Unit 12 (or unit 11, part 2)

I think the time involved in the standard install has at this point become an extra two hours or less in the overall unit exercises. When considering this exercise in light of the fact that a repository will be around for a while (theoretically), I can’t see how it is that much more time consuming than something ‘off the shelf’ in the short run. In the long run, I’m guessing it’s better to build your repository from scratch as we’ve been doing. Because if you know how something is put together, you’re far more likely to know where to look when it breaks. More than a few times when I was unable to utilize Fugu for transferring files, I had to instead utilize the command line. This was a great learning experience, and I understood the server better after having gone into the files. Once, I could not see an xml file that I had uploaded, and I was able to see the entire XML file when using the command line at the terminal. I was able to practice some of the commands to get inside a particular directory, and this was really helpful.

I have far more confidence in my abilities than I once did, but I know that I have a lot more to learn, and that some of this will be learned on the job. I'm already concerned that next semester I won’t have many opportunities to go into the command line, and I guess that I'm already somewhat wistful for this dark screen:) I do think it would have been better if 675 was right before my capstone project, because I know that in some way I would like to be involved in repository configuration. One thing I’m curious about, do you have the same ability to create snapshots outside of a VM. I’m guessing this would be the typical server backups that most institutions have. Ultimately being able to take snapshots gave me much more confidence through both 672 & 675.

Unit 11 - Omeka

I have enjoyed the end product of all the repositories so far, however I would have to say that the user interface for Drupal is my least favorite. Dspace was at the top of my list until this week following the installation of Omeka. Omeka is by far the most superior in terms of the graphical interface, as well as in the flexibility of the configuration. Its presentation piece seems to suit my collection very well, especially since the collection consists of a variety of file types, including MOV, MP3, PDF & JPEG. I also appreciated the ease with which I could add metadata fields, as well as on the presentation side how easy it was to access and play my files, especially my film & music files.

I still appreciate Dspace and its communities and collections, however I believe that for my final project I will be utilizing Omeka. I appreciate JHOVE for all the technical metadata it revealed about my objects, and I liked the interface and searching capabilities of Eprints. I also loved harvesting Metadata by utilizing the base url and the OAI harvester.

To date this has been my most challenging semester, based in part on the advanced command line configuration we’ve had to do (and the subsequent troubleshooting that has had me pulling my hair out at times☺ , as well as the more challenging, analytical and critical management discussions that we’ve had this semester in evaluating larger federated repositories. I’m starting to see light at the end of the tunnel, so to speak, and I’m really starting to think seriously about my capstone project.

Unit 10 - an effective federated repository

I looked at the OAIster database sponsored by the University of Michigan digital library. I couldn’t get a sense of what the collections were. All the landing page states is that it provides access to millions of digital objects from thousands of contributors. It wasn’t like the Europeana federated repository, where the first thing you see on the landing page is a featured Art Noveau exhibit. This was a great way to be introduced to the site. I’m not too excited to search a federated repository that features nothing. Of course, if I know something already about this repository then I guess I wouldn’t need a polished looking landing page, but really, how many people will the site lose by not highlighting one or more of its partner collections?

I then visited the Sheet Music Consortium. The funny thing is that I entered a keyword search for Swan Songs. The first two records that came up for memorium songs to Bert Williams, an African American vaudeville performer who was active during the early 20th century. There were no results for swan songs in quotations.

I also looked at Nora (Noreweigian Open Resource Archive), and thought it was interesting how you could limit the topic by what appears to be a Mac file structure. All of the results were listed in Norweigian. I thought this was an interesting interface, especially the way in which the topics were divided according to the file structure.

The interface with the best design was by far the Sheet Music Consortium, but it was the only site that seemed to have made an attempt at creating a visually appealing landing page. This is where a federated repository such as Europeana is far and away a more appealing interface, regardless of Erway’s measured criticisms.