RDVP Class: Open Source Software and e-mandi (1/24/2005)

Today’s class covered two major topics: open source software (OSS) and a new project idea for e-mandi, a virtual marketplace in India, by visiting fellow Vipal.

The first reading for the day was Eric Raymond’s The Cathedral and the Bazaar. We talked a bit about whether it was appropriate to encourage fellows to do their development work using open source technologies, given that some countries were starting to think about requiring it. Jose talked about Venezuela, where a two week debate had raged in congress over whether to REQUIRE replacement of proprietary systems with open source equivalents within two years (unless certain conditions were demonstrated to be true). Jose further pointed out that the Venezuela country gateway, one of the most successful in South America, runs on Microsoft technologies, which might be a factor (since they could focus on content rather than worrying about the technology).

I think the consensus was that while OSS offers interesting development options and is forcing Microsoft to keep on their toes regarding product innovation, the inferior documentation and support that you find with most OSS can be a hurdle. Therefore, it’s not appropriate to mandate that RDVP projects be built on OSS—understand your user requirements, but then use the easiest technology to fulfill them, so you can focus on the application rather than the underlying technology. That said, don’t ignore open source, since some very cool tools (as Carlos has demonstrated with Moodle, etc.) are available and ripe for the taking. One that he especially pointed out was RSS, suggesting that it would be a good way to re-make the RDVP site to have feeds from fellows’ content, yielding a fresh site, combining the recent updates of all of our projects. There was also some optimism about building applications from pluggable components (like a shopping cart).

Vipal has joined the RDVP as a visiting fellow with just a few weeks planned here at Stanford. His plan is to build a virtual market place for agricultural goods in India, complete with an agency to assess quality and warehouse goods until they are sold. He will use the time at Stanford to do project planning: create a business plan, timeline, and budget. He’s found an investor in India who’s willing to back a one-year investigation/initial foray into the project, and so is anxious to get started. His project fits in well with a number of the other fellows’ projects (Dipak, Mans, Moulaye, Helen) so there was enthusiasm for the chance to brainstorm with Vipal. Mans offered some interesting insights about alternative market clearing schemes (suggesting a Dutch auction, where prices descend until someone expresses a willingness to buy, which can lead to greater efficiency, as well as more of the surplus going to the producer…) There were a number of complexities to Vipal’s project, including:

• Logistics (e-mandi would pack goods for export)
  
• Financial Information
  
• Market making (aggregation of demand, splitting or aggregation of supply)
  
• Pricing as market or “limit” orders as in the stock market
  
• Re-using (by leasing) the government-owned food warehousing infrastructure
  
• Creation of a futures market as well as a delivery market for speculation in commodities
  
• Financing by using savings from the discontinuation of food subsidies
  

A lot of big ideas, and a short amount of time, and we welcome his enthusiasm!