Google To Offer 4000000 titles

[Sonist]

From Japan Today:

"U.S. online search giant Google Inc has clinched the support of almost all publishers in the United States for its digital bookstore expected to be launched as early as the end of June, company officials said Sunday. The number of authors and publishers that have agreed to participate in Google’s sales effort to distribute books online topped 25,000 as of Sunday, the officials added.

If books with expired copyright are included, Google will handle over 4 million books, including about 2 million handled by the publishers that have agreed to join what is set to be the world’s largest virtual bookstore. ..."

So, 2 million titles handled by publishers. Wonder if it means we'll get good formatting?

[Kali Yuga]

If Google's existing offerings are any guide, probably not. I'm sure they will use the same scans and poorly proofread OCR'ed versions of all their PD and scanned books.

This is mostly old news, though it remains to be seen if the number of titles has any real meaning -- especially since the book scanning settlement may get shot down. I can't imagine the mess that would result if they start selling books prior to all of that getting finalized.

The only benefit here is that Google's offering will be browser-based with some kind of cached offline option, thus more cross-platform than most other offerings.

[Sonist]

Originally Posted by Kali Yuga

If Google's existing offerings are any guide, probably not. I'm sure they will use the same scans and poorly proofread OCR'ed versions of all their PD and scanned books.

...

I was referring to the part about 2 million books handled by the publishers.

My hope is that if the publishers are involved, we might get conversions from "source" files, instead of scans. Or cleaner scans.

[Dellaster]

Originally Posted by Sonist

My hope is that if the publishers are involved, we might get conversions from "source" files, instead of scans. Or cleaner scans.

I hope so too. I've perused a couple dozen of Google's current scans and consider all those I've tried unfit for reading. Possibly the blurry, often broken type is due to flaws in the online reader and downloads will be better. Or are they currently showing low-quality raw scans, not OCR? That's certainly what it looks like.

[Marseille]

Originally Posted by Sonist

My hope is that if the publishers are involved, we might get conversions from "source" files, instead of scans.

It's not even a question. Google scanned material because it did not have cooperation of rights holders or former rights holders, and ergo, had only the print to work off of.

Originally Posted by Kali Yuga

If Google's existing offerings are any guide, probably not.

Not giving Google the benefit of the doubt with commercial titles because they dared offer a fairly unique (for a corporation that is; hat tip to gutenburg) solution on older materials is quite unreasonable. On the contrary, Google has always gone above and beyond and there's no reason to expect any different as they move forward with a mainstream bookstore.

I for one, can't wait. I hope it's not cloud-only, just cloud enhanced, but time will tell.

[Worldwalker]

Google's scans of public domain books are better than the Internet Archive's, but just barely. So it's not a rights issue. I'm not holding out a lot of hope here.

[Marseille]

Originally Posted by Worldwalker

Google's scans of public domain books are better than the Internet Archive's, but just barely. So it's not a rights issue. I'm not holding out a lot of hope here.

What isn't a rights issue? Once Google has the rights to use modern commercial titles, they won't have to scan them at all. They'll use the digital version from the publisher just like every other ebook store does.

[Kali Yuga]

By the way, PC World says Google will have 500k books at launch, not 4 million:
http://www.pcworld.com/article/19556...ly_summer.html

They also didn't have any (public/confirmed) publisher commitments as recently as May 4:
http://online.wsj.com/article/SB1000...417931818.html

In fact, everything I'm seeing about this alleged "4 million books, half provided by publishers" not only sounds highly improbable, it only has one source: one article in Japan Today. I'm not even seeing an official press release on Google Editions. So, I think I'm gonna have to doubt this figure and claim until it's officially confirmed.

It is a good thing that they've made lots of public domain titles available, but the poor quality of those texts, they certainly hasn't earned much faith when it comes to upcoming ebook quality.

Originally Posted by David Marseilles

Not giving Google the benefit of the doubt with commercial titles because they dared offer a fairly unique (for a corporation that is; hat tip to gutenburg) solution on older materials is quite unreasonable.

Good, because that's not what I'm doing.

Legalities aside, Google poured significant resources into their book scanning project, and captured millions of titles that way. I have no doubt that is the source of a very large number of books they will sell, and there is no reason to imagine that anything captured by that project will be of better quality than what is currently available. Nor will they have access to any source material that isn't used for other formats / retailers.

