Category Archives: Technology

A simple tool with a lot of power: Project Estimates

It takes a lot to build and publish digital collections as you can see from the variety and scope of the blog posts here on Bitstreams.  We all have our internal workflows and tools we use to make our jobs easier and more efficient.  The number and scale of activities going on behind the scenes is mind-boggling and we would never be able to do as much as we do if we didn’t continually refine our workflows and create tools and systems that help manage our data and work.  Some of these tools are big, like the Duke Digital Repository (DDR), with its public, staff and backend interface used to preserve, secure, and provide access to digital resources, while others are small, like scripts built to transform ArchiveSpace output into a starter digitization guides.  In the Digital Production Center (DPC) we use a homegrown tool that not only tracks production statistics but is also used to do project projections and to help isolate problems that occur during the digitization process.  This tool is a relational database that is affectionately named the Daily Work Report and has collected over 9 years of data on nearly every project in that time.

A long time ago, in a newly minted DPC, supervisors and other Library staff often asked me, “How long will that take?”, “How many students will we need to digitize this collection?”, “What will the data foot print of this project be?”, “How fast does this scanner go?”, “How many scans did we do last year?”, “How many items is that?”.  While I used to provide general information and anecdotal evidence to answer all of these questions, along with some manual hunting down of this information, it became more and more difficult to answer these questions as the number of projects multiplied, our services grew, the number of capture devices multiplied and the types of projects grew to include preservation projects, donor requests, patron request and exhibits.  Answering these seemingly simple questions became more complicated and time consuming as the department grew.  I thought to myself, I need a simple way to track the work being done on these projects that would help me answer these recurring common questions.

We were already using a FileMaker Pro database with a GUI interface as a checkout system to assign students batches of material to scan, but it was only tracking what student worked on what material.  I decided I could build out this concept to include all of the data points needed to answer the questions above.  I decided to use Microsoft Access because it was a common tool installed on every workstation in the department, I had used it before, and classes and instructional videos abound if I wanted to do anything fancy.

Enter the Daily Work Report (DWR).  I created a number of discrete tables to hold various types of data: project names, digitization tasks, employee names and so on.  These fields are connected to a datasheet represented as a form, which allowed for dropdown lists and auto filling for rapid and consistent entry of information. 

At the end of each shift students and professionals alike fill out the DWR for each task they performed on each project and how long they worked on each task.  These range from the obvious tasks of scanning and quality control to more minute tasks of derivative creation, equipment cleaning, calibration, documentation, material transfer, file movement, file renaming, ingest prep, and ingest.

Some of these tasks may seem minor and possibly too insignificant to record but they add up.  They add up to ~30% of the time it takes to complete a project.   When projecting the time it will take to complete a project we collect Scanning and Quality Control data from a similar project, calculate the time and add 30%.

Common Digitization Tasks

Task
Hours Overall % of project
Scanning 406.5 57.9
Quality Control 1 133 19
Running Scripts 24.5 3.5
Collection Analysis 21 3
Derivative Creation 20.5 2.9
File Renaming 15.5 2.2
Material Transfer 14 2
Testing 12.5 1.8
Documentation 10 1.4
File Movement 9.75 1.4
Digitization Guide 7 1
Quality Control 2 6.75 1
Training 6 0.9
Quality Control 3 5.5 0.9
Stitching 3 0.4
Rescanning 1.5 0.2
Finalize 1.5 0.2
Troubleshooting 1.5 0.2
Conservation Consultation 1 0.1
Total 701 100

New Project Estimates

Using the Daily Work Report’s Datasheet View, the database can be filtered by project, then by the “Scanning” task to get the total number of scans and the hours worked to complete those scans.  The same can be done for the Quality Control task.  With this information the average number of scans per hour can be calculated for the project and applied to the new project estimate.

Gather information from an existing project that is most similar to the project you are creating the estimate for.  For example, if you need to develop an estimate for a collection of bound volumes that will be captured on the Zeutschel you should find a similar collection in the DWR to run your numbers.

