I will be presenting at a couple of conferences in Germany later this month, and I need to get my stuff together. So I will be dedicating the time that I had been giving to this project, to those presentations. I will be back at the end of July.
The subject line says it all.
I’ve uploaded new versions for all the BookReaders, and also updated the index to include all the manuscripts in the collection. (Follow the “Walters BookReader” link in the top menu to find it) The BR are greatly improved, however there is a pretty major bug: all the 1-up images display very, very small. Doug is working on methods to fix this, but for now I would just recommend that you use only the 2-up or thumbnail views.
Also this morning, I uploaded a couple of thousand decoration images into Omeka. Some of them are now tagged (with the siglum of the ms in which they are found) and public. (Follow the “Walters Omeka” link in the top menu) This is taking much longer than anticipated, as Omeka seems unable to accept a CSV file with more than about 90-100 rows. Considering the thousands of image I’m dealing with, loading only 100 or so at a time means a lot of effort and time. I’m going to check out the forums to see if there is a way around this, but for now… enjoy!
I spent this evening finalizing the XSLT to convert the msDesc into CSV for import into Omeka. To the fields I mapped yesterday I’m adding the manuscript siglum (so we can easily find which decorations are from which manuscript), as well as the folio number on which the decorations appear (this was Doug’s suggestion; I’m hoping this will make it possible to create some kind of automatic link between the Omeka records and the corresponding folios within the context of the BookReader).
I’m generating the CSV files now. I had really hoped to be able to process everything into one huge CSV file, but I wasn’t able to get it to work (and really, that would be one huge file) so instead I’m generating one CSV file for each manuscript. There is some post-processing to be done, to keep things from getting too messy I actually put XML tags around each row of data, and those will need to be stripped out. I may see about combining some of the CSV files together, so I won’t have to strip quite as many separate files, but 200 together may be too many. We’ll see. I have a holiday tomorrow so hopefully will have some time to work on this, between cooking and wrangling the toddler.
The second part of my project, after getting the manuscripts all loaded into the Internet Archive BookReader, is to build a more extensive catalog for the manuscripts in Omeka. Eventually I’m going to experiment with some scholarly tools as well (I’m particularly interested in Scripto, which enables crowdsourced transcription, and the just released today Neatline, which supports temporal and geographic research) but for now I’m most interested in getting descriptive metadata out of the manuscript descriptions and into Omeka, where they can be searched and explored.
Tonight I generated a csv file from one of the Walters manuscripts (using XSLT), and then used the Omeka CSV plugin to import that data. I wasn’t really careful about mapping the fields (I’m using extended Dublin Core in Omeka), I’ll probably go back and take another look to make sure I’m mapping to the best DC fields. For now, I’m most interested in making sure the workflow is effective. So far, it’s great.
I’m using another plugin in Omeka that allows for hierarchical collections, so I’ve created one main collection, Digital Walters, and currently one subcollection, for manuscripts that are described according to their decoration rather than their textual divisions. I will create a second subcollection for those described according to textual divisions. I expect there are some (probably several) that have both extensive decoNote sections and msContents sections… I’ll deal with that when I get to it (ah, one benefit of experimenting, I don’t have to have all the answers before I start!).
For now, however, enjoy!