ALEG

What I've done

• Semi frantic preparation for training, including development of a basic user maintenance facility and update logging and update browsing.

  We came across a strange problem on the Windows 95 machines in the training room on Monday evening (!) where after an update was made the client would throw up an error whilst closing the "Please Wait" dialog window (something about an RPC error). We've never encountered this before, and just worked around the error by trapping and ignoring it!

• Training (Tuesday, Wednesday, Thursday). Feedback during training threw up many changes which were implemented during the week (still a few to go!)

• Partner Meeting (Friday)

• Started "hints" document

• Add general edits based on XSLT Schema (the Schematron approach).
• Add topic selection to the work/expression/manifestation UI (Marie-Louise, Kerry and Annette have produced a document containing suggested changes for me which I'm working through).
• Revisit output formatting (especially to show all the newly added attributes and relationships) and search screens.

• Implement changes based on training feedback

• Enhance expression/manifestation selection search

• Promote all existing expression titles to work level, ditto for agents

• Apply Tessa's author and publisher series selections

• Start coding the program to merge in the BAL data

• Load the multicultural agent info

• Remove the test works and agents added during training

• Meeting and working with everyone at the training was wonderful!

  I presented the following list of activity for the next 6 weeks at the Partner meeting:

  Summary of development so far

  A noteworthy aspect has been that the 'obvious' risks, such as the use of the new and radically different FRBR model, topic maps, a 'super-normalised' database structure based on topics, property types and values and browser-based distributed maintenance haven't caused great concern. Instead, it was something we all greatly underestimated - the conversion of data for loosely structured/text based systems into a highly structured model, and the merging of data from different datasets.

  Major tasks to be undertaken in the next 6 weeks:

  1. data loading:
     1. cleanup of LAW data
     2. loading of BAL data
     3. loading of multicultural database author information
     4. loading of WA data
     5. loading of Lu Rees data
     6. cleanups of subjects, publishers, place names

  2. new thesaurus load and conversion
  3. clean-out of test data
  4. various topic reassignments - ethnicity/nationality, some work-type/forms
  5. various mass data fixes/cleanups (moving statement of responsibility, representing expression and manifestation titles and agents at the work level)
  6. link to Kinetica holdings
  7. tweaking of retrieval query forms and stylesheets
  8. application of graphical design templates
  9. maintenance of customers database
  10. production documentation: system and user including tips
  11. simple update review process
  12. database tuning
  13. setup of new machine
  14. Planning for stage 2

  ---

Links of the week

• Hints for Austlit Maintainers - we will try to generate a useful hint each week, accumulating them on this page.

• Think Once, Think Twice, Groupthink - What the Bay of Pigs and the Cuban Missile Crisis didn't have in common.

• Sense and Censor-Ability - John Perry Barlow