– Microsoft vs Open Source

This week is the Microsoft Public Sector CIO summit in that village named Redmond “across the pond” from Seattle. It’s also a week of continuing rotten economic news for public and private sector alike. In this environment, it sure is tempting to chuck Microsoft’s Office and web products and their complicated Enterprise and Select Agreements in favor of open source equivalents.

But you know what, the City of Seattle is not going to do that. Why?

Regular readers of this blog – if there are any – know I’m from Seattle and most of you know I’m a serious supporter of Microsoft software and products.

Clearly, I’m prejudiced.

Microsoft provides 40,000 jobs in my area, we have hundreds of thousands of shareholders (many of whom are also constituents) living here. We benefit from the tremendous wealth which has flowed from the around the world into Puget Sound to literally thousands of people, institutions and non-profits in the region. That wealth flows elsewhere, of course, too. The Bill and Melinda Gates Foundation is doing wonderful things for schools and libraries across the nation and around the world. Microsoft research and technology centers are at many locations outside of Puget Sound – indeed there are about 50,000 Microsoft jobs OUTSIDE of Washington State.

On the other hand, all governments have budget pains. I got my first official budget cut memo three weeks ago (we’ve been doing actual budget cutting for at least 9 months). In the past I’ve had to lay off people based almost solely on seniority (or, rather,”juniority”). And I’ll undoubtedly be doing it again at some time in the future, if my job isn’t cut first!

Microsoft’s licensing costs are a large part of our budget, as are the maintenance and licensing costs we pay to Oracle, IBM, ESRI, and many other vendors. We do need to examine alternatives and options.

But I’m somewhat baffled that any CIO of a large government would seriously consider using open source software for our mission critical systems and services. This seems a little bit like using cell phones to dispatch police officers and firefighters or outsourcing your help desk to India. It will save money in the short term and work pretty well “most” of the time …

What is the advantage of using software from Microsoft – or Oracle, or ESRI, or Peoplesoft, or Hansen or … any major software vendor?

No business large or small would seriously consider writing its own financial management system, even though, with web services, database software and a spreadsheet program we could probably do it. We could probably cobble together a computer-aided-dispatch system or work management system from similar components.

The advantage of off-the-shelf or “shrink wrap” is that it is pre-written for us, the bugs are fixed, the upgrades are provided and – of increasing importance – security issues are handled and addressed.

Sure, you’ll say, Microsoft software is really prone to security flaws and attacks. Why is that? Because it is the most popular and ubiquitous software in the world! Its logical that any software which reaches significant market share will become a target for teams of hackers employed by terrorist-nation-states and crime syndicates. And the software for open source is on the web and freely available for such hackers to view!

Now, I understand that open source is supported by a developer community, and that’s good. But this developer community is nebulous. It is a difficult place to find when something serious goes wrong. Governments now rely heavily upon technology to provide critical services and interact with constituents. CIOs are responsible to elected officials keep that technology reliable and available. To depend upon an amorphous “community” of developers with no direct stake in your mission is a risky proposition.

Few businesses -other than local governments – have technology systems so important that people’s lives are actually in jeopardy when those systems fail.’ Sorry, I don’t want a “nebulous” community supporting my public safety and utility system.

Next, in an open source world, what do we do about application integration? Gee, almost every vendor writes their software to work with Microsoft Office, Exchange/Outlook and similar products. Even hardware vendors such as Nortel or Avaya or Motorola will make sure their hardware/software integrates with Microsoft. If there is an issue with the way PeopleSoft HRIS or Government Financials works with e-mail software or office software, they will always fix the Microsoft integration first. When a hot new product comes out – like BlackBerrys – the vendor will make sure it works with Microsoft software right out-of-the-gate.

Believe me, I know this first hand, since the City of Seattle was (still is) a GroupWise e-mail user. I had department directors knocking down my door to get BlackBerrys but the GroupWise version was released FOUR YEARS behind the Exchange/Outlook BlackBerry.

Furthermore, many of our applications now vitally depend upon web services for their user interface. Most of those applications vendors will not be officially supporting open source versions of web services anytime soon.

So, if we – government CIOs – move to using open source software, how do we handle the support and integration?

Answer: like everything else, we hire smart people. Highly proficient technical people who understand the bits and bytes of how this stuff works and can make it happen. Managers who can develop networks of people in other jurisdictions and in the open source community to fix the bugs, get the new releases and work with the integration. Skilled “open source” employees who are dedicated to our mission of “making technology” work for our government and the people we serve.

Well, where is our budget pressure? Yes, it is in revenues and budget dollars. But it is also in FTE – headcount. How many times have each of us been told to reduce headcount? What is the one number (again, besides raw dollars) which newspapers, the public and elected officials always watch and measure? It is “Number of Government Employees”. There is constant pressure – even in good times – to hold the line on headcount, if not actually reduce it.

And when we do reduce headcount, what positions are cut and who is laid off? It is always the last hired, which are usually the youngest, tech-saavy (at least on new software or open-source software), most connected employees.

With open source not only will we have to increase headcount, we’ll become vitally depend upon those new hires and that additional headcount to make our most critical and important applications work.

By making us MORE reliant on headcount and FTE, I think a move to open source software actually exacerbates our budget problems.

On the other hand, elected officials and those with budget oversight are much more likely to accept payments to our software and hardware maintenance vendors as necessary requirements. They all have personal experience with technology, if only their cell phone and desktop computers. They all understand the need to maintain cars and buildings and computers.

But how much of our core and critical work can really be “crowd sourced”? Do we really want to open-source computer-aided dispatch systems or records management systems which have personally identifiable data or arrest/911 call information? And I’m very nervous about open sourcing any part of SCADA (utility control), or traffic management or other control systems which are vital to our governments and targets for attack and compromise.

In these high-pressure, budget-constrained, headcount-hunting times, use of open source software appears to be a high-risk, low-return proposition at best, and a “government fails” newspaper headline at worst.