On 21/02/2012 18:46, Kevin M Randall wrote:
<snip>James Weinheimer wrote:These points demonstrate how far apart we are. Of course, we are building a product for use of the public. We are not disinterested people building some abstract tool that can be taken or left. At least everybody claims that our work is supposed to be useful for the public, and not only for us. It only makes sense to make sure instead of taking it for granted. Following your analogy, are we using space age, modern materials or thatch with wattle and daub? I don't think I would want wattle and daub. Or are we trying to build a cave?
I think that we can probably agree that if a company builds a product the public does not want, it will be exceedingly difficult to get anybody to buy that product. Therefore, the task for that company would be to convince the public to buy something it does not want. Yet, an outsider will reasonably ask, "Why not build something that people want?" while an investor would probably not want to invest in that company.FRBR and RDA are not about building a product that the user is going to be encountering. They are about building the DATA. Entities such as OCLC, Ex Libris, Innovative Interfaces, SirsiDynix (to name *only* a few by way of example) are the ones building products that the user encounters. Those products use the DATA that FRBR and RDA are concerned with.
A very rough analogy might be building construction; the analogues of FRBR and RDA would be material specifications and manufacture, and the analogues of the public catalog would be the actual buildings. In general, most people will only rarely give much thought, if any, to the materials that were used in constructing a building that they are using. But the people who are actually designing and constructing the building care very much about the materials, since the specifications are critical.
I don't understand why it is so difficult to realize this.
If you are discussing something else, that is, talking about opening up our data to general development, now that is another argument, and one that I can readily agree with. But we can open up our data now without going through the expense, hassle and *waiting* for everything to be changed to the FRBR entities. We can provide MODS or even simple Dublin Core. Lots of bibliographic agencies are doing this now. This would be following Tim Berners-Lee's suggestion: put out your data in a useful format and link what can be linked. Others will pick up the pieces and transform your data for their own purposes.
If we did this, we could learn so much about our user community. Who knows? We might find out that people actually do want FRBR type entities. I would be surprised but that would be fine--at least then we would know that it would be worth the expense. Or we might find out something entirely new.