Web services registry.

• Myth 1: UDDI is mainly for locating new (or unknown) partners
• Myth 2: UDDI must be used as a public registry
• Myth 3: UDDI will eliminate trading-partner agreement problems
• Myth 4: UDDI is only usable through machine interfaces

Web Hosting Strategy. The latest Web Hosting Matters newsletter from Internet World includes an article by Carl Burnham, in which he suggests that the services most in demand include performance monitoring, content distribution and security. Carl's book, Web Hosting, and mine Strategies for Web Hosting and Managed Services, both hit the bookstore shelves late last year. Carl's book is written for those who want to provide web-hosting services, whereas mine is for those who are web-hosting customers. But I think you'll find both books are valuable to the other constituency as well.
Posted Thursday, March 28, 2002 11:01:37 AM

Colo Customer Priorities. Potential customers using the "Find a Colo" search on colosource.com said what they are looking for in a colocation provider.

Posted Wednesday, March 27, 2002 9:36:49 AM

And Then There Were Three. In his weblog, John Robb asks about Speedera:

This is the new hotshot company in content delivery networks. Doesn't this management team look like it is a little light on technology experience? Also, this VC team looks to be a rejection of standard SV firms. Doug Kaye, what is your take on this company?

In fact, Speedera is one of only three standalone CDNs that are still around. The other two are Akamai and Mirror Image Internet. Akamai is now bearing the burden of their huge network consisting of more than 13,000 servers in more than 1,000 networks in 63 countries. They also lost a great CTO when Danny Lewin was killed aboard one of the airplanes that hit the World Trade Center on September 11. Speedera and Mirror Image Internet's networks are based on a more centralized architecture that uses far fewer locations but larger servers. (Of course it's all covered in my book. :-))

The remaining CDN players (all three of them) are hurting because their fundamental value proposition aren't currently high on customers' lists. The two primary benefits of using a CDN were (a) to lower the cost of content delivery, and (b) to deliver it more quickly and more reliably. The former isn't working because of the artificially low cost of conventional long-haul bandwidth. The latter is less important than during the dot-com era, when over-funded companies thought site performance would be a significant competitive differentiator. Oops.

In search of profits, Akamai and Mirror Image Internet are getting into the delivery of applications from the edge of the network. Akamai is the driving force behind edge-side includes (see ESI). MII wants to deliver XML/SOAP-based Web Services from the edge. Look for intriguing announcements from MII in about two weeks.
Posted Monday, March 25, 2002 8:10:47 PM

Moving Up the Pyramid. Although MSPs may give the impression they'll "take care of everything," many of them only go as high in the service pyramid as databases and application servers. They stop short of supporting applications. But Tim Wilson thinks we should look more carefully at Loudcloud's acquisition of ASP Frontera who supports higher-level applications from BroadVision, Vignette, E.piphany and ATG. Tim says, "Many corporations that had considered those applications too costly to deploy - or too risky to rely on - may reconsider their views. And other vendors, including Loudcloud’s competitors, may follow suit."
Posted Monday, March 25, 2002 7:59:38 AM

View or search newsletter archives
Email Doug or visit his site at www.rds.com
Permanent link to this issue