Google announced today a mobile version of google books

booksearch.blogspot.com/2009/0 ...

[Edit: Added last paragraph, added reference to it in title.]

Still not listed here..
m.google.com/?dc=gbackstop

You think they'll make Books its own icon? :)

I think an open blue book with yellow/white pages and black scrawl for text like in cartoons.

Something like the Notebook icon on Android/iPhone but opened up.

(I work on Google OCR)

The OCR errors you're seeing there are the result of running a book with Early Modern English orthographic conventions through a modern English OCR engine. In EME, the letter s is often written in a form that looks like and f (there are other changes, too, see en.wikipedia.org/wiki/Early_Mo ...). It's true that these are still OCR errors, and we're looking to fix this, but that passage isn't really indicative of the general OCR quality. Note that clicking on a paragraph will toggle between the page image and OCR text.

A better selection of good books for the mobile version is in the Featured Books section on the main page (or there's a non-EME Merchant of Venice at books.google.com/googlebooks/m ...).

Nothing in the SciFi/Fantasy category :(

And it doesn't work on the Nintendo DS Browser :(

Matt, thanks, for some reason, that was the only result I got (and still get) when searching Google Book Search mobile for [shakespeare]. Checking it again now, I do see there is a white on light blue "Load more..." message below that single result, but it doesn't load anything after a while. And when I check the Merchant of Venice page you refer to, for some reason I'm seeing "No mobile preview" (from Germany, if that makes a difference)...

What percentage of books in Google Books' public domain collection is Early Modern English, by the way?

PS: I do sometimes get JavaScript errors though, not sure if that is related to using the mobile version with a non-mobile browser...
"Error: this.b is null
Source File: http: //books.google.com/googlebooks/ mobile/B52[...]D9.cache.html
Line: 1149"

Yeah, I believe that the source country does matter here because it uses that to determine a date to use for public domain. I wasn't able to reproduce the non-working "load more.." button as you described (I'm on firefox3/linux).

I don't have a number offhand for the percentage of EME books in the corpus, unfortunately.

I wasn't able to reproduce the javascript error you described. What browser are you using? Is the error correlated with a specific event? The front-end team didn't think that the source country should matter for that.

(Matt, you can contact infoblogoscoped.com please so I can send you some more details on this)