Fifteen Worthy Reads from October 25, 2018

Comment of the Day: Maynard Handley in "How to Convert Document from Google Docs to Text File"Even more useful is that Google Docs provides access to what has been (in my experience) by far the best OCR system available. Upload a JPG (and I'm guessing probably other image formats like PDF scans, but I haven't yet tried that) to your Google Drive, and then in your browser in the Google Drive window, right/ctrl-click on the image and choose "Open in Google Docs". You'll get a document opened with the text of the image. I've found that this works not just for the easy cases, but even the tough stuff -- small low quality images, multiple columns, things like that. A little more hassle than the various OCR+scanner apps I've paid for but vastly higher quality. (Supposedly MS Live can do the same thing, but I tried getting to MS' OCR from a dozen different angles, on iOS and Mac, through the web and through OneNote, and gave it up. MS may have the greatest OCR scheme on earth, but they've hidden access to it so well it's useless to me.) (Apple have adopted a strange tactic WRT to OCR. Many things on an iOS13 system are automatically text OCR'd, like any scans or images you dump into Notes or Messages. And this text is indexed, so that the relevant scans/images appear in searches. But you can't get at the underlying text for other purposes, whether to edit it or just to read it. It's unclear if this was just not enough time to ship by iOS13, or if it's an attempt to warn other OCR vendors to find some other app category soon, so that there's less grumbling and the usual anti-competitive complaints when Apple does ship. Either way, this particular design choice means that, at least right now, I can't yet compare the quality of Apple's work to that of Google.)...

#commentoftheday #2019-10-27