The ISBN mess: Yet another argument against eBabel and for ePub
Just how to handle ISBN numbers when e-books appear in a bunch of different formats? The more formats, the more hassles for publishers at this difficult time.
Conversion hassles and less choice for consumer are just part of the eBabel mess. How many techies and others outside the book industry can grasp the ISBN complications when so many formats are eBabeling on?
A master format like the IDPF’s ePub standard could at least reduce the confusion, especially without proprietary DRM to clutter matters up. The best DRM remains no DRM.
Meanwhile, to appreciate the extent of the mess, check out ISBN’s on all formats, by Michael Cairns, former president of R.R. Bowker, the official U.S. ISBN agency as well as publisher of Books in Print.
Image: The Confusion of Tongues by Gustave Dore, via Wikipedia item on the Tower of Babel.
June 27th, 2008 at 5:30 pm
While it would be great if the whole world could settle on a single eBook format, I don’t see that happening in the next five years. In the meantime, allowing a single ISBN to be assigned to a single eBook (offered at a single price) regardless of the format it’s converted to makes sense.
For large publishers, ISBN numbers are cheap (a dime each). For small publishers and self-publishers, they’re not cheap at all. Different ISBN numbers for each format would also mean going through the registration process six or seven times for each title, add to the complexity of the conversion process because the original would have to be altered for each format (to include the distinct ISBN), and would add to the challenges of accounting for sales.
Rob Preece
Publisher, http://www.BooksForABuck.com
June 27th, 2008 at 7:17 pm
The fundamental problem is that ISBN is not designed, nor intended, to be used for different renditions of a book, and each different format of an e-book is a different rendition.
As a small e-book publisher myself, I am very sympathetic with the ISBN cost issue, though, and the entity to blame on this is Bowker.
If Bowker wants ISBNs to be used per the standard, then it needs to set up a better pricing structure for small lots of ISBN numbers. I’ve not heard any justification for their current pricing structure.
July 2nd, 2008 at 10:37 am
It is not just cost ISBNs, but Rob it is major point, considering the potential for an exponentially greater number of publications and a myriad of potentially new types of publications that epublishers may produce.
Small publishers are, after all, favoured by the technology — ISBNs are not designed for tiny operations, especially if they are producing many small publications.
ISBNs remain a good way to reference already print published works, but not to identify the electronic versions, which much more readily become new editions.
We need an unambiguous identity system, please see the World Wide Unique Identity system I helped to develop at Lestec or at my own site.
It is still beta, but the idea is that any server with a static IP can generate these simply, and they can never be duplicated.
Having a simple free Unique code system is relatively easy, what we also need is an international database to register the codes simply.
The system I would be suggesting is very simple, authors are given a unique code, their works another,and the particular edition another code, which need not necessarily be registered in the data base.
I am currently working (very slowly) on my first ebook publication, using it as an example the database would record the author unambiguously:
Author: Egon Erwin Kisch (Prague, April 29, 1885 - March 31, 1948) id = jl5lpcsbnleagl
Work: Author: jl5lpcsbnleagl, Australian Landfall (1937), id = uwl5lpcsbnleagl
Print: Work: uwl5lpcsbnleagl, Australasian Book Society, 104 Bathurst Street, Sydney Australia 1969 id = a4l5lpcsbnleagl
Translator: John Fisher id = c4l5lpcsbnleagl
Translator: Irene Fitzgerald id = qd5ndpcsbruwawl
Translator: Kevin Fitzgerald id = pf5ndpcsbruwawl
It has a forward by A. T. Yarwood so he would be a separate entry, along with the forward (written in 1962).
The Print edition would be separately entered as well with its publication data and given another id (it is pre ISBNs).
Then the electronic version given its own edition id, perhaps registered by including some data on the publisher and all the other relevant ids.
Just a suggestion of a way forward, but it has some useful attributes. Over the next few months we are trying to work out a way to always ensure the first character of the id is a-z, because XML identities do not accept a number.
I hope to marry this system with the XHTML/XML namespace so that electronic references can be unambiguously identified.
Hence rod43pcsbnleagl.1.23/5-6 would represent “Australian Landfall” chapter 1 paragraph 23 (p id = 1.23.), the slash represents the next grammatical unit (ie a sentence), so effectively it means sentences 5 and 6 from within the paragraph.
The idea is that we can quote freely and just a fragment is sufficient to trace back to the original work, the electronic edition and via that the original context.
Of course refinements are needed “Australian Landfall” was originally published in German and I do not have the title of that yet, nor any publication details.
Good database design could make a world wide system of reference very easily, allowing for mistakes and simple reassignments, attuned for very small publishers to effectively register their publications.
The system of Ids we are using also allows for easy look up using nested directories, we use a simple reversed time stamp to randomise the IDs and can get effective randomisation easily to 26×32x32×32x32×32 (which is beyond what my calculator can show in normal numbers - 6 levels of directories each containing 32 directories each using the id itself as a key to a lookup table.
The point being, that it is not difficult to come up with a solution that is unambiguous, unique, error-robust, and endlessly extensible.