#DivBlk: Principles in Action During a Website Migration
3) Quality Assurance
Semester: Winter 2019
Committee: Chair Michelle Byrnes (undergraduate), Lauren Cooper (librarian), Rachel Nelson (masters student), Quader’a Henry (undergraduate), and Briana Richardson (undergraduate)
Introduction: In Week 1, we discussed how important it was to have a full-time project manager to adapt work plans as the rhythm of an academic calendar changed and student skill sets varied. In Week 2, we shared how, after two years of work we thought we were ready to launch, but fresh eyes on the project helped to see some major work still outstanding. This week we get into the details of the depth of missing content (more than 100 pieces) and identify why we think this happened.
For this undertaking, several CCP team members from other committees joined the Website Committee. We were a mix of undergraduates, graduates, and librarian. Most were new to WordPress which also provided an opportunity to familiarize them with the platform and compare differences with Omeka Classic’s exhibit presentation.
The Website Committee set out to do a thorough audit of each exhibit. We documented missing pages, text/sections, images, visualizations, and note other issues such as awkward layouts. The team began the audit in late November 2018, broke for the two week for winter break, and finished in late-January 2019 during winter term.
The team audited 14 exhibits that had a total of 298 exhibit pages. We found:
- 118 missing pieces of content. This figure does not reflect the full complexity of what was missing which could be as simple as images or entire pages of content
- 43 pages needing a re-work
- 8 pages needing content to be taken out of pop up windows
- 1 URL needing updating
- 12 IT fixes
How did this happen? Why was so much content missing?
One reason was there wasn’t enough quality assurance built into the initial process. Team members had been assigned exhibits but there hadn’t been a checklist to ensure that all pages were accounted for in the transfer. [Knowing what we know now, we are designing a “Exhibit Transfer from Omeka to WordPress” work plan to share with the DH field.]
The second reason was a critical one that continued to frustrate us. We were using Omeka 2.1 and were unable to update to the current version because of changes a previous member had made to the core code without leaving documentation behind. Users who were logged in could view images that were private. These images often had captions with pertinent exhibit information. To make content visible to anyone, an item (image, document), then as now, to be marked “Public.” When the Spring 2018 Website Committee was transferring exhibits, it’s very likely that members were not logged in, and therefore, could not see images in a layout in the menu structure. A reasonable question that our exhibits and website team asked repeatedly was: “Is it possible the exhibit creator didn’t want to display that image wasn’t ready for public viewing?” This was absolutely a possibility but, given our background knowledge of the exhibits and realizing this issue between Public and Private, it was likely that the exhibit creator, being logged in, thought the item was available for Public viewing. Unfinished pages were pretty clear, and we made decisions about what was ready for public viewing and what was not, or we consolidated content with a like page to ensure the information was available.
Comparative view of the same exhibit page
The difference between a team member being logged in or logged out likely contributed to batches of content missing.
The third reason content could have been missing was because of the compressed timeline to complete the migration project. The Spring 2018 website team had 2-3 months to migrate content to two systems. Additionally, student team members left or moved on in the summer. The new project manager wasn’t starting until October so this left a huge gap when little attention was being focused on completing the project. After this lapse, our full time project reinitiated the work
.As we moved through our audit, we faced the tensions between 1) design/style 2) creator’s scholarly vision and 3) attending to our principles especially in regards to visuals and access. We speak more about this in blog installment #4) Ensuring Digital Space for Black Women and Access to Black Histories.
A scan of the extent of work left to be done revealed that we needed to develop a work plan to fix the challenges outlined above. But we were caught between systems, doing double entry which was causing waylays in other committee’s work. We needed to make some big decisions. It was too much to try to complete the exhibits and figure out Omeka S given our limitations. We decided to shift our platform choices and use WP and Omeka Classic. The question was how do we do that when we had a dysfunctional Omeka installation? This, too, would take research and a work plan to figure out.
While when weighed on scales that log efficiencies this process might not measure up, it was absolutely necessary to ensure that the scholarship was intact. We also developed a system to audit the content, manage the work flow, address technical limitations, and better understand Omeka Classic system’s functionality. To document the work we did, we are now developing a Omeka Classic to WordPress Work Flow we plan to share with the DH community.
READ MORE: #DivBlk: Principles in Action During a Website Migration
The Colored Conventions Project appreciates the support of:
The Colored Conventions Project was launched & cultivated at the University of Delaware from 2012-2020.