On a CD-R preserved in the Kenneth R. Shoulders Papers at the Science History Institute Archives, digital archivist Sarah Newhouse encountered over 1,300 files that modern operating systems could not read. The optical disc contained born-digital records belonging to Kenneth R. Shoulders, an electrical engineer who conducted research across microelectronics, aerospace projects, experimental aircraft, and energy research.
Unlocking the Legacy Disk
Shoulders's collection included floppy disks and CD-Rs spanning from the 1980s through the early 2000s. While modern file systems could list the items on the disc, almost all of the files lacked standard three-character file extensions, preventing modern computers from identifying the software used to create them.
Reading the Byte Signatures
To determine the underlying file formats, Newhouse processed the contents with Siegfried, a command-line tool that inspects a file's hex code for specific byte signatures rather than relying on external file names. Siegfried's analysis revealed that Shoulders's legacy documents were written in WordStar, a dominant word processor during the late 1970s and 1980s.
WordStar allowed users to embed inline formatting commands directly into document text, such as ^B for bold text and ^S for underlined text. The application did not require rigid file extensions, encouraging users to invent custom labels like .LET for correspondence or .ev for Shoulders's notes on energy research.
Renaming for Modern Compatibility
Identifying the file format presented a new technical hurdle: modern word processors like Microsoft Word can process legacy WordStar files, but only if the files possess a .ws extension. Renaming files manually threatened to erase Shoulders's original naming conventions.
To preserve the original structure, Newhouse used the batch utility ReNamer to convert periods in existing filenames into underscores—turning names like "beamstop.ev" into "beamstop_ev"—before appending .ws to every file. Opening these converted files in Microsoft Word exposed the document contents alongside visible header metadata and garbled inline formatting symbols.
Batch Migration via Command Line
While opening individual .ws files rendered the text readable, converting more than 1,300 individual documents one by one was impractical. Furthermore, relying on desktop word processors to support vintage extensions long-term presented a risk of future incompatibility.
To solve the scale problem, Newhouse executed LibreOffice via terminal commands to run automated batch operations across the collection. The command converted all 1,300 WordStar documents into standardized PDF files in under 30 minutes, establishing stable formats for archival discovery.
Preserving both the raw byte-for-byte originals alongside stable PDF conversions ensures future researchers can examine Shoulders's engineering records using modern software or DOS emulation environments.
AI-generated editorial illustration; not a documentary photograph.
Source: It’s Written in the (Word)Stars
SIGNAL TITANS
Leave a Reply