Lost Opportunity 1

In a previous post I described a simple networked or standalone, (depending on data definitions), document imaging processing project I led.  This system would have been perfect as an Open Source project for the following reasons:

1. The need for such capability to be provided as infrastructure was universal.  The main constraint at the time to this happening being the high cost and complexity of the current applications.  So adoption would probably have been quite rapid
2. The target user community tended to be very network centric, as they were often charged with distributed access to large central collections of information.
3. My company had no interest in the software,  its interest was in an efficient way to capture and distribute its image data as efficiently as possible
4. The system was very extensible, allowing additional storage drivers, scanner drivers, printer drivers, viewers and databases to be added by other developers.  The automatic maintenance of these components in line with new hardware advances and the increased deployment reach would have been very valuable
5. There were many areas of potential improvement.  Some of the concepts were not fully realised, and many functional/feature improvements such as OCR were missing

 It’s also likely that the community would have seen the natural evolution of the product to a web application, and created that UI and made the necessary database access module changes, saving my company the need for that re-write.  Finally vendors of specialist image viewers, scanners and printers/plotters may have sponsored development as it would have enabled them to provide a working system, rather than a component. 

In addition to the raw savings in development cost for my company the potential benefits for the community were greatest. The provision of document imaging capability, a truly universal need on every device and in every business, has never happened at the infrastructure level, but rather in an ad-hoc way at department level in most cases. This system not because it was the best, but because it was very simple, made few assumptions about its data, and few assumptions about its environment could have been the start of making this happen.