On Tuesday, September 24, 2002, at 05:51 PM, T. Kenji Sugahara wrote: > Greg: "The open source community is unlikely to produce massive and > specialized software systems such as what the IRS or FAA or State of > Oregon need." > > But would companies that were historically non-open source produce > open source material to obtain a contract? I don't see why not, if they could make money and have some way to protect their IP. I don't think software developers are against open source, they are just not familiar with it. But in a market-driven capitalist economy based on profit, getting everyone to give their development time and IP away is, to me, unreasonable. I'm straying pretty far off topic, but since some earlier posts on this subject seem to imply that simply trading Windows for Linux will solve the problems in government agencies, I'm trying to expand the discussion. The software of interest at government agencies and big companies is not the commodity stuff, it's the real expensive stuff that has very small vertical markets. I can think of three ways software products make money. Maybe there are more: 1. Commodity products, sold as-is or bundled with hardware. Examples: Windows, MS Word, Red Hat Linux. 2. Base products customized for the customer and enhanced with add-on modules and/or custom programming. Examples: Oracle Financials, SAP, PeopleSoft. 3. Customer contracts for development of software specific to their needs. In the first two cases the developer owns the software and probably wants to protect their IP (which represents a significant investment and a source of future revenue). Keeping the source closed in one way to protect IP. Copyrights, licenses, NDAs, contracts, and even public opinion are another. The market may impose big barriers to entry for competing products (like the customer base Oracle and Windows enjoy) and that barrier may mean open sourcing would have little real effect on sales (one could argue that MS Windows is in that category). In the third case the end user owns the software and the IP (unless they gave that away to the developers). The end user may intend to protect their IP so they can resell their application. They may believe the source code contains trade secrets. Imagine what happens if everyone is forced to open their source, either by mandate or market pressure. Whoever financed the development now has to protect their IP without the simplest defense of not letting anyone else see it. The value of the software drops but the development costs don't. Developers don't have as much incentive to take chances on commodity products or customizable packages, because they risk competing with other companies who didn't have to spend anything on development. If copyright laws could be enforced accurately and efficiently that would be some protection, but as it is now a copyright case is too slow, too expensive, and in the case of complicated software, too hard to prove. End users with money would continue to pay for custom development, but even if they didn't care about competitors they would still worry about trade secrets or business practices embedded in their source code. Open source advocates propose various alternatives, but the business models so far seem to come down to (a) volume sales during the short competitive advantage window, i.e. the hype surrounding new releases, and (b) selling services to end-users. These are both viable models but not likely to produce an Oracle or Microsoft (not that those are companies to emulate). My point is that open source companies will always be nickel-and-dime compared to closed source companies because the open sources don't have IP or a significant competitive advantage. Service companies are necessarily limited by people and time, and unless you are EDS charging General Motors $10K to change a firewall rule you can't make a lot of money selling services to people who paid $300 for their PC and $50 for the software (that they could download for free). I'm trying to make my points based on business rather than politics. I don't have anything against open source, but as a business model I think it has some big problems. Sorry for going so far out into the weeds. -- Greg Jorgensen PDXperts LLC, Portland, Oregon, USA
This archive was generated by hypermail 2b30 : Tue Sep 24 2002 - 20:35:47 PDT