Microsoft does not appear in any imminent danger of losing much market share on the desktop productivity front to any Linux or open-source competitors.
In a presentation at this weeks Gartner Symposium/ITxpo here in San Francisco titled “Client OS and Office – Charting a Course to Longhorn (or Linux),” Michael Silver, a Gartner analyst, said that just because Linux is free does not mean it is cheap.
Among the impediments to using open-source office software products among businesses are compatibility and fidelity issues.
“You just cant move all of your users to StarOffice/OpenOffice.org. You will have to keep some Microsoft Office. You will have to look at big swaths/large departments and groups of people relatively isolated from others and who pretty much only send documents between themselves as potential users,” Silver said.
Asked whether the Macintosh might be a better choice than Linux on the desktop at the moment, Silver said it could be, as there is Office for the Mac and it has a better, more intuitive user interface.
But Silver said this is also a good time for business and enterprise customers to take a close look at Linux and StarOffice/OpenOffice.org on the desktop “as we now have a better idea of whats coming in Longhorn [the next version of Windows] and how compelling that will be.”
“Longhorn will include a single worldwide binary that can be used to reduce the number of images companies have to deploy across the world,” he said.
Longhorn would also see the concept of LUA (Least User Access), where users were no longer given administrative rights and applications did not break as a result of those lesser rights. Longhorn would also bring better search, better ways of categorizing and searching documents, Silver said.
Linux on the desktop for mainstream business users has also “passed the peak of hype” and real deployments are starting, so “well now see what some of the holes actually are,” he said.
But Silver also cautioned the audience not to believe all they hear about Linux on the desktop, listing what he sees as the 10 myths around this. These are that:
- Linux is be less expensive than Windows because StarOffice/OpenOffice.org can be used instead of Microsoft Office
- Linux is free
- There are no forced upgrades. (“We expect there to be as little support for older versions of Linux as for older Windows,” he said.)
- Linux requires significantly less labor to manage
- Linux has a lower TCO than Windows because of the available management tools
- Applications are inexpensive or free
- Skills are transferable
- The hardware can be kept longer if Linux is used, or older hardware can be used;
- Linux should be deployed as soon as a Microsoft Enterprise Agreement expires, and
- Linux on the desktop is an all-or-nothing equation.
Microsoft Office
Turning to Microsoft Office, Silver said Office 2003 is a far more compelling version than Office XP has been, which is reflected in the fact that XPs adoption peaked at about 30 percent of users.
Silver said he is expecting the next version, Office 12, in the late 2006 time frame.
Given that many customers have an enterprise agreement for Office, if Microsoft pushed the release of Office 12 out more than three years after the release of Office XP, some of these customers might not get an office upgrade, which would cause a lot of dissatisfaction and unhappiness, he said.
The version of Office beyond that, Office 13, could be expected around the time WinFS (the Windows File System, which was pulled from Longhorn) is released and will likely take advantage of some of the features and functionality in that release.
“The big question is whether you will want to roll out Office at the same time as you roll out Longhorn. The majority of uses move to a new version of Office when they move to a new version of Windows,” Silver said.
“So, what is Office 12? I dont know,” he said, to much laughter. “Microsoft is not saying very much about it at the moment, as they want you to still buy the versions currently available. But Id expect more Outlook improvements in Office 12 as well as better search, more XML, more integration with collaboration and a lot more around tying Office to a users business applications,” he said.
With regard to the Windows operating system, Silver said that just because Microsoft has committed to supporting a version of Windows for 10 years does not mean that users should stay with that version that long.
In a fall 2004 survey of the client operating system installed base, Gartner found that in the United States, 60 percent of users are still running Windows 2000 on their desktop, but the year-end outlook for XP looked good for desktops and even better for notebooks as upgrades continued, he said.
Microsoft would also continue to support Windows XP until 2013 in some way or another, with support for Windows 2000 running until 2010.
As most people keep their desktop for four years and their laptops for three years, Gartner suggested customers start buying Longhorn machines starting in 2008.
But Silver cautioned that this would probably mean they would have to support several versions of Windows for some time.
“We have a lot of customers who intend on skipping Windows XP, which is not unusual, particularly in the pharmaceutical industry, where they have to certify applications to the operating system and need to fulfill legal and other requirements,” Silver said.
He then listed five reasons why a customer might want to consider upgrading to Windows XP if they have not already done so:
- They can reduce dependency on Windows 2000 support;
- They can reduce their dependency on the Longhorn ship date;
- They can begin a leisurely migration to Longhorn; and
- Windows XP Service Pack 2 brought improved security.
But if customers are planning to skip Windows XP on new computers, there are a few things they need to do, Silver said.
Among these is ensuring that their ISV will support Windows 2000 through 2011; having contingency plans for critical applications that require Windows XP; beginning serious testing of applications on Longhorn during its beta cycle, especially beta 2; and ensuring that their ISVs will support Longhorn within 12 months of its shipping.
“You also need to budget to migrate all users off Windows 2000 by mid-2010,” Silver said, adding that he did not expect Longhorn to be released before the second half of 2006 at the earliest and that it would take some 18 months after that before 10 percent of the installed base had migrated.
If Microsoft wants to ship Longhorn in time for the 2006 holiday season, it needs to release this to manufacturing sometime in September, with a November launch.
“I would expect to see the release candidates next summer, following the second beta, which will be the first time we will really see the new user interface and plan around that, expected January 2006.
Gartner also did not recommend that users skip the first version of Longhorn and wait for the next, which could contain WinFS, which Microsoft pulled when it decided Longhorn would be driven by release date rather than by the feature set, he said.
Check out eWEEK.coms for Microsoft and Windows news, views and analysis.