Some IT organizations' doubts about the strength of some vendors' support for commercial open source products is legitimate, but don't paint all vendors with the same brush, says SearchOpenSource.com resident expert Julie Hanna Farris. They may be young, but some commercial open source vendors have come up with hybrid models that include both free software and black ink.
In this interview, Farris offers tips on how to assess commercial open source vendors' technologies and stability. She also describes the emergence of a new hybrid product model -- some parts being free and open, some not -- and how to make the most of it. Farris is founder of Scalix Corp., a Linux messaging systems vendor, and several other IT companies, including Onebox.com and Portola Systems (now part of Sun ONE).
Worries about the support provided by commercial open source software (OSS) vendors turned up as a top adoption barrier for IT managers we recently surveyed. Is that concern well-founded?
Julie Hanna Farris: It's a legitimate concern because the market is so very much in its early stages. It's much the same as doing business with a young company, whether it's open source or not.
After the dot-com crash, companies became painfully aware that just assessing a product or technology is not enough; they have to assess the vendor.
What questions should IT organizations ask when evaluating the viability of an OSS vendor?
Farris: Customers need to ask new commercial open source companies about risks in the product technology, the technology base and the financial/business model risk. If vendors don't have tangible proof points available and customer references then they're not giving you enough information.
Ask suppliers to show how they're going to do business, make money and get to profitability and a sustainable existence. Financial backing is important, [as is] the quality of the investors and how much financial runway the company has. All of those things help to address the risk question.
Besides being concerned about their youth, though, aren't businesses uncertain that commercial OSS vendors can succeed while offering free software?
Farris: Yes. Businesses believe they need to pay something to ensure that there's a viable organization behind it.
IT organizations don't want free software from their suppliers because that actually represents risk to them. It represents risk because it instills doubt about whether there's a sustainable entity behind this technology that's going to maintain it, support it and ensure that it moves forward. If I become dependent on the software, I want to know that there's a viable organization behind it.
So, what's wrong with the assumption that commercial OSS vendors are going to fail because they give away software?
Farris: Open source vendors are not saying, 'We're not going to try to make money.' They're saying, 'We're now going to make some of the core software or the technology free, but we're going to charge for value-added bits and support.'
There are a several hybrid models for commercial OSS vendors:
- One says the software's free, but support, maintenance and updates are not;
- Another is the subscription model, which Red Hat has proven can work;
- The dual licensing model lets you license it under the open source license for free or under a commercial license to get the support, documentation, etc. MySQL has proven dual licensing to be viable;
- Then, there's the model in which a vendor offers some parts of the software open source and free, but other parts of the software are sold under a commercial license and are not free. For the free software, you can pay for support and services and you can also pay for extensions and other parts of the software that are fee-based. This is what Scalix does, and it's proving successful.
Would you say that these hybrid models, as you call them, are catching on with the proprietary vendors that have open sourced some core products and/or code?
Farris: I think that's true in some cases, and it makes sense. There are some things in software that should become commoditized. They're ubiquitous. They've been around long enough. Basic e-mail software, for example, has been on the market for years and is now freely available.
It's the value-added and innovative software that should be paid for, the stuff that a company makes a big intellectual property investment in. I think that's a model that works because customers feel good about getting value for their money and the vendor can be financially viable.
What doesn't work anymore is customers paying for the same basic technology over and over again and being locked into that technology.
Aren't some proprietary companies releasing, or dumping, failing products to open source, rather than just open sourcing core, commodity parts of products?
Farris: Yes, there are examples of companies releasing things into the open source community as the final last ditch attempt to revive it or as an exit strategy. I think those are efforts that have not succeeded because the open source community is so savvy.
It's more legitimate if a vendor thinks there's a community of open source developers that would take the technology, extend it, port it to different platforms, localize it in different languages, integrate it with other things and extend it in ways never anticipated. If that community can add value to the software, then there's a compelling reason to put it into the open source domain.
Also, a company could release a product to open source in order to help lower the barrier to adoption. They put something in the open source domain and that software now can be freely downloaded, used, integrated, extended and customized on a global basis.
There are some companies that may release products to open source because it's hot, it's in, and they want the open source seal of approval, if you will. The benefits of that strategy are short-lived because it's not about adding value.
Open source should be thought about in the business context as a means to an end, an end where value is added, not an end unto itself.