It's an old headache that seemed to be going away, when most everyone in the enterprise was running Internet Explorer 6 on Wintel XP.   The good thing about the Microsoft desktop monopoly was that UI developers could count on a predictable "easel", or foundation on which to deploy fat clients or web apps.  Back when I was at Oracle and we had one monolithic suite of web e-business applications, we could get away with testing solely on IE, at least until a growing set of customers insisted we support those rogue Mac users.

But if you're building rich UI apps and trying to leverage the goodies available for PHP, Flex, Silverlight et al, that headache has become a full-on migraine.   Enterprises can't dictate desktops anymore, nor is it cost-effective to try.

Now the landscape consists of:

• XP, Vista and cowboys running pre-releases of Windows, with all kinds of memory and driver configurations.
• At least 4 likely browsers.
• Macs, and Netbooks
• iPhones and other mobile devices with umpteen additional OS and version possibilities.   Okay, the Citrix receiver for the iPhone is not yet production, but could you justify building iPhone support into your enterprise apps over other projects?

Sure, we know the natural tendency for IT is to publish an SLA with supported hardware, OS versions and browsers.  This is a CYA tactic that tries to ignore reality.   Correspondingly, VDI's growth is probably being hampered because its main value is perceived to be about keeping users from messing with desktop environments.

VDI combined with User Environment Management is not just a boon for IT administration.  It provides a contained, controlled environment insulated from platform volatility.  Here are just some of the newer gotchas I've accumulated from the UI development blogs, consistent with my past experience building multiplatform enterprise applications:

• "JA (Joomlart) templates break in IE8"
• "[Silverlight] Javascript detection code isn't digging Safari on Windows"
• "File Upload [using FileReference and URLRequest] was working fine in Internet Explorer, but nothing would work in Firefox."

Of course, that's the engineer's perspective.  However, it's just as likely that the end user is the first to notice the problem, which simplifies to "My app stopped working [and I can't do my job]" with help desk responding, "It's working just fine on MY machine".

One answer is to develop code that you regression test until it runs error free on every platform.  That's the best and only way to go when you're building commercial public apps.  Even then you need to make a business decision to withhold support for the fringes: Windows 2000, IE6, Opera, etc.

If you are supporting an intranet or otherwise captive audience, you can try to limit your certification efforts to a select set of platforms.  Even so, end users will always be ahead of you in trying to use a new browser patch release, or accessing your apps on some new platform...and those new platforms will keep coming.

This is where the true, underrated value of VDI and UEM (User Environment Management) come into play.  You get more than just a desktop environment that is consistent with your limited certified platform set.  UEM's (such as Appsense) preserves users' freedom to personalize their environment, balanced with IT's need to restrict how they access your apps.  So there are no surprises when:

• Your apps are run on hardware outside your control.
• Your users find everything on their desktop where they expect it to be, and behaving the way they expect.   They use your lovingly crafted apps within the one VDI / view manager environment you deploy, whether they're running Mac, Vista or a netbook,

If developers can worry less about how their frameworks and code will run on different platforms, they can focus on making users more productive with the richest possible user experience.

How much more justification do you need?