So I’ve reached a stalemate at this point. Does anyone out there have experience with making Drupal play nice with TIFFs?

There are still lots of technical metadata specs for different formats to deal with but UH really hasn’t got to far into media, YET. I’m making some headway with my prototypes and looking forward to handing stuff over to my staff in December. In the meantime there is quite a bit of reading left to do.

Some key resources I’ve been using to educate myself and make decisions.

MODS

• Library of Congress MODS site
• LOC MODS Implementation Registry – sites using MODS
• MODS in action – Open Search of Georgia PINES catalog for “pottery”
• MODS Editor at Brown University – helpful for getting an idea what a MODS editor interface can look like

METS

• METS website at Library of Congress – examples here that are useful to see what METS looks like
• METS Implementation Registry – see who else is using METS and for what
• The Australian METS Profile – A Journey about Metadata – an article at D-Lib which discusses implementing METS at the National Library of Australia. Really good diagram

The best thing about this page is the readable error messages and suggestions on what to do to fix things. I also like that it is segmented so that you can see which pieces of the puzzle are working and which aren’t. Me I’ve got a few problems to fix.

1. Something is wrong with JHove
2. I’ve got some issues with where Fez thinks PHP lives
3. Fez isn’t connecting to Fedora properly

The great thing about this page is that even though stuff isn’t working, I know where to start looking to solve the problems. Stuff like this makes all the difference when you are trying to install and debug a new open source application. So big time kudos to the Fez folks for adding this little bit in.

If We Build It Will they Come? – Phillip Davis and Matthew Connolly

Examine usage of Dspace at Cornell and look at why or why not faculty were using it.

Upward trend of growth at Cornell. About 1000 items per year.

To get usage data, looked at server log files. Also used web crawler to look at numbers of items added.

Most collections are slightly populated in Cornell’s DSpace

Only a handful have more than 50-100 items.

What’s active

• Thesis and dissertations
• Senior seminars
• Presentations by Cornell Librarians
• Image collections
• Reports by Librarians at Cornell
• Videos being used by classes

Growth patterns of collections

• Great Plains

• Conference proceedings
• image collections
• archived newsletters

• Senior seminars
• cooperative extension publications

• Thesis and dissertations
• articles

Compared growth patterns of collections at Cornell to other schools with DSpace

Most school’s collections are in the Great Plains or Terraced Gardens model of growth.

What Faculty said?

• Access to literature is a non-issue for faculty
• Using alternatives to institutional repositories

• Have their own websites and like using them better
• Subject based repositories (prefered over IR if available)

• faculty move from school to school and want their stuff all in one place (why subject repositories appeal)

• culture does encourage use of IR
• no awareness of IR

Reasons for using/not using repositories

For

• Permanence
• Timeliness
• Registration of discoveries
• Policies of funding agency

Against

• Redundancy
• Learning curve
• Copyright
• Quality association
• Plagiarism and being scooped
• Reputation and accuracy
• Publishing “original” work

Publication forthcoming in March 2007 issue of D-Lib Magazine – (http://hdl.handle.net/1813/5195)

DSpace: Manakin Themes and Applications – Alexsei Maslov

This interesting session discussed the architecture behind Manakin Themes. Manakin is a project of the Digital Initiatives group at Texas A&M, Manakin is an XML UI for DSpace that uses SAX & the Cocoon framework to enable communities and collections to establish a unique look and feel that is distinct from the default installation of DSpace.

Manakin themes are made up of three components: the sitemap, the XSL, and the CSS. The sitemap is the configuration portion of the theme. It references the components the theme needs. The XSL converts the DRI (Digital Repository Interface) schema formatted version of the page to HTML. Lastly, the CSS file styles the output for presentation.

The nice thing about the architecture of Manakin is that it does a nice job of separating presentation from content. It also takes this one step further and let’s you change the information being presented by changing the XSL. There is a simple tutorial on creating themes at the Manakin website as well as a Developer’s Guide which has more extensive information.

Honestly, I can’t do the presentation on Manakin themes justice. This is a very powerful add-on for DSpace and definitely worth checking out if you want to run a single DSpace instance for multiple institutions. I’m really interested in spending some more time looking at Manakin but right now I’ve got other higher priority projects on my plate.

A Manakin Case Study: Visually geospatial metadata & complexity – Adam Mikal and Scott Phillips

This presentation followed up on the presentation about Manakin themes. It discussed a project at Texas A&M which used Manakin to change the user interface for a collection of maps, texts and photographs from the Geologic Atlas of the United States. The problem with the current interface was that couldn’t effectively deal with the complexity of the items in the collection (the original interface is optimized for items with fewer bit streams), facilitate searching the geospatial metadata, nor display search results in context.
The goal of the project was to make the user interface more usable, particularly for browsing and leaverage unique properties of the collection (geospatial metadata). What this means is that the item view, collection view, and search interface needed to be changed.

The item view was changed to display a gallery style set of thumbnails, with lightbox style previews. In addition, the item view included a JPEG as well as the TIFF file. The collection view was altered to contain a map that places every item into geographic context. This is done using Yahoo maps API. Lastly the search was changed to allow searching of geographic location and the results were mapped to geographic context.

