Apols if you've already seen this courtesy of
ar_gemlad, but I thought it was too good not to circulate.
Everyone knows and loves CAPTCHAs, those distorted pieces of text that you have to decipher and type in to prove you're not a spambot, right? You may even use them on your own sites -- we do in a few places, because some of our forms were getting exceedingly spammed.
Well, now you can turn them to good use -- instead of generating text and distorting it for the purpose, why not use pieces of extant text from old printed works which are sufficiently unclear that they've baffled OCR attempts to digitize them? You can get plugins for various common blog etc applications, and libraries for various common programming languages, to allow your own CAPTCHAs to contribute to Carnegie-Mellon's effort to digitize the many scans of books held at archive.org.
(There's all sorts of clever stuff about how they make sure what the user types in is right, but I won't go into that here.)
![[livejournal.com profile]](https://www.dreamwidth.org/img/external/lj-userinfo.gif)
Everyone knows and loves CAPTCHAs, those distorted pieces of text that you have to decipher and type in to prove you're not a spambot, right? You may even use them on your own sites -- we do in a few places, because some of our forms were getting exceedingly spammed.
Well, now you can turn them to good use -- instead of generating text and distorting it for the purpose, why not use pieces of extant text from old printed works which are sufficiently unclear that they've baffled OCR attempts to digitize them? You can get plugins for various common blog etc applications, and libraries for various common programming languages, to allow your own CAPTCHAs to contribute to Carnegie-Mellon's effort to digitize the many scans of books held at archive.org.
(There's all sorts of clever stuff about how they make sure what the user types in is right, but I won't go into that here.)