next up previous contents
Next: Bibliography Up: Conclusions and Future Research Previous: Educational research

INDIE and tool design

The INDIE development model represents one model of a ``GBS factory''--a few programmers maintaining and extending a code base while knowledge engineers use the code through a tool to construct new and vastly different applications in many domains. This increases the overall power of GBS builders, but doesn't seem to provide the ultimate GBS solution. GBSes like INDIE's are still extremely expensive and are only useful in some contexts. We still need to hire ``scribes''--content staff--to take what experts know and turn it into a GBS.

This is fine for very broad areas, such as required courses for medical students, where tens of thousands of people need to learn this domain each year. Tools like INDIE make it possible to turn out three or so a year with a very small programming maintenance force, which means GBSes are getting cheap enough (and interesting enough) to justify their creation. If our true goal is to overturn both traditional education and the current educational software market, though, we need to lower the time-per-GBS ratio another order of magnitude, and make it so we can turn out twenty, thirty, or even a hundred GBSes a year. This speedup, then, is the ultimate future work left to do on INDIE.

Works Cited