Gather data from an existing project:

Scanning

  • Number of scans = 3,473
  • Number of hours = 78.5
  • 3,473/78.5 = 2/hr

Quality Control

  • Number of scans = 3,473
  • Number of hours = 52.75
  • 3,473/52.75 = 8/hr

Apply the per hour rates to the new project:

Estimated number of scans: 7,800

  • Scanning: 7,800 / 44.2/hr = 176.5 hrs
  • QC: 7,800 / 68.8/hr = 113.4 hrs
  • Total: 290 hrs
  • + 30%: 87 hrs
  • Grand Total: 377 hrs

Rolling Production Rate

When an update is required for an ongoing project the Daily Work Report can be used to see how much has been done and calculate how much longer it will take.  The number of images scanned in a collection can be found by filtering by project then by the “Scanning” Task.  That number can then be subtracted from the total number of scans in the project.  Then, using a similar project to the one above you can calculate the production rate for the project and estimate the number of hours it will take to complete the project.

Scanning

  • Number of scans in the project = 7,800
  • Number of scans completed = 4,951
  • Number of scans left to do = 7,800 – 4,951 = 2,849

Scanning time to completion

  • Number of scans left = 2,849
  • 2,849/42.4/hr = 2 hrs

Quality Control

  • Number of files to QC in the project = 7,800
  • Number of files completed = 3,712
  • Number of files left to do = 7,800 – 3,712 = 4,088

QC hours to completion

  • Number of scans left to scan = 4,088
  • 4,088/68.8 = 4 hrs

The amount of time left to complete the project

  • Scanning – 67.2 hrs
  • Quality Control – 59.4 hrs
  • Total = 126.2 hrs
  • + 30% = 38
  • Grand Total = 164.2 hrs

Isolate an error

Errors inevitably occur during most digitization projects.  The DWR can be used to identify how widespread the error is by using a combination of filtering, the digitization guide (which is an inventory of images captured along with other metadata about the capture process), and inspecting the images.  As an example, a set of files may be found to have no color profile.  The digitization guide can be used to identify the day the erroneous images were created and who created them. The DWR can be used to filter by the scanner operator and date to see if the error is isolated to a particular person, a particular machine or a particular day.  This information can then be used to filter by the same variables across collections to see if the error exists elsewhere.  The result of this search can facilitate retraining, recalibrating of capture devices and also identify groups of images that need to be rescanned without having to comb through an entire collection.

While I’ve only touched on the uses of the Daily Work Report, we have used this database in many different ways over the years.  It has continued to answer those recurring questions that come up year after year.  How many scans did we do last year?  How many students worked on that multiyear project?  How many patron requests did we complete last quarter?  This database has helped us do our estimates, isolate problems and provide accurate updates over the years.  For such a simple tool it sure does come in handy.

Web Accessibility: Values and Vigilance

The Duke Libraries are committed to providing outstanding service based on respect and empathy for the diverse backgrounds and needs in our community. Our guiding principles make clear how critically important diversity and inclusion are to the library, and the extent to which we strive to break down barriers to scholarship.

One of the biggest and most important barriers for us to tackle is the accessibility of our web content. Duke University’s Web Accessibility site sums it up well:

Duke believes web content needs to be accessible to people with a wide range of abilities, including visual, auditory, physical, speech, cognitive, language, learning, and neurological abilities.

Screenshot of Duke Web Accessibility homepage
The Duke Web Accessibility website is a tremendous resource for the Duke community.

This belief is also consistent with the core values expressed by the American Library Association (ALA). A library’s website and online resources should be available in formats accessible to people of all ages and abilities.

Web Content

As one of the largest research libraries in the U.S., we have a whole lot of content on the web to consider.

Our website alone comprises over a thousand pages with more than fifty staff contributors. The library catalog interface displays records for over 13 million items at Duke and partner libraries. Our various digital repositories and digital exhibits platforms host hundreds of thousands of interactive digital objects of different types, including images, A/V, documents, datasets, and more. The list goes on.

Any attempt to take a full inventory of the library’s digital content reveals potentially several million web pages under the library’s purview, and all that content is managed and rendered via a dizzying array of technology platforms. We have upwards of a hundred web applications with public-facing interfaces. We built some of these ourselves, some are community-developed (with local customizations), and others we have licensed from vendors. Some interfaces are new, some are old. And some are really old, dating all the way back to the mid-90s.

