Earlier quoted context omitted.
Out of curiosity, would it not have been a better idea to build your experiments/software on Linux or a BSD variant? Was it a lack of software drivers for your equipment? Given the precise control you can get over the operating environment on these OSes, you can pretty much ensure that the environment your programs run under will never change.
They weren't "my" experiments - I was the station network engineer with double-duty as helpdesk/IT person. Imagine this: you're a broke grad student and you're given $xx grant money to conduct a certain experiment in N months. What are you going to spend your VERY limited time+money on? Finding a good Linux hardware developer to roll some custom solution that no one can understand or support? Or just throw together s…
Just install some distro and "just throw together some hacky Python". If you had some hardware and measurement instruments that require Windows for drivers then I could understand it, otherwise it's just negligence.
> or .net MVP that will run on Windows which practically anyone in the world can troubleshoot with basic knowledge?
In my experience people who claim that they could troubleshoot Windows is usually just that - they claim it but rarely can they actually fix anything.
I don't know, maybe Windows is that much better and easier to use for everyone and I just can see it - who knows...