Not sure who downvoted your GGP comment, you did have a valid point. Nonetheless, I think there's hardly ever a justification for asking other people to perfor

Not sure who downvoted your GGP comment, you did have a valid point. Nonetheless, I think there's hardly ever a justification for asking other people to perform "a day or two" of work to set up something, even in exchange for years of solid deployment. A day or two, multiplied by thousands of Windows developers, adds up to several man-years of entirely avoidable work. Entirely avoidable, because it's Google's job, not ours, to package up this Omaha thingy into a neat binary with a neat API that any project can easily use. It's not too difficult to ship a single binary (or collection of binaries) that will work out of the box on every Windows Vista/7/8 PC, with the only dependencies being the Win32 API, vcredist, and perhaps a recent version of .NET Framework. Wrap it all up in an .msi and you're good to go. No need to tell a C# developer, for example, to meddle with Visual C++ or Python. If the Omaha devs can spend a few weeks to make the setup process as simple as that, now that would have a real chance of adoption.