Ensuring that this content is equally accessible to everyone is important, and it is indeed a significant undertaking. We must also be vigilant to ensure that it stays accessible over time.

With that as our context, I’d like to highlight a few recent efforts in the library to improve the accessibility of our digital resources.

Style Guide With Color Contrast Checks

In January 2019, we launched a new catalog, replacing a decade-old platform and its outdated interface. As we began developing the front-end, we knew we wanted to be consistent, constrained, and intentional in how we styled elements of the interface. We were especially focused on ensuring that any text in the UI had sufficient contrast with its background to be accessible to users with low vision or color-blindness.

We tried out a few existing “living style guide” frameworks. But none of them proved to be a good fit, particularly for color contrast management. So we ended up taking a DIY approach and developed our own living style guide using Javascript and Ruby.

Screenshot of the library catalog style guide showing a color palette.
The library catalog’s living style guide dynamically checks for color contrast accessibility.

Here’s how it works. In our templates we specify the array of color variable names for each category. Then we use client-side Javascript to dynamically measure the hex & RGB values and the luminance of each color in the guide. From those figures, we return score labels for black and white contrast ratios, color-coded for WCAG 2.0 compliance.

This style guide is “living” in that it’s a real-time up-to-date reflection of how elements of the UI will appear when using particular color variable names and CSS classes. It helps to guide developers and other project team members to make good decisions about colors from our palette to stay in compliance with accessibility guidelines.

Audiovisual Captions & Interactive Transcripts

In fall 2017, I wrote about an innovative, custom-developed feature in our Digital Repository that renders interactive caption text for A/V within and below our media player. At that time, however, none of our A/V items making use of that feature were available to the public.  In the months since then, we have debuted several captioned items for public access.

We extended these features in 2018, including: 1) exporting captions on-the-fly as Text, PDF, or original WebVTT files, and 2) accommodating transcript files that originated as documents (PDF, Word)

Screenshot of an interactive transcript with export options
WebVTT caption files for A/V are rendered as interactive HTML transcripts and can be exported into text or PDF.

Two of my talented colleagues have shared more about our A/V accessibility efforts at conferences over the past year. Noah Huffman presented at ARCHIVES*RECORDS (Joint Annual Meeting of CoSA, NAGARA, and SAA) in Aug 2018. And Molly Bragg presented at Digital Library Federation (DLF) Forum (slides) in Nov 2018.

Institutional Repository Accessibility

We have documented our work over 2018 revitalizing DSpace at Duke, and then subsequently developing a new set of innovative features that highlight Duke researchers and the impact of their work. This spring, we took a closer look at our new UI’s accessibility following Duke’s helpful guide.
In the course of this assessment, we were able to identify (and then fix!) several accessibility issues in DukeSpace. I’ll share two strategies in particular from the guide that proved to be really effective. I highly recommend using them frequently.

The Keyboard Test

How easy is it to navigate your site using only your keyboard? Can you get where you want to go using TAB, ENTER, SPACE, UP, and DOWN?  Is it clear which element of the page current has the focus?
Screenshot of DukeSpace homepage showing skip to content link
A “Skip to main content” feature in DukeSpace improves navigation via keyboard or assistive devices.
This test illuminated several problems. But with a few modest tweaks to our UI markup, we were able to add semantic markers to designate page sections and a skip to main content link, making the content much more navigable for users with keyboards and assistive devices alike.

A Browser Extension

If you’re a developer like me, chances are you already spend a lot of time using your browser’s Developer Tools pane to look under the hood of web pages, reverse-engineer UIs, mess with styles and markup, or troubleshoot problems.
The Deque Systems aXe Chrome Extension (also available for Firefox) integrates seamlessly into existing Dev Tools. It’s a remarkably useful tool to have in your toolset to help quickly find and fix accessibility issues. Its interface is clear and easy to understand. It finds and succinctly describes accessibility problems, and even tells you how to fix them in your code.
An image from the Deque aXe Chrome extension site showing the tool in action.
With aXe testing, we quickly learned we had some major issues to fix. The biggest problems revealed were missing form labels and page landmarks, and low contrast on color pairings. Again, these were not hard to fix since the tool explained what to do, and where.
Turning away from DSpace for a moment, see this example article published on a popular academic journal’s website. Note how it fares with an automated aXe accessibility test (197 violations of various types found).  And if you were using a keyboard, you’d have to press Tab over 100 times in order to download a PDF of the article.
Screenshot of aXe Chrome extension running on a journal website.
UI for a published journal article in a publisher’s website after running the aXe accessibility test. Violations found: 197.

