One row in the historical-cipher acquisition record gives an image index of 265 and a manuscript-side label of 131r. Image 265 was not manuscript page 265. Covers and blank sides can be included in the image sequence, while the front and back of a leaf carry separate labels. A later comparison needed to preserve both numbers to return to the same side of the same document.
I followed public collection information and digitized manuscripts to assemble locations for possible comparison material. Working with Gallica at the Bibliothèque nationale de France, I read manuscript image inventories and retrieved selected views while recording image order separately from the labels attached to the manuscript sides. The retained output documents that acquisition path. It did not reach a transcription or decipherment of the cipher.
From an image inventory to a local file
The work used IIIF manifests. IIIF provides a common way to describe and serve digital cultural-heritage images; a manifest lists the views belonging to an object and their order. A canvas is a presentation space for a view. It does not identify an individual cipher passage.
For each selected view, I kept its sequence index, displayed label, canvas address and image address, then connected them to the downloaded filename. The response status and received byte count were recorded too, making the path from inventory entry to file traceable.
I recorded the size of each manuscript’s full image inventory separately from the views selected for retrieval. Confusing the two would turn a download of selected sides into a claim to have acquired the entire manuscript. Keeping the inventory and retrieved files separate allows later comparisons to be tied to the material actually collected.
Two numbering systems needed separate fields
Even the labels differed between the two selected sources. One set of stored rows uses “NP”; the other uses numbers followed by r or v, distinguishing the front and back of a leaf. I therefore preserved both the viewer’s image sequence and the manuscript-side label instead of treating the sequence number as foliation.
That fixes the location of a possible comparison. It does not yet establish a relationship between texts. A document associated with the same person or event might be useful, but its relevance to a particular cipher passage still depends on the writing, date and context. The passage itself has to be located and read before the relationship can be assessed.
What remains after retrieval
Successful response codes and file sizes describe an acquisition run. They cannot establish that a file contains a cipher, that the writing is legible, or that a comparison document corresponds to it. Those judgments require reading the images. A downloaded file does not supply a transcription or a cipher key.
The next question is which passage on the selected sides should be compared. That requires a transcription retaining uncertain readings, a corresponding passage in the comparison material, and a check against existing decipherment research. The record connecting 265 with 131r provides a way back to that location. It cannot establish the meaning of the cipher or a new historical interpretation by itself.