Landscapes of Injustice

While I was in the UK, I rather hurriedly created at JSR's request a clone of the live titles database for duplicate-entry checking. I ended up doing it at 2am on the night I was leaving, because it was the first chance I got and it was urgent. Not the best time to be working, especially with jetlag and at the end of several long days of work. It turned out that the cloned db did not inherit a number of triggers which were essential to the functioning of the interface. SF reported that drop-down lists were empty of content, although they had items. I've now copied over the triggers on the owners, titles, lawyers and props tables, and I think everything is fixed.

The "loi" space on nfs is now available and working, and I've completed a set of detailed instructions for users, including screenshots, and distributed it.

With SA, wrote some CSS so the team can easily proof their work, and then tweaked the schema and encoding practices a little based on some of the peculiarities were were seeing.

With SA, went through the questions arising out of the street-index encoding task and tweaked the schema a bit to provide more guidance and more constraints.

With SA went through a list of queries from NH about the person list entries, and tweaked the schema where necessary to come up with encoding strategies for things not predicted before.

Finally got the last of the netlinks for project members sorted out, so I've sent the request in to sysadmin for the file storage, with the usual arrangement for home directories inside the project dir, and SetGID/2775. Rewrote an existing set of instructions from another project to suit our setup, and as soon as the new filespace is set up I'll create screenshots with Filezilla showing the correct login stuff and finish the document, then circulate it.

JC and I met today, and she showed me some of the GIS work. We talked about some principles for file/folder naming conventions and identifiers, and she's going to work these out in detail and circulate them so that when we eventually bring all the GIS databases together they'll make sense, and will mesh with our other data.

By the middle of next week, we hope to have the shared filespace set up so you can store your data centrally where it will be backed up. For that, I need UVic netlink ids from the cluster; still waiting for those.

JC and I agreed on a set of identifiers for the streets, so that we don't have to worry about "Street" versus "St." and upper/lowercasing etc.:

main
gore
dunlevy
jackson
princess
easthastings
eastcordova
powell
alexander
railway

We noted that the GIS folks are including Pender, whereas we haven't so far, because it doesn't appear on the Nihonmachi map we have, so I'm checking with JS-R about that. We also agreed that since there are some buildings with entrances on the alleys, which are unnamed, we should also have a convention for identifying the alleys; this is based on the initial letters of the streets between which they run, and their position E/W or N/S on the 1941 plan, so we have:

Running N/S:

alley_mg1 (between Main & alley_ap1, from Alexander to Powell)

alley_mg2 (between Main & alley_pc1, from Powell to East Cordova)

alley_mg3 (between Main & alley_ch1, from East Cordova to East Hastings) 

alley_jp1 (between Jackson & Princess, from Railway to Alexander)

Running W/E:

alley_ra4 (between Railway & Alexander, from alley_jp1 to Princess)

alley_ap1 (between Alexander & Powell, from alley_mg1 to Gore)

alley_ap2 (between Alexander & Powell, from Gore to Dunlevy)

alley_ap3 (between Alexander & Powell, from Dunlevy to Jackson)

alley_ap4 (between Alexander & Powell, from Jackson to Princess)

alley_pc1 (between Powell & East Cordova, from alley_mg2 to Gore)

alley_pc2 (between Powell & East Cordova, from Gore to Dunlevy)

[no alley_pc3: park intervenes]

alley_pc4 (between Powell & East Cordova, from Jackson to Princess)

alley_ch1 (between East Cordova & East Hastings, from alley_mg3 to Gore)

alley_ch2 (between East Cordova & East Hastings, from Gore to Dunlevy)

alley_ch3 (between East Cordova & East Hastings, from Dunlevy to Jackson)

alley_ch4 (between East Cordova & East Hastings, from Jackson to Princess) 

Did a quick intro to XML and TEI for the Directories folks and also people from the VPN project.

We now basically have working schemas for the documents we're encoding. The differing document types lead to differing encoding strategies, which is perhaps not ideal, but they will work pretty well. Spent a long time figuring out what the different components of the Japanese directory ads are, looking up kanji etc., to get some basic examples encoded. They use a lot of obscure or obsolete characters.

SA and I spent most of the day working on the ODD file for the Directories encoding project. We're using some of the new TEI ODD elements in preference to RelaxNG where we can; so far I've filed two minor bugs on these, and I have one difficult query on the Council list to see if I'm misunderstanding something. We're about three-quarters done, though, and it's going to be a very well-constrained schema.