Now, let’s look at the open access copy of that same article that resides in our DukeSpace site. With our spring 2019 DukeSpace accessibility revisions in place, when we run an aXe test, we see zero accessibility violations. Our interface is also now easily navigated without a mouse.

Screenshot or DukeSpace UI showing no violations found by aXe accessibility checker
Open access copy of an article in DukeSpace: No accessibility violations found.

Here’s another example of an open access article in DukeSpace vs. its published counterpart in the website of a popular journal (PNAS).  While the publisher’s site markup addresses many common accessibility issues, it still shows seven violations in aXe. And perhaps most concerning is that it’s completely unnavigable via a keyboard: the stylesheets have removed all focus styles from displaying.

Concluding Thoughts

Libraries are increasingly becoming champions for open access to scholarly research. The overlap in aims between the open access movement and web accessibility in general is quite striking. It all boils down to removing barriers and making access to information as inclusive as possible.

Our open access repository UIs may never be able to match all the feature-rich bells and whistles present in many academic journal websites. But accessibility, well, that’s right up our alley. We can and should do better. It’s all about being true to our values, collaborating with our community of peers, and being vigilant in prioritizing the work.

Look for many more accessibility improvements throughout many of the library’s digital resources as the year progresses.


Brief explanatory note about the A11Y++ image in this post: A11Y is a numeronym — shorthand for the word “accessibility” and conveniently also visually resembling the word “ally.” The “++” is an increment operator in many programming languages, adding one to a variable. 

Mythical Beasts of Audio

Gear. Kit. Hardware. Rig. Equipment.

In the audio world, we take our tools seriously, sometimes to an unhealthy and obsessive degree. We give them pet names, endow them with human qualities, and imbue them with magical powers. In this context, it’s not really strange that a manufacturer of professional audio interfaces would call themselves “Mark of the Unicorn.”

Here at the Digital Production Center, we recently upgraded our audio interface to a MOTU 896 mk3 from an ancient (in tech years) Edirol UA-101. The audio interface, which converts analog signals to digital and vice-versa, is the heart of any computer-based audio system. It controls all of the routing from the analog sources (mostly cassette and open reel tape decks in our case) to the computer workstation and the audio recording/editing software. If the audio interface isn’t seamlessly performing analog to digital conversion at archival standards, we have no hope of fulfilling our mission of creating high-quality digital surrogates of library A/V materials.

Edirol UA-101
The Edirol enjoying its retirement with some other pieces of kit

While the Edirol served us well from the very beginning of the Library’s forays into audio digitization, it had recently begun to cause issues resulting in crashes, restarts, and lost work. Given that the Edirol is over 10 years old and has been discontinued, it is expected that it would eventually fail to keep up with continued OS and software updates. After re-assessing our needs and doing a bit of research, we settled on the MOTU 896 mk3 as its replacement. The 896 had the input, output, and sync options we needed along with plenty of other bells and whistles.

I’ve been using the MOTU for several weeks now, and here are some things that I’m liking about it:

  • Easy installation of drivers
  • Designed to fit into standard audio rack
  • Choice of USB or Firewire connection to PC workstation
  • Good visual feedback on audio levels, sample rate, etc. via LED meters on front panel
  • Clarity and definition of sound
MOTU 896mk3
The MOTU sitting atop the audio tower

I haven’t had a chance to explore all of the additional features of the MOTU yet, but so far it has lived up to expectations and improved our digitization workflow. However, in a production environment such as ours, each piece of equipment needs to be a workhorse that can perform its function day in and day out as we work our way through the vaults. Only time can tell if the Mark of the Unicorn will be elevated to the pantheon of gear that its whimsical name suggests!

