On Fri, Sep 4, 2009 at 7:21 PM, Lars Aronsson<lars(a)aronsson.se> wrote:
...
For example, a new project might download the database dump from
OpenLibrary and start to weed out the "junk records". But that
junk sorting could also take place inside OpenLibrary. Why not?
Because metadata without digital objects are next to useless. Worldcat
already provides a directory of where physical books are held.
A database of metadata with lots of duplicates and no means for the
reader to fix them, and discuss them, is disrespectful.
If a new project goes to a library to ask for a copy
of their
catalog, they might get the question "we already gave (or didn't
give) that to OpenLibrary, so how is your project any different?"
And what should the new project answer to that?
See above. I dont see any value in going back to the libraries.
Doing that would only end up with the same result that OpenLibrary
has; it would be simpler to take the OpenLibrary dump.
I want to encourage wikipedians and wikisourcerers to
join the
OpenLibrary project, just like you should also join OpenStreetMap
and other good projects for free knowledge and information. Bring
your experience. If you get tired of one project, as I do
sometimes, work on another one for a while.
Tell me _one_ thing that I can do at OpenLibrary that I can not do at
Wikisource.
OpenLibrary has author pages for 6.5 million author
names. Some of
these are "junk" duplicates that should be merged, but still there
are quite a large number of authors there. These have a field for
a Wikipedia URL, but only 1100 records have a value. Connecting
author pages in OpenLibrary to Wikipedia biographies is just one
way where we can do a lot, without needing to start a new project.
_Most_ of them are duplicates.
http://openlibrary.org/search?q=Jules+Gabriel+Verne
I have an account at OpenLibrary, and I am responsible for 0.2% of the
Wikipedia links :P
I am not keen on becoming attached to a project that is littered with
so much crap, especially when I am not given the tools required to fix
the crap, nor do I have any say in whether more crap can be imported.
http://openlibrary.org/user/jayvdb
These two need to be merged.
http://openlibrary.org/a/OL2296708A/Charles-C.-Nott
http://openlibrary.org/a/OL2544127A/Charles-Cooper-Nott
Both of them look terrible, because I have no control over the
presentation of the pages. Dups, lack of sorting, etc.
I haven't found the OpenLibrary coolaid; I'll stick with Wikisource,
for good or ill.
--
John Vandenberg