new features include addition of non-xref'd links in search results (now you can see _everything_ jake knows about), links for per-database title lists, better layout of search results including links and providre names, and a number of bugfixes. try it out at jake.med.yale.edu...

seen on many lists recently is XMLMARC, a "Java client/server program [which] converts MARC to XML based on flexible maps and simplified, yet detailed DTDs for bibliographic and authorities formats." Its license is free for non-commercial use but requires consent for redistribution, and their broader experiment is described here. They seem to be aware of MARC.pm... it would be great to see a side-by-side feature comparison of the two.

as seen at freshmeat: "Over 14,000 (65%) of serials cross-referenced now have Library of Congress subject headings, classifications, or Dewey classifications, and jake may be searched according to these, initial support for a complete data localization framework so libraries may experiment with customizing front-ends and results according to local holdings, and several bugfixes and UI tweaks, including easy-to-download database content lists." more at the jake site...

30 new databases (mostly from Ebsco) are in jake; see the list or grab the whole tarball at the all about jake page.

gnujake is the GNU Jointly Administered Knowledge Environment. It seeks to support the management of and linking between online resources and descriptions thereof.

There are now thousands of auto-correlated records in the union list. Try it out at www.med.yale.edu/library/gnujake/. Note that the authority editing piece is under revision for a few days...