While Microsoft Corp. continues raising Longhorn, rivals are seizing the operating system"s extended adolescence to develop competing feature sets of their own.

The already scaled-back version of Longhorn is still roughly 18 months from shipping, and with the expected technical advances by Linux competitors during that time, Microsoft"s estimable industry influence to sell the product as currently constituted will be severely tested.

Even with the much anticipated WinFS file system chopped out and sent back to its ostensibly eternal beta cycle, Microsoft believes the remaining technology goodies in the Longhorn bag will compel many business users and consumers to upgrade.

Microsoft, indeed, has put considerable resources into the Avalon graphics subsystem and Indigo technology for connecting subsystems via Web services. In addition to Avalon and Indigo, the first release of Longhorn is expected to have new information management tools including a built-in desktop search capability, management tools designed to significantly reduce deployments costs (including capabilities for image creation, editing and installation), and better all around reliability through a diagnostic infrastructure that detects and fixes problems faster.

But will even the sweetest of those treats taste a little stale by late 2006 as rivals deliver competitive offerings? And given that it has already said its unified file system will not appear in Longhorn, will Microsoft be able to deliver any killer feature that will inspire customers to upgrade?

Still far off into the future

With IT shops not anticipating any bold new features in the first version of the product, many are not planning to make a meaningful commitment to Longhorn any time soon after its release.

The IT department that runs MoMA (Museum of Modern Art), for instance, has standardized on Windows, but is not yet even thinking about Longhorn.

"We haven"t had time to evaluate Longhorn," said CIO Steve Peltzman. "We change things like that when we have to. If I can spend money to save money, I"ll do it, but unless it"s significant I try to stay away from that."

What increasingly is shackling Microsoft to deliver innovative capabilities is the enormity and diversity of its user base, which makes sliding innovative features into Longhorn difficult even over an extended period of time.

"In my view they are stuck between a rock and a hard place. They can"t offer best-of-breed in a few areas because they are so wrapped up in trying to satisfy the needs of all their different constituencies. Their attitude is "we will do only the things we do well and do them the best we can," whereas their competitors can concentrate on targeted and sometimes more interesting things," said Bill Cornfield, president of WSG Systems.

In the meantime, many experts believe that leading Linux distributors such as Red Hat and Novell, along with their top-tier business partners IBM and Hewlett-Packard, will continue to make inroads against Windows, particularly on the server, over the next 18 months. They suggest that by pushing technologies such as virtualization, a number of other planned network-based strategies built around Web services, and improved security, their offerings will be as technically competent as those of Longhorn"s by the late 2006, early 2007 time frame. With Apple"s current momentum, some believe it can nibble away at the low end of the desktop market.

"Linux is a credible platform for a lot of users on servers and even for some on the desktop. But Windows is ubiquitous and enterprises have so many resources surrounding it that users have to seriously considering upgrading. It will be important for [Microsoft] to continue with Longhorn what they started with Windows XP Service Pack 2, namely crisp execution of Trustworthy Computing and security initiatives. Otherwise they leave an opening," said Stephen O"Grady, senior analyst at RedMonk.

Killer features

What Longhorn lacks as of now is a handful of features so compelling that customers simply will not be able to afford, financially or competitive advantage-wise, not to upgrade shortly after the operating system becomes available.

If there is one feature that would make Longhorn more compelling to most users in lieu of dazzling innovation, however, it would be rock-solid security. In fact, the lack of such security is the number one reason many IT shops may be driven to more seriously consider Linux on servers and Apple on the desktop. If it delivers, Microsoft can further cement its operating system monopoly; if not, it stands to lose significant market share.

"[Longhorn] is Microsoft"s chance to design in security from the bottom up. If they do a good job at it, they will be forgiven for a lot of things like the file system not in there and it being very late. If they screw it up, you are going to see IBM and a lot of other people waving their Linux flags," WSG"s Cornfield said.

There was hope among some that with Microsoft"s acquisition last week of Groove Networks the company would be stitching some of Groove"s peer-to-peer communications and other collaborative features into the first version of Longhorn. At the news conference announcing the deal last week, in fact, Microsoft chairman Bill Gates said his company would be melding the best of Groove"s "peer-to-peer and authentication capabilities," in with its own similar technologies to strengthen Longhorn.

