[log in to unmask] writes


> **PROJECT DETAILS**  
> The UC Davis University Library is launching a project to digitize the
> [Amerine wine label collection](https://www.flickr.com/photos/brantley/sets/72
> 157655817440104/with/21116552632/)

  Some look like hard to read.

> and engage the public to transcribe the information contained on the
> labels and associated annotations.

  This may take a long time. I suggest rather than doing that, take
  somebody in a low-income country who speaks French, say, and who will
  type all the data in. That way you get consistency in the data.  I
  live in Siberia, I can find somebody there. Once this data is in a
  simple text file, you can use in-house staff to attach it to the
  label images in your systems.

  Crowdsource sounds cool, but for 4000 label it makes no sense.
  If the typist gets $10/h, and gets 20 labels done in 1h, we
  are talking $200. The visit you are planning for your developer
  will cost that much. 
-- 

  Cheers,

  Thomas Krichel                  http://openlib.org/home/krichel
                                              skype:thomaskrichel