ALEG
Weekly Report - Week 22, 28 September 2001
What I've done
- Some queries started taking noticably longer than they should. After reviewing
these queries in detail, I changed the SQL query generator to generate simplier
queries (same information, but better structured) which the Oracle SQL query
optimizer interpreted more efficiently, greatly speeding up many queries.
A few other queries used in the maintenance dialogs to search for reviews and
manifestations were also optimized, so hopefully it is now possible to search
for common journals and untitled reviews (as long as some other information
such as date or author is provided) and get a fast response.
- Added holdings group of attributes to an agent (for archival holdings).
- Added a website work template.
- Added a note to the source ("part manifestation") group of attributes.
- Many changes to agent and work summary and complete display stylesheets.
These are not yet finalised, but are getting there.
- Many changes to the home page and static page stylesheets as the design from
the graphic designers is fine-tuned.
- Mapping of thesaurus terms is almost completed. I mapped the
"Aboriginal tribes & groups" subjects to new thesaurs terms, created agents for
all the Literary Festivals and Conferences (about 90 new agents with about
220 references), and mapped many of the award-as-subject references. Everyone
else has been busy mapping and there are now no unmapped terms and only
132 pendings (of which about 71 are awards). We are hoping to "push the button"
on the new thesaurus early next week.
- The AustLit machine was getting an unexpected number of collisions and errors on its connection
to the local switch. The network card was also operating in a very slow
mode (10 megabits/second, half duplex), although the cable and switch to which it was
connected were capable of 100 megabits per second, full duplex. Geoff Colin
from ITSC reconfigured the switch to force 100 megabits/full duplex but this
just made things worse, and a reboot didn't help. However, once the switch
was further configured to negotiate its settings with the AustLit machine's network
interface, everything worked fine and the link has been error free since.
The whole problem seems to have been caused by the switch being originally configured to force a
10 megabits/second connection because of a concern that the general cabling in the Library could not accomodate
the faster 100 megabits/second. But as the AustLit computer is physically adjacent to the
switch, the quality of cabling around the Library is not an issue for the AustLit server.
- Server patches were applied Friday afternoon to bring the AustLit machine
up to Sun's recommended patch level. A temporary workaround to a boot-time problem
which reduced performance has been removed.
Next Week
- Implement the searching of collection subsets of the database.
- More work on the static pages content and styling.
- Move to the new thesaurus.
- Multiple creation events for a work as a mechanism for allowing date ranges
to be associated with agents responsible for works, eg editors of a periodical.
- More result formatting from simple and guided searches.
Next few weeks
- First known dates (expression level).
- Advanced search screen design.
- William noticed that some of the place-of-publication data for
loaded records is wrong where the name of the place of publication (town or
city) occurs in more than one state (eg Richmond, Glebe). I'll investigate
when we map the spatial thesaurus. Dan, Chris and Terry have also noticed
similar incidents, so I think a very careful look at all place assignments
to places occuring in multiple states/countries is warranted.
Link of the week