Reality, Promise, And Web ServicesReality, Promise, And Web Services

With much of the hype over Web services finally fading into the background--being replaced by real projects and products, or so I hear--I'm wondering how things are going out there in IT Land.Reason I'm asking is because a couple of recent news stories have made me curious about how Web services are being used these days. In separate announcements, IBM and Microsoft said they're connecting some of their wares to SAP's enterprise software. The idea is to manage SAP's human resources, accounting, and/or inventory applications from the familiar front-end environments of Notes or Office.

Despite the similarities, the respective firms are taking somewhat different tacks, of course. According to an SAP spokeswoman, the SAP-Microsoft linkup is based on something called ESA, or Enterprise Services Architecture. It's Web services all grown up into a comprehensive framework that's managed and tracked.

As Microsoft moves further into Web services with Longhorn and other environments, not to mention Grandchildren of .Net, I'm betting we'll be hearing more about this as Web services and/or SOA become increasingly more sophisticated and enterprises can customize even more pieces of the integration.

For its part, Lotus Notes access for SAP promises to help customers integrate SAP data and business processes with Lotus Notes tasks, including vacation/leave workflow, time reporting, contact management, and report generation.

Interesting that the vendor community is starting to use Web services to (finally) link up various software packages. There will likely be even more of these types of announcements ahead and some interesting bedfellows indeed.

But what of the user community? Sure, we've all heard about the huge early adopters, but what about the other 99% of the companies out there? I'm hearing that Web services fit into an enterprise's middleware scenario to help with some things, but they're certainly not the all-singing, all-dancing cure-all that was first hyped.

As some of you may recall, Web services were promised to do pretty much anything that needed to be done from an enterprise development standpoint. Updating legacy systems, integrating front-end and back-end software, developing new types of software that could take a customer query and build an answer from various Web sites on the fly--all was supposedly possible with this magical technology.

Some early adopters, however, discovered that performance wasn't all it could be, so using Web services to try to power a search engine that got 300,000 hits a day, for instance, wasn't a great idea.

Also, as my TechWeb colleague Max Fomitchev pointed out in a blog entry earlier this year, the XML and SOAP specifications at the heart of Web services are sufficiently vague to allow for different interpretations. That means applications created with different tools may not be interoperable after all, which was kind of the whole point of Web services, wasn't it?

And so I'm wondering how the reality is living up to the promise. If it's like most technology, there are problems and implementation headaches and things you wish you'd known about Web services before you began. But you're probably slogging through anyway because the alternatives are even worse.

What types of projects do you have going, do you consider them mission-critical or still in the proof-of-concept stage, and how did you or your IT folks learn to use this technology? How long did it take (really) to reach ROI, assuming you did? What about the technology has surprised you--for good or for ill? Please weigh in below.If it's like most technology, there are problems and implementation headaches and things you wish you'd known about Web services before you began. But you're probably slogging through anyway because the alternatives are even worse.