On Fri, Dec 10, 2010 at 1:57 PM, Don Stanley <[email protected]> wrote:
> > If you want to see an example of the core group going commercial, take a > look at Red Hat and their resulting open source Linux effort "Fedora". > > Those who wish to go commercial are already doing so with no disruption to > the EMC2 effort (only enhances it). > Since I have an opinion on what makes Fedora successful, I thought I'd comment on the similarities between EMC and Fedora/Redhat. Simple economics says that in the long term, the sellers must control the value of the goods they are offering. Since EMC itself is not controllable this way, the value must lie elsewhere: the EMC vendors must offer their customers excellent service, or integrating expertise, or other skills. The EMC software is the foundation and the tooling, and it's the services that is the real product. Similarly, the professional trades such architects and builders aren't selling CAD software and building materials; they just use them to construct a house for their customer. Redhat's modus operandi is similar: they essentially offer a maintenance contract (support, updates, assurance, etc.). The Fedora project is their R&D division: they put a lot of engineering and other resources in it, with the expectation that the results will directly go into into their commercial product. In that sense, everyone who uses Fedora is doing Redhat's QA---it's a tradeoff of some possibility of breakage, justified by all the goodies that come down to Fedora much earlier than to the safe but stolid Enterprise version. The novel principle that Redhat understands is that OS software is now a service rather than a manufacturing industry. They can't differentiate on the software: they could maybe get more customers if they put some proprietary enhancements and divergences in their commercial OS---but overall they are better off by developing all the new features in the open/free version. Any hypothetical business offering EMC-driven machining hardware would need to follow that model, but it's a hard psychological argument to make to a person coming from the old, 'software as a manufacturing industry' worldview. The caveat is that Redhat is really putting a lot of resources into Fedora, by contributing engineering and administrative/managerial resources, not to mention infrastructure. Even the governance of Fedora is pretty amazing: they actually have MEETINGS which result in concensus ACTION ITEMS that someone FOLLOWS UP on :). Their software workflow is pretty impressive too and quite automated: Bugzilla bug and issue tracking and Abrt automatic bug reporting, Koji build and packaging system, and the AutoQA testing rig they are working on. The bottom line is that it would take humility and strength for a company to have a successful relationship with EMC. After all, they would need EMC as much or more than EMC would need them, so they would have to build trust and prove their usefulness. Difficult, but doable. ------------------------------------------------------------------------------ Oracle to DB2 Conversion Guide: Learn learn about native support for PL/SQL, new data types, scalar functions, improved concurrency, built-in packages, OCI, SQL*Plus, data movement tools, best practices and more. http://p.sf.net/sfu/oracle-sfdev2dev _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