Check out the revised interface to the collection and compare it to the original interface. In my opinion, it is a tremendous of improvement and goes to show how an interface can be altered to leaverage the unique metadata within a given collection.

Installation

The installation instruction for ETD-db while complete are not particularly enlightening if you want to install the software in a different place from what the software reccomends. Particularly if you want to install the Perl scripts of ETD-db in a different place (ie your cgi-bin). You can move things around but if you do you need to edit the etd-master-lib.pl which typically lives at the top level of your ETD-db cgi-bin files. Also if you change where you put the files this will also change the instructions for creating Script Aliases

Another installation problem I encountered was where ETD-db thought Perl was located on my system. This can quickly be fixed by creating a symbolic link from where ETD-db wants Perl to be to Perl’s actual location.

It is also important to note that when you change the location of the files, this may break Perl code that includes other Perl files (typically these are lines that say ‘required …’). Most notably this happens to the non-cgi-bin files. This is very fixable. Look for lines that say:

unless ($cgi_system_root) { $cgi_system_root = “/export/ETD-db”;}
require “$cgi_system_root/non-cgi-scripts/etd-non-cgi-lib.pl”;
and change them to

require “etd-non-cgi-lib.pl”;

Once you’ve installed ETD-db you probably want to customize it in some ways. However, the documentation on customization is a bit sparse. Below are some tips I’ve garnered via our customization process.

Customization

There are several places one should look in order to customize various screens in ETD-db. Where you look to customize a screen depends on if that screen is an HTML page or being generated by a Perl script. Most of the screens within ETD-db are created by Perl scripts. However some are HTML files which you need to edit. The most important of these are the index.html file which controls the main ETD-db page and the HTML file located in the help directory.
If the screen you want to alter is created by a Perl script, the first place is the /language/english directory. There are many “message” files here which govern what the screens look like and the information that appears on them. The first thing you need know is that ETD-db has groups its screens by the “process” which is taking place. These processes corresponded to particular Perl scripts that live in the ETD-db directory structure. Below is a list of the ETD-db processes, the corresponding directories and message files.

Remember that if you want to change a something about a particular screen for example you want to change how the “Add Advisor” screen works that there is an add advisor screen present in all of the following processes (Submit an ETD, Manage available ETDs, Review submitted ETDs, and Manage withheld ETDs) so you will need to change it in all these places.

The /language/english directory also contains a global_messages.pl file that control some universal things like the header and footer for each of the screens. It also contains the non-cgi_messages.pl file which is used by the scripts in the non-cgi-script directory and controls the output created by these scripts.

Most of the HTML generated by ETD-db is HTML 4.01 so if you want your ETD-db installation to be XHTML compliant then you will need to do some major work rewriting and cleaning up the message files. This process is tedious but not overly time consuming if you know which files you need to alter. Plus a good text editor with the ability to globally find and replace with several files will make the process go smoother.

Changing the screens in ETD-db is relatively simple as long as you aren’t adding new fields. If you do decide to add new fields (as we chose to do) then this will be a more complicated process. Below are some basic steps for adding a new “text” field.

1. Add new fields to all the relevant ETD-db databases (in most cases this is etd_submitted, etd_available, and etd_withheld)
2. Add new fields to global variables (language/english/global_variables.pl
3. Edit all of the forms so that the reflect the new field (these forms exist in the submit_messages.pl, review_messages.pl, maint_messages.pl – In submit_messages they are called manage_etd_table, modify_etd_form, enter_title_page_form. In the other message files they are modify_etd_main_table, view_etd_main_table)
4. Edit the scripts which process the form data and add it to the database. (ETD-submit/add_etd, ETD-submit/update_etd, ETD-review,maint,withheld/update_etd)
5. Edit form that generates title page for each dissertation (language/english/non-cgi_messages.pl)

The process becomes slightly more complicated if you want to add a new drop-down menu.

1. Add new fields to all the relevant ETD-db databases (in most cases this is etd_submitted, etd_available, and etd_withheld)
2. Add new fields to global variables (language/english/global_variables.pl
3. Create array of items in the new drop-down in languages/english/global_variables.pl
   Example:
   @department_list = (
   “College of Music”,
   “College of Natural Sciences”
   );
4. Edit all of the forms so that the reflect the new field (these forms exist in the submit_messages.pl, review_messages.pl, maint_messages.pl – In submit_messages they are called manage_etd_table, modify_etd_form, enter_title_page_form. In the other message files they are modify_etd_main_table, view_etd_main_table)
5. Edit the perl script that generates the form to create the drop-down menu (ETD-submit/enter_title_page , ETD-submit/modify_etd , ETD-review,maint,withheld/modify_etd)
6. Edit the scripts which process the form data and add it to the database. (ETD-submit/add_etd, ETD-submit/update_etd, ETD-review,maint,withheld/update_etd)
7. Edit form that generates title page for each dissertation (language/english/non-cgi_messages.pl)

If you want more information on customizing ETD-db, feel free to contact me. I’m not a Perl expert. However, I’m spent a lot of time figuring out how the system works to get it working right for us.

I’ve also worked out the issues with ETD-db and authentication. Coming soon a post on mod_auth_mysql!