Why Open Source Isn't Adopted by Business
I was reading an article calling for the government to make more use of open source software and it rang some bells. I'm the IT Manager of an SMB, so I've usually got one eye on our spending and am looking for ways to cut those costs. As such I looked at converting our business to run on open source alternatives, possibly not across the board, but as we have a split between what those at head office use their PCs for and the much simpler tasks required at the branches, I thought it could potentially help save money.
As we've already bought the hardware and they came with licenses for Windows there wouldn't be any direct saving in the costs of the operating system if we switched to another one (Linux for example). Even on new machines, the cost difference between a machine supplied with Windows and without is about £70 (assuming you can find a vendor who sells machines without it) which equates to about £1 a month over the typical three-year life of hardware. Not exactly massive.
We could potentially (perhaps unadvisedly) save the cost of our anti-virus software. Longer term there are potential cost savings in being able to continue using the existing hardware when Microsoft stops supporting XP in 2013, certainly the existing machines will not support even Windows 7.
Against it I had to weigh the cost of the extra time it would undoubtedly take to support users on a new OS.
Something that fell against any open source software was the fact that most computer users have used Windows and Microsoft's various other products, notably the Office suite. Although the branches don't use Office (with the exception of a few copies of Outlook) they do use viewers and, in some cases, OpenOffice.org (or derivatives) to open MS documents.
I've definitely found OpenOffice.org (or derivatives) harder to work with and in some cases slower to use (partly due to familiarity, partly because the product is just slow).
Again, any questions would be coming my way, so that would mean more time spent supporting users.
Talking of support, there's not just the issue of users, but also the administrators having to maintain an unfamiliar system. That has potential security, performance and time implications.
It's much less of an issue than it has been, but alternative operating systems mean you have to be aware of hardware support issues. There are still problems with certain hardware not being compatible, much more so for any legacy hardware.
That may mean you have to throw out or spend time making existing hardware compatible, which may actually end up costing you money. You will also have to be careful in the hardware you purchase going forward.
As I mentioned, in our case, we would probably end up with some machines still running the Microsoft platform, which opens you up to problems when supporting hardware, software and exchanging things between the systems.
You may suddenly find you can't reuse systems or components as they're not compatible, you end up supporting, administering and maintaining two separate platforms and doing something simple like sending a file to a colleague may become much more of a challenge (the newer MS formats still aren't well supported by anything non-Microsoft and even if a package states it does support a certain file type I've found advanced features often don't work).
Even if we could control this internally, we're still working with other businesses who will continue to send us files and information in formats we may not be able to open correctly, or at all. It doesn't make you look very professional.
Microsoft get a lot of stick for their various products, but I haven't come across an open source alternative to beat Outlook and Exchange. It has a good range of compatibility with devices, a web version, and its very common meeting invites work inside and outside the organization. No real bad points. It's not a deal breaker necessarily, but it's very handy.
Compatibility with Windows
This is similar to hardware support, but we use some software and even websites which only work on Windows (ActiveX controls, for example), services which we have little or no control over and which we have to use. That's before you consider some of our server-side software. So we would need to maintain some sort of access to Windows regardless of what we decided to move to. That takes you back to the whole duplication argument.
In fact, stepping away from Windows means you say goodbye to a whole lot of software and while you may be able to find open source/compatible replacements for some, maybe even most, Windows offers a far greater array and (generally) guaranteed compatibility.
Another area some open source systems seem to suffer from issues with multimedia support, partly down to licensing of codecs and such, which a paid operating system can afford to include. Again, probably not a deal breaker, but another mark against them.
On the Plus Side
Well, there are potential cost savings on software going forward. OS, Office and other software licenses aren't cheap. Once XP disappears we're probably looking at having to replace all of our hardware at the same time, with the sort of outlay no business wants to hit their cashflow.
Likewise costs for Office, anti-virus and various other pieces of software are ongoing so not having to pay for them would mean more capital saved that could be ploughed into training for users and administrators.
As I mentioned, XP becomes unsupported (as it stands) in 2013. At that point we'd be forced to look at a different OS anyway. To make use of our existing hardware we could switch to a lighter, open source OS, which we could strip down and either run that or use it as a thin client and visualize the desktops if we wish. We haven't made any decisions on that yet, but we may have to start testing in 2012 to see how each option holds up.
Longer term I wonder if open source desktop software will largely be needed, or most desktop software for that matter, with companies buying hosted services instead, with the exception of a few applications that deal with data that is too large to handle over an internet connection.
Having looked at it, I eventually decided that the costs savings, especially in our hybrid environment, we're not going to be enough to justify the move, let alone the cost of supporting and administering a new system. Most of our costs are already paid, with the investment made. Going forward it's issues with user acceptance and interoperability that concern me the most, likewise the possibility of having to support two OS environments.
Those concerns only really multiply with larger businesses. If the quality of some of the open source tools and their interoperability improves then they may become viable. For basic tasks they're currently okay, but getting users pulling data into Excel from a data warehouse running on SQL Server is currently a lot easier using Microsoft or other closed tools than any open source equivalent I've seen. And that's just an example. Until open source can replicate every function currently provided by existing software and platforms, they're never going to break through.