Last month a Memorandum Against RDA Test was posted to the web and a call for signatures was put out on the cataloging discussion lists. As of this writing there are 298 signatures.
So, what's the complaint? There is concern that the RDA test records being input into the OCLC WorldCat database are going to cause problems for local libraries' cataloging workflows and systems.
There is particular concern about the authority records and the new headings created using RDA as the cataloging code instead of AACR2. People are concerned that the new headings being used in the test bibliographic records (when there is already an established AACR2 form) will cause a lot of conflict and, if RDA is implemented, a lot of authority changes will need to happen in local systems.
Since this is only a test that's adding a limited dataset in WorldCat, it does not seems like an insurmountable problem in the short run. If we're going to be practical when implementing RDA, however, we should just pick a date, stop using AACR2 rules to establish headings, and start using RDA rules for anything new. The problem the test has revealed (and why catalogers are so upset) is that it makes no sense to use new headings when, they argue, there's a perfectly good heading already established.