In 2009 Jeffrey Tucker led the Mises Institute in open-sourcing this library and putting its books under Creative Commons. As Mises.org webmaster, I was the technical side of that release: roughly 140 GB of books, journals, and recorded lectures, published for anyone to copy, host, and rebuild.
What the release could not include was a way to read it. Scanned books were image-only PDFs no search engine could index, lectures were WMV files no browser plays, and almost nothing carried a transcript or usable metadata.
This archive is the rebuild. Nearly every recording is transcribed now, 12 million words of lectures, searchable for the first time. Every book carries its own rights record, read off the work itself rather than inherited from a blanket claim, and the scans that were page images with no text at all were OCR'd until a search engine could read them. Read online or download PDF, EPUB, MOBI, and audio at no cost. The statistics page states every total, and every gap, with its denominator.
For current work, my economic research and policy comparisons connect these primary texts to modern data, explicit measures, and source-backed comparison methods.