Is there an app for that? The seemingly endless quest to make discovery easier for users

Contributed by Assessment & User Experience Department Practicum Students Amelia Midgett-Nicholson and Allison Cruse 

Duke University Libraries (DUL) is always searching for new ways to increase access and make discovery easier for users. One area users frequently have trouble with is accessing online articles. Too often we hear from students that they cannot find an article PDF they are looking for, or even worse, that they end up paying to get through a journal paywall. To address this problem, DUL’s Assessment and User Experience (AUX) Department explored three possible tools: LibKey Discovery, Kopernio, and Lean Library. After user testing and internal review, LibKey Discovery emerged as the best available tool for the job.  

LibKey Discovery logo

LibKey Discovery is a suite of user-friendly application programming interfaces (APIs) used to enhance the library’s existing discovery system.  The APIs enable one-click access to PDFs for subscribed and open-source content, one-click access to full journal browsing via the BrowZine application, and access to cover art for thousands of journals.  The tool integrates fully with the existing discovery interface and does not require the use of additional plug-ins.

According to their website, LibKey Discovery has the potential to save users thousands of clicks per day by providing one-click access to millions of articles.  The ability to streamline processes enabling the efficient and effective discovery and retrieval of academic journal content prompted the AUX department to investigate the tool and its capabilities further.  An internal review of the system was preceded by an introduction of the tool to Duke’s subject librarians and followed with a preliminary round of student-based user testing.

Current DUL interface
Current DUL discovery interface
LibKey interface
LibKey discovery interface

Pros

  • One-Click Article and Full Journal Access

Both the AUX staff and the subject librarians who performed an initial review of the LibKey Discovery tools were impressed with the ease of article access and full journal browsing.  Three members of the AUX department independently reviewed LibKey’s features and concluded the system does provide substantial utility in its ability to reduce the number of clicks necessary to access articles and journals.

  • Streamlined Appearance

The tool streamlines the appearance and formatting of all journals, thus removing ambiguity in how to access information from different sources within the catalog.  This is beneficial in helping to direct users to the features they want without having to search for points of access. The AUX department review team all found this helpful.

  • Seamless Integration

LibKey Discovery’s APIs integrate fully into the existing DUL discovery interface without the need for users to download an additional plug-in.  This provides users the benefit of the new system without asking them to go through extra steps or make any changes to their current search processes.  Aside from the new one-click options available within the catalog’s search results page, the LibKey interface is indistinguishable from the current DUL interface helping users to benefit from the added functionality without feeling like they need to learn a new system.  

Cons

  • Cost

LibKey Discovery carries a relatively hefty price tag, so its utility to the end-user must be weighed against its cost.  While internal review and testing has indicated LibKey Discovery has the ability to streamline and optimize the discovery process, it must be determined if those benefits are universal enough to warrant the added annual expenditure.

  • Inconsistency in Options

A potential downside to LibKey Discovery is lack of consistency in one-click options between articles.  While many articles provide the option for easy, one-click access to a PDF, the full text online, and full journal access, these options are not available for all content.  As a result, this may cause confusion around the options that are available for users and may diminish the overall utility of the tool depending on what percentage of the catalog’s content is exempt from the one-click features.

LibKey Discovery User Testing Findings

An initial round of user testing was completed with ten student volunteers in the lobby of Perkins Library in early April.  Half of the users were asked to access an article and browse a full journal in the existing DUL system; the other half were asked to perform the same tasks using the LibKey Discovery interface.

Initial testing indicated that student users had a high level of satisfaction with the LibKey interface; however, they were equally satisfied with the existing access points in the DUL catalog.  The final recommendations from the user testing report suggest the need for additional testing to be completed. Specifically, it was recommended that more targeted testing be completed with graduate-level students and faculty as a majority of the original test’s participants were undergraduate students with limited experience searching for and accessing academic journal issues and articles.  It was concluded that testing with a more experienced user group would likely produce better feedback as to the true value of LibKey Discovery.

LibKey Summary

LibKey Discovery is a promising addition to Duke’s existing discovery system.  It allows for streamlined, one-click article and full journal access without disrupting the look and feel of the current interface or requiring the use of a plug-in.  Initial reviews of the system by library staff have been glowing; however, preliminary user testing with student participants indicated the need for additional testing to determine if LibKey’s cost is sufficiently offset by its utility to the user.

