Low hanging fruit is often picked first. That could happen“Gartner reckons there are eight million Windows Server 2003 OS instances in operation, and SI Avanade reckons that of those instances, a full 20 per cent â€“ 1.6 million â€“ will blow past the 14 July end-of-support date.” with millions of servers running that other OS. The cost of migrating thousands of server-based applications is a huge cost to organizations large and small. Even if all they have to do is move applications to a new server, it’s a big job and M$’s 2003 version is on its last legs.
Typically, organizations are forced at least to review their applications. They have to check out licensing, duplication of function, unused, not useful and other categories. Almost certainly, they will think which, of all these, can be done on GNU/Linux? Not surprisingly, many server-applications can be run on GNU/Linux either as is or with substitutions. The more upgrading can be done by distros rather than local IT, the greater will be the on-going savings due to those who make the effort to migrate stuff to GNU/Linux rather than that other OS.
I migrated an application once. It had a Java front-end on MySQL. Remember Java, the write once, run everywhere software system? Well, some idiot had embedded “C:\” all over the place. I just moved the data to a GNU/Linux system running MySQL and the job was done. Instant access to data, no need for paying M$ for a server-licence, no need to install that other OS, just a few minutes work and it was good to go. Another application was not supported on GNU/Linux clients but GNU/Linux with FireFox worked immediately. Some guys just don’t know about open standards.
No matter. Millions of servers will not be properly migrated to M$’s next release on time so they might as well migrate to GNU/Linux and take charge of their futures rather than be dependent on M$’s schedule and desire for cash. Why should your organization uproot itself because another organization wants a fresh influx of cash? Do the right thing. Go to Debian GNU/Linux and FLOSS. It will work for you, not M$.