Those present included representatives from Cornell, Princeton, Brandeis, Stanford, and NYU, as well as Ed Glazier and Karen Smith-Yoshimura from RLG.
A new version of the Authority Record Assistant will be installed in early February. It will fix some bugs, i.e. you will only be able to GENerate an authority record from an account enabled for NACO contributions; you will not be able to GENerate from vernacular script fields (when determining sequence number of a field, you should count both the vernacular and romanized). The aggregator in Chinese headings currently is transferred when generating a record, but will be removed in the future.
Subfield b with eng will be prompted in 040. STH will be changed to a on the template (since the record will be in save, a status of d for provisional is over-protective). RFE will remain blank on names and n on series. On the list of developments will be RFE of a if a reference is generated.
A 642 field will be generated from the $v in the series field on the bib record from which you are generating (requested change). 642 and 645 will get both $5 DPCC and $5 for your NUC code.
Terminal punctuation will be fixed in the next release, i.e. semi-colon will be removed from 1XX in series; before date in 670 $a; inside parenthesis in 670 $b.
UPN will be linked to the 100 field, i.e. a when 100 and n otherwise (requested change).
For series records, the 670 $b will come from the 4XX if the record is generated from 4XX and will be blank if the record is generated from 8XX (requested change). For the moment, name records will continue to have 245 $c prompted in 670 $b.
Currently, if you pass or put a record, the bib record will be passed/put unless you have moved to the authorities module and found your record. RLG is aware that this is not necessarily the desirable result but did not promise any quick fix.
It was suggested that a Reformat command would be helpful to redisplay the generated authority record before it is entered (since the active authorities window closes when you enter, it is not currently possible to enter and then redisplay the record with field rearrangement, etc.). If this change is approved, it will probably be a "dis" or "ful" command.
The Z39.50 capability will be tested soon -- beta testing will take place in February. You must still be using RLIN terminal software for Windows, in the CAT function. You will SEL FIL LOC (with "loc" being a surrogate for the local system of the logged-on LI for which the address, port, etc. will be stored by RLG). You will do a FIN ID search to retrieve a particular bib record. Once the record is retrieved and displayed, you will use the GEN command as you now do.
Another new process which is being discussed would allow the loading of authority records from local systems. The command would work via Z39.50 and might work like tra der (transfer derive) currently works in the bib file. The capability would be on a one-record-at-a-time basis, though batch loading is also being discussed. LC has more intraedit requirements than RLG and a record can be rejected even if it passes RLG edits.
Technical requirements will be checked first at NYU and Stanford. The University of Oxford has also expressed an interest already in the Z39.50 transfer. All NACO participants (including funnelisti) will soon be queried about technical capabilities and interest in Z39.50 transfer.
Notes by Sherman Clarke, NYU
with input from Ed Glazier at RLG