124
submitted by nothingberg *

Or it was to foil search engines and AI in the future..?

Nah I like my narcissist demon empowerment idea.....don't say his name, he'll just get stronger!

Jarnes Corney

Or it was to foil search engines and AI in the future..? Nah I like my narcissist demon empowerment idea.....don't say his name, he'll just get stronger! Jarnes Corney
Comments (6)
sorted by:
You're viewing a single comment thread. View all comments, or full comment thread.
Lambinator (1 point)

Yes, I can think of two technical issues that could cause that:

One is scanning and ocr’ing. Ocr is ‘optical character recognition’, i.e. turning scans of printed documents into editable text. I use one of the world's leading ocr softwares regularly, and with some fonts it has great difficulty distinguishing between m and rn. Many of the science journal papers I scan and ocr contain multiple instances of the word “modern”, and my ocr software renders a very high proportion of these as modem instead of modern (grr!) This is the opposite error to what happened in this Comey-to-Corney case, but it still seems plausible to me that it could be an ocr error.

A second technical quirk that could conceivably be involved is ligatures. Ligatures are when two touching letters are joined into a single exotic dual-character, sometimes with an extra little stroke added. In olden days it was standard to join certain combinations of letters as ligatures. Think æ instead of ae in “Encyclopædia Britannica”. Even though use of ligatures was pretty much abandoned 50 to 100 years ago in the West, the world’s number 1 publishing software comes with ligatures turned on as the default setting (!?!) It is a real headache for many publishers, and can prevent searches of pdf documents from working because the document contains exotic dual-characters ligatured-together rather than the ordinary two letters in your search term. Anyway, it seems conceivable to me that with some fonts, changes between m and rn could be a ligature issue.

That said, I strongly suspect this Comey-to-Corney case was a deliberate mischievous global search and replace by some prankster office staff. ☻

Or maybe someone wanted to prevent searches on the term “Comey” from working?

Lambinator (1 point) *

Also, if their ocr software was set to ‘auto-correct spelling’, then unless they explicitly added the word “Comey” to the ocr software’s dictionary, then it would treat every instance of “Comey” as a spelling error and convert it to the nearest legitimate word it could find in its dictionary, which seems to have been—appropriately enough—the word “corney”. ☺