Gartner predicts, errs on open source

Speaking at an afternoon session today at Gartner Symposium/ITxpo in Orlando, FL,  Gartner research vice president Mark Driver estimated the probability that four key enterprise open source trends would come true as follows:

• 80 percent probability: By 2010, 75 percent of mainstream IT organizations will have formal open source acquisition and management  strategies.
• 80 percent probability: By 2008, open source software solutions will directly compete with closed source products in all software infrastructure markets.
• 70 percent probability: By 2010, mainstream IT organizations will consider open source software in 80 percent of their infrastructure focused software investments.
• 60 percent probability: By 2010, mainstream IT organizations will consider open source software in 25 percent of their business software investments

Personally, I find that last probability number to be a little low.  But then again, five years from now, Google may have turned the entire industry on its ear and maybe nobody will be using open source applications because they won' be using locally installed software.

The well-attended session was clearly an icebreaker for IT pros who are trying to get up to speed on open source for the first time.  Very little if any of the material covered ground that those well-steeped in open source would have found useful.  In fact, Driver spoke out of school himself on a couple of occasions. For example, he said that of the 17 or so different types of licenses that Microsoft uses for its software, that none of them are actually open source licenses (he said some come close).  This isn't true.  Microsoft actually offers some of its code under the Common Public License (an IBM-authored license that, in my estimation, contains unacceptable terms of mutually assured destruction that are explained here). 

Driver also blamed Microsoft's failure to embrace open source on the immature antics of the Open Source Initiative -- also patently false in my opinion.  In discussing the risks of using open source software, Driver was also too cavalier in his use of the terms intellectual property, copyrights, and patents.  While he seemed to have a pretty good grasp of the copyright issue as it relates to open source code, he tossed the word patent into the conversation as though it was the same thing as a copyright or that patents could simply be grouped with copyrights as something that all open source licenses cover when they don't. 

As a reminder, in the context of open source, copyrights apply strictly to source code (the code that most all software is written in).   Just because an open source license has granted you the right to copy or modify a program's source code doesn't mean that the program itself isn't infringing on a patent.  In other words, when you agree to an open source license that comes with some chunk of code, your chances of being sued for willful infringement on a copyright are pretty low (although if the developer of that code copied it from copyrighted code-base, he or she could be sued).   But, if that code (a.k.a.: an implementation) infringes on someone's patent, the copyright will do nothing to protect you from patent infringement.  For that, you need a patent grant and even then, such grants only assure you that the benefactor won't sue you.  It doesn't mean that someone other than the benefactor who thinks that the implementation infringes on their patent won't sue you. 

Driver also said that there were 40-something open source licenses.  There are 58 (a number that pales in comparison to the number of commercial "closed source" licenses).   Perhaps Driver (and Gartner) would be better off sticking to a discussion of or open source trends and value propositions while leaving the legal part to an experienced open source attorney.