Planning his Knight News Challenge entry, Joe Boydston followed his own advice: think small. He won a modest $10,000 to develop a desktop application that will allow web-challenged journalists to drag text files into a folder and have them automatically published online. “Someone at the conference asked me if I regretted not requesting more money,” he wrote in an e-mail exchange. “My answer was ‘Why would I? It doesn’t cost that much.'”

During the conference, though, Boydston, vice president of technology and new media at McNaughton Newspaper Group in Northern California, found his ambitions for his project shifting. His original goal was simply to help small community papers — many of whom were using 10-year-old technology like Mac OS 9, he noted — publish stories to the web with less hassle.

“Early on the second day of the conference (during the wiki/semantic-web barcamp led by Benjamin Mako Hill) it really hit me that the value of this project is not found in efficiencies gained by news organizations,” Boydston said. “The real value is in surfacing valuable data online, so that it can be shared with our communities. It’s not about publishing faster or cheaper, it’s about publishing more.”

Boydston started to envision newspapers using his CMS uploader “to create a wiki of historical documents from a newspaper’s archive.” This trove of old articles could be the starting point for crowdsourcing a project on local community history.

“Taking an asset largely forgotten (old archive stories) and building an interactive, engaged online community around it is raising the bar pretty high for small community news organizations,” Boydston said. In this way, he said, he hopes the CMS uploader “will serve both the technology laggards and the innovators.”

The text files published to the web using his CMS uploader will be indexable and searchable. (At this point, the application won’t work for PDFs, video or audio, Boydston said.) “You might think of the CMS uploader as poor man’s DocumentCloud,” Boydston said, “but only in that it facilitates the creation of an easy to use, online repository of information.”

On his web site, Boydston has a sample upload of articles to a WordPress account. For 900 stories, Boydston notes, “The keyword indexing took about 4-5 minutes, but the actual data import was under 10 seconds.”