This point is not new, and has just been made on the WeoGeo blog for example. I just found enough public data online to illustrate the second keynote in my second-last post, and begin an ROI spreadsheet for proprietary vs. open-source webGIS:
- Gerry James posted some figures for proprietary vs. open source application software costs
- I found on CNET some indicative figures to get a hardware and operational software costs
- Keith Weber posted capital vs. operating cost percentages for proprietary software
- Paul Ramsey also gave a rough breakdown of similar cost percentages for open source
To reiterate: given the right tools, entire systems can be put together for the cost of just the framework of others - $50,000 / €33,000 / £30,000 appear to be a watermark shown in the chart here in yellow, for available budgets at small operators and agencies. This is a crucial point in times of credit crunch for both parties, as their budgetary ceilings are fixed and not getting higher: That made it exceedingly difficult to justify GIS IT; even if the software could be sourced, the remaining cost often went over available funds - no reflection on either software or service, it's the simple truth in business today. In other words even if the capital costs proved affordable, operating costs most often did not.
This is a brave new world, where capital and operating costs are lowered, not client expectation or vendor delivery. In jest, no need to dust off old bookshelves and bring back old hardware to achieve this either:
(also on my Youtube channel)