On 29-Dec-2010, at 10:27 PM, Sarah Breau wrote:
> The way things stand now, there seems to be only one way to do a clean author
> merge. Before merging several author records into one, you need to go
> through all the works and point them to the same author record first. The
> merge will work properly only when it is done with one author record with
> many works and all the other author records with 0 works. When this is done,
> only one author record will be returned for every discrete author when a
> search is requested.
>
> If, however, there are still works spread across different author records
> when the merge is done, then all the other author names linger, and the merge
> is not clean. For example, if you search on the author name "Hasluck" you
> will see that the author Frederick William Hasluck shows up 11 times.
> However, since someone has done a merge on this author WITHOUT putting all
> the works into the master author record first, there is no way to move those
> works around now. There is no way for me to move all the works to the master
> record for Frederick William Hasluck, which is OL1435705A, which would
> eliminate all the other author names. So if you do a search on Hasluck, you
> still get 11 results for Frederick William after the merge, even though there
> is just one author with that name. Further, anyone entering a new work by
> Frederick William Hasluck will have to sort through those 11 author records
> and try to figure out which one is the correct one (which is kind of a joke,
> since they are all "merged" and ultimately point to the same author record).
Author merge takes care of updating the author filed of the effected works.
Usually the search engine takes a while to catch up with these updates.
In this case, it looks like the search engine didn't get these updates.
http://openlibrary.org/search/authors?q=Frederick+William+Hasluck
Thanks for reporting it. We'll look into this issue.
> Can anyone think of a way to split out the authors so the merge can be
> re-done cleanly? I thought about going through all the works that aren't
> associated with the master author record and maybe assigning them somewhere
> else (i.e. a dummy author record) and then assigning them back to the master
> author record for Frederick William Hasluck. That be a feasible workaround.
> The problem is, there is no way of telling *which* works are associated with
> the master record and which are associated with the phantom author records,
> so I would have to do it for all 14 works...and I am hesitant to put that
> much work into this author without knowing if it will work out in the end.
We don't have a way to split the authors yet. But if you notice a bad merge,
please let us know. We have an admin tool to undo the merge.
Anand
_______________________________________________
Ol-discuss mailing list
[email protected]
http://mail.archive.org/cgi-bin/mailman/listinfo/ol-discuss
To unsubscribe from this mailing list, send email to
[email protected]
