In part two of this interview, Dave Uhlman, CEO of Unversa Inc., names the apps he considers enterprise-ready, such as MySQL, and how IT managers can cash in on the savings to be gained reduced licensing and support fees by strategically pitching open source applications to higher-ups. Uhlman recommends targeting that IT managers target specific IT needs and choose the best possible fit for their environment.
What about the concerns that open source software may not be enterprise-ready for deployment?
David Uhlman: It's true of any software. It all comes back to the vendor. Are you comfortable working with that vendor? Does the vendor have the capacity to serve your needs? Have they been around long enough? Software companies, as a rule of thumb, have come and gone regardless of whether they're open source or proprietary, so there are other decisions that come into the process that aren't purely open source. What makes an application enterprise-ready is not necessarily the software license.
What qualifies an application as enterprise-ready?
Uhlman: One of those things is feature set. An app should have a rich enough feature set that it meets business user needs 100%, not 10% or 20%. There are a lot of devils in the details that can torpedo an entire conversion.
Second, it has to do with the experiences of other people either in a similar industry or situation and being able to draw on those to look at
It's easy to do something once or twice. It's very hard to do it on a larger scale. If there's something else out there that's working for a 5,000-seat installation, there's a very good chance that it will work for you. At the same time, we need pioneers in open source. We need companies that are looking for a competitive edge and willing to absorb some kind of risk to do that. That's also where open source plays in. It's potentially a huge strategic advantage for companies that are willing to take some risks and exploit it.
Which proprietary apps do you think are the best candidates for replacements with open source software?
Uhlman: Your typical office, e-mail suite are definitely targets. You generally see applications with very specific, very small market segments. A printer or desktop publishing company with only five or 10 users might find people who've written software that services their need directly. It's probably not applicable to anyone else.
There are smaller companies that have decided to solve their own problems and make their products open source. If you're in certain types of agriculture, there are solutions for that. There is some good irrigation control vertical market software that is open source and available. It's not very broadly applicable, but it can make a big difference to a greenhouse farmer looking to gain that competitive edge.
How should IT directors go about trying and pitching these applications?
Also, there are people who are 'best of breed' decision makers who might have a mixed environment of several different tools. In that sort of environment, open source is always a fair consideration. You can use the same process you use to decide whether any piece of software -- open source or not -- and choose those alternatives that are out there. Certainly using resources like Sourceforge.net is a great way to find products that you might not be aware of that are open source, partly open source or at least depend on open source stack.
Which applications should IT managers be paying attention to?
Uhlman: IT managers should look at the interplay with MySQL and how that's really making the biggest difference in a number of companies. You will probably have to do some application re-engineering, but the payoff in savings over a five- or 10-year period is in hundreds of thousands or millions of dollars while improving the performance that you're seeing.
MySQL has done an excellent job in approaching that market, and it's just one of those back-office things that you don't care what the brand is on the front. If it does its job and does it well, that's all that you need.
Go back to part one of this interview.