Hi all, I am the digital architect for the new Omeka site for Thomas Edison Papers. I would appreciate any feedback or comments you have! There is a feedback form linked on the landing page, but you can also post here.
I have intentions to publish a “recipe” of this site for use as a template for future digital editions, including a generic version of the theme, etc… but for now, simply “launching” is taking enough time.
Site:
http://edison.rutgers.edu/digital/
Theme:
Based on Seasons theme
Includes decent responsive breakpoint customization, looking for input here
Plugins that have been used:
- Admin Images (Version 1.3 by UC Santa Cruz University Library)
- Clean Url (Version 2.16.1 by Daniel Berthereau)
- CSV Import+ (Version 2.3.3 by Daniel Berthereau - improvement of CSV Import of Center for History & New Media)
- Date Search (Version 0.1 by Gero Zahn)
- Default Sort (Version 1.2 by The Digital Ark)
- Hide Elements (Version 1.3 by John Flatness)
- History Log (Version 2.10 by UC Santa Cruz University Library and Daniel Berthereau)
- Reference (Version 2.4.4 by Daniel Berthereau)
- Search By Metadata (Version 1.2.1 by Roy Rosenzweig Center for History and New Media)
- Simple Pages (Version 3.1 by Roy Rosenzweig Center for History and New Media)
- Sitemap 2 (Version 2.3 by UC Santa Cruz University Library)
“Baked in” Javascript (inserted in theme):
fancybox3 - http://fancyapps.com/fancybox/3/docs/
clipboard - https://clipboardjs.com/
ISSUES
Due to some query speed issues, I have taken the Reference plugin HTML output and copied it into Simple Pages in certain cases, as the collection is mostly static and the page load time greatly decreases. I am going to look into adding a “render to Simple Pages” option for Reference plugin, this seems like a good pattern for large historical collections that are not changing much over time.
The sort time is a little bit slow, maybe 3-4 seconds, on the entire “Browse Documents” result set. I have a separate thread open on this, but it will probably require a custom SQL table to be loaded and then a custom SQL query to be written into the theme.
In order to conserve server space, we have symlinked all “derivative” images to the one original source image directory - the images are highly compressed JPG files (e.g. 1000px width or less, 50KB). In some cases the images do not load for a particular document, so I have to trace and see where the issue lies.
Thanks in advance for any feedback - bug reports, UX issues, suggestions, examples of other sites that do things similarly or better… whatever!
BEN BAKELAAR
Digital Architect
Thomas A. Edison Papers
Rutgers, The State University of New Jersey