Originally Posted by David Marseilles

On the contrary, Google has always gone above and beyond and there's no reason to expect any different as they move forward with a mainstream bookstore.

Above and beyond, eh? *cough* Dejanews *cough* Orkut *cough cough*

Google, for all their virtues, are basically a bunch of engineers who also have the capital to buy neat toys. Half of their Cool Stuff were acquisitions -- SketchUp, Google Earth, Picasa etc. Oh, and as Business Insider points out, in case you have any doubt where Google's revenue comes from....

Originally Posted by David Marseilles

I for one, can't wait. I hope it's not cloud-only, just cloud enhanced, but time will tell.

It will have offline caching, but I wouldn't expect too much more than that. Google loves the cloud. I'm also sure that, much like Amazon or Apple, they can yank a title at any time if it was necessary.

And don't forget, pretty much everything you do that's connected to Google is another information stream to them, largely to feed their advertising services. That's not necessarily a bad thing, just keep it in mind.

[Worldwalker]

Originally Posted by David Marseilles

What isn't a rights issue? Once Google has the rights to use modern commercial titles, they won't have to scan them at all. They'll use the digital version from the publisher just like every other ebook store does.

Google's lousy scans aren't a rights issue.

Whatever the reason is that they don't think they need to make the book, y'know, readable, it's not rights. Their PD books are just as awful as, if not worse than, their non-PD books.

Also, given the number of publishers selling ebooks that seem to be made from bad scans of their own books, I'm not holding out a lot of hope that any but the most recent books will be something other than scans, with or without the cooperation of the publishers. It's been claimed elsewhere in these forums that the publishers themselves don't have digital copies of any but their most recent (relatively speaking) books.

I don't see how it can possibly be a bad thing to have another source of e-books. Especially if they are made available in format(s) not tied to any particular device.

Yes, there versions of the OCR output of scanned public domain books are terrible, but they are free so it is hard to get too worked up about that. What I like is that in addition to the OCR output version they always provide the original scan, in PDF format. I wish P.G. did this given the number of errors that often remain in the various text versions released by P.G.

My big hopes for Google Books is that a lot of 'orphan' titles, those written to late to be in public domain, but not current enough that any publisher thinks it worthwhile to go to the expense releasing these titles, will at least become available on Google in some form of text output as well as the PDF.

[Marseille]

Originally Posted by Worldwalker

Google's lousy scans aren't a rights issue.

Whatever the reason is that they don't think they need to make the book, y'know, readable, it's not rights.

No OCR is perfect. The only way to correct OCR errors is by hand. Man power is the reason Google can't correct it all. There are some techs intended to alleviate that -- like the captchca's that use internet users to interpret bad scans. I don't know if Google uses that, but either way, it would take a long time to correct a library that large two or so words at a time.

Their PD books are just as awful as, if not worse than, their non-PD books.

Up till now, they haven't had the rights to not use scanning to get even their nonPD books. Once they have agreements with publishers, they won't need to scan books. Some titles won't have digital versions available at all, and scans will still be there for that. But yes, the reason there is scanning at all for titles that are new enough is a rights issue.

[mantas9]

My big hopes for Google Books is that a lot of 'orphan' titles, those written to late to be in public domain, but not current enough that any publisher thinks it worthwhile to go to the expense releasing these titles, will at least become available on Google in some form of text output as well as the PDF.




----------------------------
phentermine
levitra

[Worldwalker]

Originally Posted by David Marseilles

Up till now, they haven't had the rights to not use scanning to get even their nonPD books. Once they have agreements with publishers, they won't need to scan books. Some titles won't have digital versions available at all, and scans will still be there for that. But yes, the reason there is scanning at all for titles that are new enough is a rights issue.

In one of the other threads on backlists and ebooks, it was mentioned that the reason so many publishers' own ebooks are of utterly horrible quality is that they are in fact scanning their own books -- they don't have electronic versions of books going back more than a few years. And it's not the new books most of us are looking forward to. It's the books that nobody -- including the publisher -- has in electronic format. Rights or no rights, they still need to be scanned, OCR'd, and proofread, and so far there's no reason to believe that this will be done any better than it has been for books in which no rights issues exist (PD books).

Maybe Google will be good to us ...