Given that nfk might be green and wart-laden and live under a bridge, his comments need to be addressed to prevent confusion among the newer members who browse this section. So, let me chime in just a tad.
Everyone talks about how we professionals should move away from Windows and move to LINUX, but they need to consider economics in multiple flavors - and there is also a flavor of reality to be addressed.
First, we and our companies probably have invested a LOT of time and effort into solutions that are based on Windows and Office. Right, wrong, or indifferent originally, once you have a huge amount of money invested, you need to amortize your investment before you decide to upgrade and replace.
In the case of the U.S. Navy, without naming too many names I can tell you that they have a program for the management of the U.S. Naval Reserve personnel that was upgraded to an OpenVMS platform in the late 1980's and early 1990's. They tried seriously twice to upgrade it to get off of OpenVMS and on to ANYTHING else, but both projects died under their own rapidly growing scope bloat.
They wanted to web-enable it, they wanted to centralize it, they wanted to switch to a more modern database management schema, but the project growth made the powers-that-be step back once they had their cost estimates. That old engineering adage, "If it ain't broke, don't fix it" kept coming into play.
The security aspects of the project got more and more stringent, but economics often trumps modernization, particularly with so many projects for new threats, new issues, new concern. Nobody wanted to upgrade something because they thought it would be totally replaced. That project went through three changes of physical platform, four or maybe five, depending on how you look at it, changes in disk infrastructure, and not less than nine O/S and six DBMS software versions. It was still going when I retired, though I hear that some progress has been made to bring at least parts of it into a more modern web-based presentation.
Second, getting off of Windows is enticing, except that many people will correctly say that it is a case of "the other pasture's grass is greener but from a distance, you can't see how deep the manure was that made it look greener." LINUX and its variants are no piece of cake either.
I was in the department that did system administration for over 1000 servers of various flavors. ALL of them... NO EXCEPTIONS... had security issues that were enough to drive us nuts even as it kept us employed. In a world of threats for ransomware, trojans, viruses, and the like, ALL repeat ALL O/S and platform combinations have their serious vulnerabilities. I saw the security notices including which O/S was affected for each such notice. Windows, UNIX variants, OpenVMS, and Apple were ALL being targeted. So before anyone jumps through the UNIX/LINUX hoop, decide WHY you want it. Look at the security features of the O/S you are jumping to, because if it is a "stripped down" O/S to make it more efficient, how much was stripped?
If you can't field a solid TCP/IP stack, you are a gone goose right there because that is the vector through which most infectious code enters your system. If you don't have a solid file protection scheme that includes Access Control List methodology, you are going to fight tooth and nail to maintain decent protection. If you don't have internal protections to keep process A from interacting with process B (unless that was their designed goal) then you will find yourself struggling to assure that you have good backups of every file on your system. If your O/S doesn't support hardware protection against code modification from an untrusted user, you are toast. So jumping from O/S A to O/S B requires very careful consideration.
Third in the hit parade: It seems like you have tons of problems with Windows and I don't disagree - but the advantage is that everyone and his brother-in-law has developed packages for Windows. LINUX, because of its more "engineering" orientation in its earliest years, was not a developer-friendly environment. To be blunt, the O/S was bare-bones because its original purpose was phone-company switching management, for which extra overhead and a requirement for a lot of resident code were simply not supportable. Developers had to "roll their own" utilities because there were no commercial packages available for what they wanted to do. UNIX variants have started to catch up, but it is still harder to find certain things as layered apps for the things that companies want to do.
At first, if you wanted Computer Aided Drafting - the only CAD packages were Windows based. Typesetting? Windows (or OpenVMS, but nobody could afford the early VMS platforms). Web GUI tools? Windows started it. (OK, UNIX picked that up quickly, but they weren't first.) Keep going. You might want to go away from Windows, but your apps (none of which were cheap) also contributes to the investment that you have to amortize before you can talk about moving to a new environment.
The reality of the situation is that hackers will code for a popular O/S - but if they see that what they want is moving to a newer O/S, they WILL re-target their efforts. And don't forget that there is a HUGE negative out there. UNIX code for the Kernel O/S has been published. Hackers can read the code and FIND those vulnerabilities in LINUX and its ilk a lot quicker than they can find it in Windows code, which has at least NOT been published in its entirety.
The ONLY O/S I know of that has never been truly hacked is OpenVMS, because of the way it was originally designed to prevent code alteration. You CANNOT hack the kernel through external means. You have to become privileged first, and that means social hacks would be required.
Oh, you can hack your own process memory and thereby kill your own process, but there are layers to the Kernel code including hardware memory protection. You know that feature Windows has to protect critical code? They got the idea from OpenVMS. At the DEFCON 9 convention some years ago (which is a hacker convention), an OpenVMS v8.2 was set up for a week. Nobody could hack it. At best, they were able to impose a Denial of Service attack, but nobody actually got in.
So... to nfk and anyone else who is thinking about moving away from Windows and Office, your FIRST question must be "to what?" and you need to investigate the answers to that question thoroughly before doing ANYTHING towards a move. Look at available layered packages that do what your company does. Look at the need to modify existing apps to the new environment, because there WILL be some engineering changes in the move. Look to how it will fit in with your security requirements. Don't jump ship because Windows is an antiquated ship. Jump because your targeted new platform will do the job with some tangible advantage.
... puts away soapbox...