SAP -- Behind the OSS Curve and Loving It?

From FTPOnline, quoting SAP's Shai Agassi on open source software and intellectual property...

Open source will fail to deliver innovation and is more likely to break applications...

First, open source is good at "opening" mature software patterns. For example, about the time Unix was 30 years old, Linux became relatively easy to create. There are some innovations, but for the most part Linux has opened access to the Unix operating system approach. GCC did the same for compilers. Open Office for desktop applications. GCC actually became a vehicle for compiler research. Linux and Open Office less so, to my knowledge.

Second, there is actually a good bit of innovation in OSS. Look at the Java industry. EJB is a disastrous combination of innovation by entrenched IP vendors. As the market realized that, individuals stepped up with better approaches. The Java industry is more rich and I would suggest has had its life extended through this innovation. Not to mention Ruby on Rails, Seaside, etc. beyond the Java industry per se.

It is ironic that these comments are coming from SAP rather than BEA or IBM. SAP's J2EE environment is noticably behind the industry in general and would be advanced greatly by adopting some of the OSS Java innovations.

Third, it is unclear to me that IP goes away with OSS. Google is a fairly successful company based on leveraging OSS to get its IP into the market. Back to SAP, why should SAP invest large resources into a proprietary Java environment when its clear advantage is in its business capabilities, i.e. its IP? SAP has a long history of innovation in virtual machines and server utilization, that is one reason. But 80 percent of their Java environment could probably be OSS and in doing so SAP would approach and begin to surpass the leading J2EE vendors in their innovativeness.

I would consider myself an OSS "realist" in some ways... I agree with comments that many companies are underwriting OSS because they can leverage it for other revenue streams. OSS has to be funded somehow, and I have questions about how sustainable or broadly OSS models can be applied. I for one am happy that Cincom is profitable with Smalltalk. They are a clear example of having met OSS on their own terms. In a position to choose software tools for one purpose or another, there are several instances where I would choose Cincom's non-OSS Smalltalk product. There are others where I would choose OSS tools like Erlang, Ruby, Python, or Squeak Smalltalk.

"But if you look at the most innovative desktop today, Microsoft's Vista is not copying Linux, it is copying Apple."

This seems like kind of a silly quote in the context of OSS and IP. Mac OSX is based on Unix, BSD, and Mach. Darwin, the underpinning of Mac OSX with this BSD/Mach heritage, is itself OSS. Linux is just another implementation of Unix. Less innovative than Mach but unarguably successful and innovative in its industry impact.

And so the other apparent absurdity in this statement: MSFT's Vista has not innovated significantly beyond Mach. Has it? I guess we could argue about its "significance" and its "innovation". I think Vista neither exceeds Mach at the OS level nor other OSS innovations at the user level by an order of magnitude no matter how you measure it. Microsoft is near the top of software IP companies and yet they have not demonstrated innovation well beyond the OSS community.

At least, they have not created innovative software so "compelling" (to use their own adjective) that I believe I *have* to have it, or it *has* to be part of a software infrastructure that I would be responsible for.

Shai -- get busy on your own environment which is well behind the technology curve. And a good way to do this would be to leverage OSS.