The Loken Report

The Report of the APS task force on electronic information systems , from the task force headed by Stewart C. Loken of LBL, was originally published in the Bulletin of the American Physical Society, Vol. 36, No. 4, p. 1119 (April 1991). The Appendices to the report are available in a separate file.

Bringing the Loken report on-line

In the six years since the APS task force on electronic information systems published its final report, much has changed. The World-Wide-Web, and, closer to the principal recommendation of the report, the Los Alamos e-print Archive, have appeared on the scene and accomplished much that the reports' authors feared might take another 25-30 years. No doubt the publication of this report had at least a small effect on bringing these efforts to fruition. But much remains to be done, and it seems appropriate at this time to revisit the recommendations and thoughts of this task force on a subject so vital to the future of physics. A recent sampling of APS efforts in this area is available in the November issue of the APS News.

One might think that a report on "electronic information systems" would have been preserved electronically over the 6 years since its publication in the Bulletin of the American Physical Society. Unfortunately and ironically this was not the case, or at least no useable electronic copies could be located after questions were placed at the APS, AIP, and among the report's authors (the issue of long-term archiving is actually only indirectly addressed in the report). The following is a brief account of how an online version was produced.

Locating a clean copy

After several poor copies were examined, a clean reprint was finally located (thanks to Charlie Muller).

Scanning and OCR'ing

Adobe's "capture" and Caere's "Omni-pro" were both used on parts of the document, with an HP Scanjet scanning the pages at about 300 dpi. Both did an adequate job - Omni-pro was much faster, but had a severe failure reading "th"'s, requiring considerable proof corrections, so "capture' seemed more reliable (but in total took about 10 minutes per page on a 486-66 PC).

Storing as ASCII.

After proof corrections (both OCR packages highlighted words they thought were probably in error, allowing immediate fixes) an ASCII file was generated and stored.

Re-formatting.

The OCR'd text contained a large number of hyphenated words due to the 3-narrow-column format of the original document. Hyphen's were removed, and other minor errors fixed (the original document strangely had "commerical" for "commercial" everywhere).

Conversion to HTML

The document was then modified with HTML tags identifying the various titles and lists (not entirely consistent) in the original document.

Hyperlinking

The final step was to add in hyperlinks within the document, and links to a few external sources.

Arthur P. Smith (apsmith@aps.org) January 13, 1997