1. Momentum.

2. Custom Windows-only applications

3. Really, really good price on support contracts - which they'd insist on getting if they switched to any *nix. Oracle (formerly Sun), IBM, RedHat (?). Are there any other major players in the OS world that can offer anywhere near the level of support that the government has come to expect?

4. And as Igglyboo said, MS Office. So many workflows in corporate environments depend on that damned suite, it'd be difficult to move away from it at the moment.

---

There are many offices that could make the transition. They use all web applications and various office suite tools (Excel, Word, PowerPoint mostly). The other applications they use would be things like Acrobat (but office suites can generate PDFs directly these days) and software to fill out forms (not handled as DOC/DOCX files or PDFs). The latter may be able to transition to signed/fillable PDFs. But these aren't the majority. Too many offices (particularly anything involving project management or inventory management) just depend too much on custom desktop applications that would take too much to port over or rewrite. It's like trying to modernize the avionics systems of an aircraft: in theory it'd be great, in practice your developers will kill themselves.

Because you have to understand the context. For nearly everyone one earth, a computer is a physical tool that exists for the sole purpose of helping them accomplish other tasks. In many cases, the most complex tasks people tackle are those they're being paid to work on. Those people just want a tool that works, that works they way the expect, that is supported and supportable by a vendor and internal IT, and is replaceable with something similar enough that things don't break.

Look, I'm a programmer turned engineering manager and I totally get where you're coming from with this comment, but you just don't seem to get how and why things are the way they are.

Speaking personally, I managed a project a few years ago to replace MS Office 2003 with Libre Office 4.0 (we actually started with OpenOffice but when Oracle acquired Sun and The Document Foundation was created to house the OO.o fork, we switched out of antipathy for Oracle). The mission was to remove MSO from as many machines as feasible, and the decision of what was feasible was, for the most part, left to me. We managed to convert about 20% of users, and these were composed of about 75% machines that never executed any Office app anyway and about 20% of machines whose users were just consumers of files other people created. Only about 5% of the machines converted belonged to creators. We found that most macros and custom formatting and blahblahblah were fairly straightforward to convert to work in LibreOffice, but the UI/UX was horrific, sometimes things just didn't work, and productivity was adversely affected far beyond the cost of the licenses.

We migrated to Gmail and Google Apps from Exchange+Outlook at the same time and saw similar adoption issues with Google Drive ("Docs" at that time -- in 2008). We (I) didn't make the mistake of forcing the issue this time and relied on coercion and organic growth to build the foundation for broad support. Today we are seeing about 1500 Sheets and ~500 Docs & Presentations created weekly (~22,000 users). Most of them still have MS Office on their machine, too.

Because, guess what, the right tool for the job is important.

(another commenter noted the haphazard and often ridiculous release schedule and evolution of RedHat / Fedora, BSD and Ubuntu over the past ten years or so. In my company, the Linux guys are mostly on their own to build their own environment because it became too unpleasant for corporate IT to create and maintain a standard image that worked well enough for most people most of the time. The devs these days tend to prefer Arch and Ubuntu and the sysadmins often run Fedora (our prod servers are almost always RHEL ... because guess what -- support is kinda important to most companies, especially big ones that aren't technology companies and where internal support is often understaffed and underequipped to DIY everything.)

This probably comes off rantier than I had hoped, and for that I apologize. Perhaps I'm just getting to CIO'y in my old age. :(

Because Linux or BSD is not a drop in replacement for Windows XP. It comes at a cost of rewriting their existing working applications and not to mention the time and money required to retrain the existing employees. Does anyone know how good the corporate support is for Linux/BSD distributions? Does it match upto what Microsoft provides?

>How is this a responsible use of taxpayer money?

The same way the govt decides to purchase Cisco's $25,000 wifi routers for city libraries.

At the scale the IRS is running, a copy of Windows is a few tens of dollars, which is about the cost of the necessary replacement keyboards over the lifespan of the computer. Ongoing patch support, however, is where MS actually shines here. Windows XP had support available for 13 years, with subsequent OSes being supported for 10+ years. In comparison, for OSes with support lengths as long as windows, like RHEL, the per-year support seat costs start at $49, which would lead to support costs of near $500 per seat over that same lifespan. So even with the IRS botching the rollout and needing to pay MS an extra year, they're still cheaper than the few Unix offerings that provide the duration of support MS does.

I think it would be great for them to switch but I can think of 2 huge reasons they wouldn't. 1. Retraining the users, very few people use linux so it would be a massive undertaking to retrain their thousands of mostly non-technical workers to use a new, unfamiliar, operating systems. 2. Microsoft Office. Say what you want about LibreOffice and OpenOffice but when it comes to enterprise grade software MS Office is unrivalved, especially concerning Excel.

It would be much more expensive in terms of hours lost to make the switch.