IDs in zhook

[Francisco Treacy]

Hi all,

How do I declare a book's ID in a zhook? I presume some meta property,
but I can't seem to find this in the spec.

Thanks,
Francisco

[Joseph Pearson]

Short answer:

<meta name="bookid" content="xyz">

Long answer:

The unique book identifier is an EPUB concept that has no analogy in
Zhook. I'm not really a fan of the idea — it seems like half a
solution to a not particularly significant problem. For our project I
just use file hashes to identify identical ebook versions, and
internal database identifiers to link ebook versions to a
"publication".

So it really comes down to how your Zhook->EPUB conversion tool
obtains a unique id. Peregrin uses the "bookid" metadatum and
generates a 12-char random string if it's not found. Peregrin's
predecessor, chook (which is still active on publisher.ochook.org for
now), used the 'identifier' metadatum, else the 'isbn', else a URL-ish
construction based on the internal identifier for the upload.

I'm happy to take patches for Peregrin's behaviour — probably it
should fall back to ISBN like chook did, but I haven't given it enough
thought.

- J

--
Joseph Pearson | software inventor | inventivelabs.com.au | +61394163198

[Francisco Treacy]

On Wed, Sep 1, 2010 at 4:23 AM, Joseph Pearson
<jos...@inventivelabs.com.au> wrote:
> Short answer:
>
>    <meta name="bookid" content="xyz">

I see. I should have looked into Peregrin... I blame my lack of sleep.

> Long answer:
>
> The unique book identifier is an EPUB concept that has no analogy in
> Zhook. I'm not really a fan of the idea — it seems like half a
> solution to a not particularly significant problem. For our project I
> just use file hashes to identify identical ebook versions, and
> internal database identifiers to link ebook versions to a
> "publication".

We agree it's not a particularly significant problem. We also use file
hashes. An ID field could come in handy, however, in the case of
having multiple versions/publications if you need to identify them as
siblings (don't know how you go about that with your internal IDs).

> So it really comes down to how your Zhook->EPUB conversion tool
> obtains a unique id. Peregrin uses the "bookid" metadatum and
> generates a 12-char random string if it's not found. Peregrin's
> predecessor, chook (which is still active on publisher.ochook.org for
> now), used the 'identifier' metadatum, else the 'isbn', else a URL-ish
> construction based on the internal identifier for the upload.
>
> I'm happy to take patches for Peregrin's behaviour — probably it
> should fall back to ISBN like chook did, but I haven't given it enough
> thought.

I might fix that. Not a priority for now.

Francisco

[Joseph Pearson]

On 01/09/2010, at 6:45 PM, Francisco Treacy <fran...@widescript.com> wrote:

internal database identifiers to link ebook versions to a

"publication".

We agree it's not a particularly significant problem. We also use file
hashes. An ID field could come in handy, however, in the case of
having multiple versions/publications if you need to identify them as
siblings (don't know how you go about that with your internal IDs).

Yep, I know what you mean. More or less we leave it up to the user to update a publication with a different version of the book. But that may just be an artefact of our own design/architecture.

Happy to consider a small, obvious set of conventional metadata, among which we'd find a unique id. This kinda quickly gets caught up in weighty Dublin Core distractions, I think. Still, some solid ground would be useful to book designers. Is it reasonable to have a simple list like Title/Creator/Identifier/...?

- J

[Francisco Treacy]

On Wed, Sep 1, 2010 at 12:20 PM, Joseph Pearson
<jos...@inventivelabs.com.au> wrote:
> Happy to consider a small, obvious set of conventional metadata, among which
> we'd find a unique id. This kinda quickly gets caught up in weighty Dublin
> Core distractions, I think. Still, some solid ground would be useful to book
> designers. Is it reasonable to have a simple list like
> Title/Creator/Identifier/...?

It is, but it should remain as simple as possible. I wouldn't require
any of those meta fields, but suggest authors including a title,
creator and identifier:

<meta name="title" content="The Elements of Style">
<meta name="creator" content="William Strunk Jr.">
<meta name="id" content="urn:uuid:6f82990c-9394-11df-920d-001cc0a62c0b">
<meta name="description" content="The Elements of Style is an
American English writing style guide. It is one of the most
influential and best-known prescriptive treatments of English grammar
and usage in the United States. It originally detailed eight
elementary rules of usage, ten elementary principles of composition,
and &quot;a few matters of form&quot; as well as a list of commonly
misused words and expressions. Updated editions of the paperback book
are often required reading for American high school and college
composition classes.">

among perhaps a handful of metadata that are common and make sense for
*all* publications. In the end, it's just a recommendation. You can
add whatever fields you like to your meta tags.

If authors don't provide all or part of them, it's up to the RS to
handle that case.

Francisco