HCMC Blogs

Here's my summary of the major points coming out of the meeting on Friday 25 Sept (PD, JL, MH, GN, SA):

1) modify 1891 dataset
notes:
- add 500 records
- modify occupation in a couple of hundred records
to do:
- SA to check if the postgres is one large table or is one table
per "census"
- SA and PD to see what, if any, mapping there is between the
record ids in the access dbs and the record ids in the postres db(s)
- PD to provide data strcutured as ageed
- SA or GN to write the GREP, SQL queries etc. to execute
necessary modifications

2) include "check" census (1892?)
notes:
- field structure similar that of business directory
- searchable on its own
- include in global search
to do:
- PD to confirm list of fields and type of data in each field
- SA to research documentation on how to add a new dataset

3) add 1911 dataset
notes:
- field structure similar to that of 1891 census
- searchable on its own
- include in global search
to do:
- SA to research documentation on how to add a new dataset (new tables and/or new fields to existing table)

4) make it easier for users to find the annotation feature
notes:
to do:
- SA to check structure of data to see which dbs include annotation field, and thus where to put GUI element linking to submission form

5) add ability to send street address to related site (?in Kingston or London) and get back KML file which can be used to locate address on a map
notes:
- any implementation work at this end only after the interface is set
to do:
- PD to talk to people in Ontario about exposing a programming
interface to the GIS data that we can query
- we will consult on the specification for that interface

Tasks #1 and #4 are probably small enough that we can work on them in "on-demand" hours this term, likely completing them by December 2009.
Tasks #2 and #3 are large enough to require formal project status and scheduling. Implementation work would be scheduled beginning January 2010 at earliest.
Task #5 any further implementation work is dependent on others providing a public interface and technical documentation. If it proceeds, it will be a small, formal project.

Link: http://www.lexilogos.com/francais_langue_dictionnaires.htm

The "lexilogos" site is very cool -- many dictionaries in French (including a link to the Furetière on Gallica. It really is the most interesting 17th-c. dictionary).

As for the question of repeated references to the same word, I think it is a good idea, given that readers may enter the text at different points as the result of a search. But (Martin) can this be done automatically? I take it Emily was inserting each reference manually.

Toujours des points intéressants!

Churned through the bibliography. One more article to go...

An extremely helpful user has been sending regular error reports, which we're working through steadily. Made some changes to Verbs 2D this morning; there are more in the queue, awaiting OK from LB.

At CC's request, changed "Complet" to "Tous les documents".

Racing a deadline on the Adaptive DB project...

The XSLT code which allowed access to 1858 and 1846, while blocking the other years with an explanatory message, was rather ad-hoc, and didn't easily lend itself to extension as we add new years, so I've turned it all into a neat variable that has a sequence of xs:string values for the years allowed. The error message is generated from the same variable, so now "publishing" a year is just a question of adding it to the sequence.

Looking back through our correspondence, I realized that JW had mentioned in passing that one of the custom fields she wanted was a lookup field (a string field which has a drop-down of all previous distinct values applied to that field as a prompt). This wasn't part of my original spec, so it's an unexpected piece of extra work; ended up pushing into the evening in an effort to get it done, but it's actually quite complicated because the SQL that gathers the previous values is more complex than that for a normal lookup field. I've managed to get it half-done. When editing an existing record, the previous values show up, and work, but when editing a new field, or using the search, they don't get populated, and I can't figure out why so far. I can put one more day into this, but after that it'll have to be shelved, I think. Lots of interruptions from other projects and work coming in have made it very difficult to concentrate today, so I didn't get far until everyone had gone home.

DR wrote with a new program for the 2009 Colloquium, and some related updates.

Significant changes (mainly simplifications, linking to the main UVic calendar in place of enumerating info on the site itself) done at the request of DR.