- From: LAURA DAWSON <ljndawson@gmail.com>
- Date: Tue, 16 Sep 2014 07:24:19 -0400
- To: Liam R E Quin <liam@w3.org>
- CC: "Siegman, Tzviya - Hoboken" <tsiegman@wiley.com>, Ivan Herman <ivan@w3.org>, W3C Digital Publishing IG <public-digipub-ig@w3.org>, Bill Kasdorf <bkasdorf@apexcovantage.com>, Madi Weland Solomon <madi.solomon@pearson.com>
Yes, it¡¯s odd that ebook systems can¡¯t seem to ignore the leading article when the retailers who develop these systems mastered that some 14 years ago. Indicates a bifurcation in product management teams. These are all terrific suggestions; and I agree that metadata is one of those topics that takes two forms: the simplistically reduced and the deeply complex. Somehow we¡¯ll have to muddle through. :) On 9/15/14, 7:15 PM, "Liam R E Quin" <liam@w3.org> wrote: >On Mon, 15 Sep 2014 18:11:39 -0400 >LAURA DAWSON <ljndawson@gmail.com> wrote: > >> Yes, and I wouldn©öt expect the retail sites to change that. How they >> ingest data and express it on their sites is at the core of the value >>each >> retailer brings to the table > >+1 > >> Page count is another one of those troublesome fields. :) > >I have my trusty copy of McKerrow on hand for bibliography and citing >collations :-) > >But you are right. > >Clearly metadata for ebooks (and for Web sites that may be used offline) >has several properties that are needed... some likely examples: > >* mixture of embeddable metadata (title, author, category/facets), > updatable metadata (price, latest edition...) and > pointers to remotely updatable metadata (BCIP, OCLC, Amazon category >names). > >* Retraction (e.g. ability for publisher to correct errors, or when a >category > is split so that Shelf Zero now has occult and computing separately >instead of > intermixed (real example!) > >* Marking of every item with source and date - e.g., according to Allen & >Unwin > colume 3 of Lord of the Rings goes in Adult Fiction, as of >such-and-such a date; > if you extract single "triples" and use them out of context you're >asking for > a mess. > >* Handling complex "fields" - e.g. book or journal titles in mathematics > often contain formul©¡; in the humanities you'll get titles with >fragments from multiple languages (Nielsen and other metadata >organizations notwithstanding). The "CDATA excaping" mechanism for this >is lunacy. > >* User-supplied metadata (e.g. "this book is really about computers, not >the occult, > and that's where I want it on my virtual shelf" or "sort this book >under "R" for > "Really Hard", not under "M" for "McKerrow"... although the ebook >systems I've > seen have enough trouble ignoring "The" when sorting titles...) > >Onyx seems to be (1) complex enough to handle these cases, with work in >some areas perhaps, and (2) complex enough to make a lot of people run >screaming. But then, try going to the middle of your friendly reference >library and saying, "can anyone help me, I want to have a friendly chat >about MARC records" :-) and one has to remember that, just as XML tends >to be used at the boundary between the "world" and the computer, metadata >is used by people who are metadata experts but not necessarily computer >experts. > > >-- >Liam Quin - XML Activity Lead, W3C, http://www.w3.org/People/Quin/ >Pictures from old books: http://fromoldbooks.org/
Received on Tuesday, 16 September 2014 11:25:00 UTC