There has been a good deal of discussion about my July 17 draft of the
Resource Type element, and in keeping with my weekly review schedule I've
considered all the messages on this topic and am ready to respond and
redraft. Please excuse the length of this message, but I thought it would
be best for you to receive one long one than a number of short ones.
* John Kunze raises an excellent point about functionality that I don't
wish to ignore while addressing his specific suggestion. That is, we need
to think carefully about WHY we are doing this, and not allow ourselves
to be caught up in the excercise of trying to slot every little thing we
can envision coming down the pike. What types are essential to
information discovery and retrieval? Or perhaps more specifically, when
and how will users want to limit their searches to items of particular
TYPES?
Nonetheless I disagree with John's suggestion to begin with "online,"
"interactive," and "offline" as the first level (frankly the idea of
using DC to describe offline resources is new to me). It seems almost too
high-level to be useful, since I think in reality nearly everything
described by DC records will be online, thus rendering online
("fetchable") vs. interactive the only distinction. Given that, I think
the existing categories largely imply either fetchability or
interactivity by their nature. But what do others think?
* Arthur Chapman suggests that Environment is not the appropriate term. I
can understand that, coming from Environment Australia and the particular
usage of the term in that context. But the first definition for
environment in the dictionary I have at hand is "something that
surrounds; surroundings" which is exactly the meaning of what we are
describing. That is, "interactive" doesn't cut it with me, since many
things can be interactive without being encompassing in the way that a
game or virtual reality environment can be. I do not want to put forms
(simply "interactive") in the same class as virtual reality (an
"environment").
* "Collection" has been suggested. I am rather at a loss to determine
what I think of this, since either "text.collection" or
"collection.image" makes sense; that is, it can either be a top-level
category (that then one could presumably qualify with text, image, sound,
etc.) or a qualifier for the current top-level categories. I think it
would be better to use multiple resource types to describe the elements
present in a collection, and use the DESCRIPTION element to describe it
as a collection.
* Andrew Chapman suggested a subtype for Data of Structured-Text, to
include database reports and bibliographic data, which I implemented.
* Jordan Reiter suggested a hypermedia type to be added to Environment. I
am suggesting multimedia as being more appropriate in that category (that
is, hypermedia can be strictly text and therefore is probably more
appropriate as text.x-hypertext, whereas multimedia objects can be hyper
and still be classed as multimedia)
* Debbie Campbell pointed out that proceedings can be either monograph or
serial, thus I moved it out from under serial.
* A comment from Jon Knight prompted me to change Media to Promotion and
add a description with examples of appropriate instances.
* On the top-level page (http://sunsite.berkeley.edu/Metadata/types.html)
I added a link to Andrew Daviel's summary list of all proposals.
Thanks to everyone for their great feedback and ideas. If you don't see
your favorite one here, keep at it. I do try to reflect the will of the
group as much as I'm able to determine what it is. Thanks,
Roy
|