Other Microsoft and Groove officials, however, were less than optimistic about planting meaningful Groove-based technology into the first release of Longhorn, which would mean it would not appear until at least 2007.

Although he thinks Microsoft needs to add something to Longhorn and, in fact, is not ruling out that they won"t, Michael Cherry, lead analyst for operating systems at Directions on Microsoft, believes Microsoft should not attempt to integrate something as significant as core technology from Groove.

Citing Microsoft"s own formula for approaching operating system development, Cherry said there are three things to be considered: quality, time, and features, although the company can only change one and still succeed. In Longhorn"s case they can choose quality and have a complete set of features, which means it takes a longer period of time. Or they could choose to ship it on a certain date, allowing them to vary quality or the number of features.

"With Longhorn two things have happened. One is they have fixed time and they have fixed the quality of things like Trustworthy Computing. So no matter how cool Groove is I would vote not to include it. Time is not going to move but quality will," Cherry said.

Microsoft declined to comment on exactly what the final feature set of Longhorn would be, saying only that it is "still under development" but that it will deliver major improvements in user productivity and "significant advances in security, deployment, and reliability," according to a company spokesperson.

API Challenges

One issue that could hold Longhorn back, especially the server version, is the new set of APIs Microsoft is planning for it, which will cause those having bought fairly recent versions of the company"s popular server applications to buy new versions or to do a significant amount of rewriting to those applications for them.

"You have always been able to run all kinds of obscure pieces of older software in a compatibility box, but Longhorn is filling in all that backwards compatibility with new APIs. This is a problem for users who recently bought new server versions, but we hear it is also causing Microsoft"s server development teams like BizTalk and SharePoint to ask, are we not going to do another release until after Longhorn so as to be compatible?" said Mike Drips, an independent consultant in San Francisco who works with a number of large end-user companies and government agencies.

Others do not think that the new APIs are substantially different than the existing ones and, even if there are critical differences, they believe Microsoft will improve their compatibility or make both sets available within Longhorn.

"The new API sets look nice and clean and they are much closer to the .Net Framework. Right now they are holding their cards close to the vest, but at WinHEC (conference in early April) they will have to lay those cards on the table so hardware manufacturers will know what is going on," Cherry said.

A Microsoft spokesman would not comment on the company"s plans to offer enlightenment about Longhorn"s APIs, disclosing only that Microsoft will discuss the upcoming operating system in more detail at its Windows Hardware Engineering (WinHEC) conference in late April.

Inevitable Migration

There are more than a few industry observers who believe that no matter what Microsoft delivers or does not deliver, or what their archrivals do or do not do over the next 18 months, the Redmond-based software giant is guaranteed a certain amount of initial success given its enormous technical and financial resources, marketing machine, and political sway in the industry.

MoMA CIO Peltzman explained that as much as he is a self-proclaimed "big proponent of open source" and would like to tap the open source community, internal IT resources prohibit that and, in turn, are part of the reason the museum standardized on Windows.

"I would love to use it," he said, referring to open source. "But we don"t have the bandwidth for trying open source and hiring" personnel with Linux expertise.

That said, Microsoft, shackled by its vast user base, may have to move slowly in terms of technical innovation, but they may eventually find a way to wield that user base as a lethal weapon against smaller competitors.

"Microsoft is both in an enviable and difficult position. They are the dominant supplier of desktop and server operating systems and so many people have built five-year IT strategies around Windows and have no plans to move away. The difficult part is that no matter what they deliver some contingents are not going to be happy because they have been handed this monolithic, unified block of an application platform," said Dan Kusnetzky, vice president in charge of system software research at IDC.

As for the army of Linux and other open source software competitors marching towards Microsoft"s castle with torches and pitchforks, they will have great difficulty even crossing over the moat.

"Most companies are going to eventually move to it because [Microsoft] will get a tremendous renewal rate among current subscribers through things like Software Assurance, and among individuals just because they want to be the first kids on their block to have it," Drips said. "You can talk about the open source competition all you want but they will amount to a flea on a rhino"s rear end."