Ski report and the Welsh Names Database
Dec. 27th, 2008 05:58 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
hrjThe oldest and youngest brothers have flown back to California (since they have a play rehearsal for 12th night). I'll still be here in Maine until the 1st and have shifted home base from the middle brother's house (formerly Sibling Central) to the parents' house. The great skiing has passed -- yesterday I went out for a couple hours but the packed trails had turned to ice and the wind had blown all manner of organic material into the tracks. The best runs were across the open meadow where the grainy crust had warmed up just enough to give the right balance of tooth and drag. There was a small shower of new snow last night but nothing useable. The forecasts suggest there may be more snow Monday night and Tuesday, so I may get a couple more decent days in. Or, I suppose, I could drive somewhere that has better conditions, but I could do that back home. The novelty here is the whole "walk down the street to the ski trail" aspect.
Other than that, I've been messing around on the computer a lot. Mostly doing a bunch of really tedious coding in the Medieval Welsh Names database. When I get around to making my New Years' Irresolutions, I think one of them will be to get the names database online in interactive form. It will always be a work in progress -- and people who want to use it for researching and documenting names for the SCA are going to need some serious training in "you can't just say 'I found it on hrj's website'." When I started this project, I saw that I could go one of two directions. I could either make the data "safe" to use, with highly filtered and interpreted content and eliminating anything with any degree of uncertainty. Or I could present the material simply as a data resource, with all the interpretations flagged as to degree of confidence and rationale, and with all the background information about context dumped into the user's lap, and with the onus for responsible usage also dumped on the user. And I concluded the only sane path was the second one. The first approach would only reinforce the illusion that there are simple answers and absolute truths. The second approach reminds people that there is always uncertainty and interpretation and you just have to lump it and do your best.
The one nod I'll be making to user "safety" is that the planned report form for search results will have all the background data on the source and all the explanations of confidence levels included in the format. It'll be repetitive and redundant, but it will make it much harder for an end-user to take snippets of data out of context and claim that I said they mean things that I never said or meant.
Other than that, I've been messing around on the computer a lot. Mostly doing a bunch of really tedious coding in the Medieval Welsh Names database. When I get around to making my New Years' Irresolutions, I think one of them will be to get the names database online in interactive form. It will always be a work in progress -- and people who want to use it for researching and documenting names for the SCA are going to need some serious training in "you can't just say 'I found it on hrj's website'." When I started this project, I saw that I could go one of two directions. I could either make the data "safe" to use, with highly filtered and interpreted content and eliminating anything with any degree of uncertainty. Or I could present the material simply as a data resource, with all the interpretations flagged as to degree of confidence and rationale, and with all the background information about context dumped into the user's lap, and with the onus for responsible usage also dumped on the user. And I concluded the only sane path was the second one. The first approach would only reinforce the illusion that there are simple answers and absolute truths. The second approach reminds people that there is always uncertainty and interpretation and you just have to lump it and do your best.
The one nod I'll be making to user "safety" is that the planned report form for search results will have all the background data on the source and all the explanations of confidence levels included in the format. It'll be repetitive and redundant, but it will make it much harder for an end-user to take snippets of data out of context and claim that I said they mean things that I never said or meant.
no subject
Date: 2008-12-28 05:02 pm (UTC)no subject
Date: 2008-12-28 06:21 pm (UTC)The Welsh names database (like the surviving garments database) currently lives in FileMaker (then exports to something else for the web interface). Both of them started in Excel, which was dreadfully clunky but was useful for the initial stages. Actually the names database originally started in dBase, but when I moved to a Mac that wouldn't work, which was when I switched to Excel.
And then, just for fun, on my Day Job I'm developing a project-tracking database for my department in Access.
No, less coding is not a goal. :)
no subject
Date: 2008-12-28 09:28 pm (UTC)You have a very curious definition of fun!
It's ok, I put 3 years of a mine's ledger data in Access (not enough rows in Excel) and started using pivot tables in Excel to do data analysis. Amazing what those little tools can do!