More than one blogger is wary of the report simply because they haven't heard of this research firm before. Others aren't sure how the company came up with its numbers, since open source adoption is so hard to quantify. As Navica CEO Bernard Golden told me not long ago in a conversation about the Open Source Census:
One of the key things that enterprises always want to know is "How do we stand in comparison to our peer organizations in terms of what we're doing with open source?" Ordinarily you go to an IDC or a Gartner and say, "What are companies like ours doing when it comes to, I don't know, call center software?" With open source, it's anonymous, so there's no way to really assess that. Companies have often felt, "Well, in the absence of data, we're reluctant to go forward."
So if companies that are using open source can't get hard data to compare what they're doing, how can anyone really estimate what the damage is to proprietary software vendors? That's what the census is about -- gathering hard data. But it's still so early in the process, the numbers don't say much. And like InformationWeek's Serdar Yegulalp points out, even if we know how many companies are using open source, can we really be sure that those numbers also represent losses for traditional software shops?https://o1.qnsr.com/log/p.gif?;n=203;c=204663295;s=11915;x=7936;f=201904081034270;u=j;z=TIMESTAMP;a=20410779;e=i
Especially in this age of the heterogenous IT environment, it seems that a company's use of open source in one area may simply enable it to spend more on proprietary solutions in another area where the open source offerings are not yet mature enough to suit the company's needs. Is that really a loss for the proprietary vendor? I don't know. But it certainly makes one think.
As for questions about Standish Group's legitimacy, I don't have any more information than other bloggers, but I did see a notice Tuesday that its chairman, Jim Johnson, is addressing the SOA Consortium at the end of June. Hopefully by then we'll know a lot more.