On Wed, Dec 08, 2010 at 10:07:46AM -0500, Soulliere, Robert wrote: > Some documentation areas of mystery or areas needing enhancement:
<snip> > Limits of vandelay (staff client importer): I was wondering if there are any > general guidelines we can provide for how many records vandelay can import at > one time to help users decide the size of MARC files to divide there records > into or to decide when to use optional external scripted approaches. I > understand that some of the vadelay limits are dictated by system capacity > etc, but general guidelines would be great and require feedback from > developers or very experienced users of the staff client importer. I've been using a rule of thumb of about 5,000 records per batch for Vandelay imports. > Importing Authority Records: Again, providing guidelines for Vandelay would > be nice here. Also, an up-to-date best method using an external approach > would be great. As part of my effort to write up the changed authority support in 2.0 (http://evergreen-ils.org/dokuwiki/doku.php?id=authorities:2.0), I've recently added some documentation on how to import authority records using both the command line and MARC Batch Import/Export at http://evergreen-ils.org/dokuwiki/doku.php?id=authorities:2.0:importing_authority_records Note that for 1.6.x the command-line approach will need direct_ingest.pl inserted between marc2are.pl and pg_loader.pl. Otherwise, I don't think things have changed much for either approach. > Adding languages in 1.6.1: To encourage international growth for Evergreen, > we need a step by step procedure for adding a new language which do not > currently exist and define which languages are supportable (e.g. > right-to-left issues language issues). Hmm, how much do you want to document here - the current Launchpad translation approach? I think we should have translators put their heads and agree to some sort of process before we try to write something up, because otherwise it will be fiction. Right now the process is basically: * Get an account on Launchpad (http://launchpad.net) * Register your account for the languages you're interested in translating * Go to the Evergreen translations page (https://translations.launchpad.net/evergreen) * Select the language you want to start translating and begin supplying translations for the terms (permissions are open, so anyone can start translating any language whenever they want) For every release tarball: * Release maintainer generates the translated files from the supplied POT files And somewhere in time at some interval (not really very well defined yet): * i18n maintainer takes translations from Launchpad translation export branch and updates or adds them in Evergreen >From there, http://evergreen-ils.org/dokuwiki/doku.php?id=server_installation:adding_localizations covers what needs to happen to add a set of translations to the database. I'm not sure the manual is the right place to document a team process like this, especially as it is likely to evolve as more people join and need to start collaborating (for example, multiple people collaborating on the Spanish translation). Where does the DIG and the Web site intersect? > Adding formats in SuperCat in 1.6.1: Dan Scott wrote something up in his > Coffee Code blog in January 2009: > http://coffeecode.net/archives/181-Adding-a-new-metadata-format-to-Evergreen-in-a-dozen-lines-of-code.html > Is this still applicable in 1.6.1? I suspect the best thing to do on this item would be for someone to try reproducing these steps as documented with a slightly different metadata format and find out if something breaks; then we can move on from there. _______________________________________________ OPEN-ILS-DOCUMENTATION mailing list [email protected] http://list.georgialibraries.org/mailman/listinfo/open-ils-documentation