Kopernio logo

Kopernio is a free browser plug-in which enables one-click access to academic journal articles. It searches the web for OA copies, institutional repository copies, and copies available through library subscriptions. The tool is designed to connect users to articles on and off campus by managing their subscription credentials and automatically finding the best version of an article no matter where a user is searching.

Given the potential of this tool to help increase access and make discovery easier for students, the AUX department initiated an internal review process. Four members of the department independently downloaded the Kopernio plug-in, thoroughly tested it in a variety of situations, and shared their general and specific notes about the tool.

Pros

  • OA Content + Library Subscription

By its design, Kopernio has an advantage over other plug-in tools that serve a similar function (i.e. Unpaywall). When users first download Kopernio they are asked to register their subscription credentials. This information is saved in the plug-in so users can automatically discover articles available from OA sources, as well as library subscriptions. This is an advantage over other plug-ins that only harvest from freely available sources.

Screenshot: Kopernio sign-in page
Kopernio sign-in page
  • Branding

Kopernio has highly visible and consistent branding. With bright green coloring, the plug-in stands out on a screen and attracts users to click on it to download articles.

  • One-Click

Kopernio is advertised as a “one-click” service, and it pays off in this respect. Using Kopernio to access articles definitely cuts down on the number of clicks required to get to an article’s PDF. The process to download articles to a computer was instantaneous, and most of the time, downloading to the Kopernio storage cloud was just as fast.

Cons

  • Creates New Pain Points

Kopernio’s most advertised strength is its ability to manage subscription credentials. Unfortunately, this strength is also a major data privacy weakness. Security concerns ultimately led to the decision to disable the feature which allowed users to access DUL subscriptions via Kopernio when off-campus. Without this feature, Kopernio only pulls from OA sources and therefore performs the same function that many other tools currently do.

Similar to data privacy concerns, Kopernio also raises copyright concerns. One of Kopernio’s features is its sharing function. You can email articles to anyone, regardless of their university affiliation or if they have downloaded Kopernio already. We tested sending DUL subscription PDFs to users without Duke email addresses and they were able to view the full-text without logging in. It is unclear if they were viewing an OA copy of the article, or if they were seeing an article only meant for DUL authenticated users.

Screenshot: Sharing articles through Kopernio
Sharing an article through Kopernio

Running the Kopernio plug-in noticeably slowed down browser speed. We tested the browser on several different computers, both on campus and off, and we all noticed slower browser speeds. This slow speed led Kopernio to be occasionally buggy (freezing, error messages etc.).

Screenshot showing a buggy screen from Kopernio
Buggy screen while using Kopernio
  • Many Features Don’t Seem Useful

When articles are saved to Kopernio’s cloud storage, users can add descriptive tags. We found this feature awkward to use. Instead of adding tags as you go along, users have to add a tag globally before they can tag an article. Overall, it seemed like more hassle than it was worth.

Kopernio automatically imports article metadata to generate citations. There were too many problems with this feature to make it useful to users. It did not import metadata for all articles that we tested, and there was no way to manually add metadata yourself. Additionally, the citations were automatically formatted in Elsevier Harvard format and we had to go to our settings to change it to a more common citation style.

Lastly, the cloud storage which at first seemed like an asset, was actually a problem. All articles automatically download to cloud storage (called the “Kopernio Locker”) as soon as you click on the Kopernio button. This wouldn’t be a problem except for the limited storage size of the locker. With only 100MB of storage in the free version of Kopernio, we found that after downloading only 2 articles the locker was already 3% full. To make this limited storage work, we would have to go back to our locker and manually delete articles that we did not need, effectively negating the steps saved by having an automatic process.

Lean Library Logo

Lean Library is a similar tool to Kopernio. It offers users one-click access to subscription and open access content through a browser extension. In Fall 2018, DUL staff were days away from purchasing this tool when Lean Library was acquired by SAGE Publishing. DUL staff had been excited to license a tool that was independent and vendor-neutral and so were disappointed to learn about its acquisition. We have found that industry consolidation in the publishing and library information systems environment has lowered competition and resulted in negative experiences for researchers and staff. Further, we take the privacy of our users very seriously and were concerned that Lean Library’s alignment with SAGE Publishing will compromise user security. Whenever possible, DUL aims to support products and services that are offered independently from those with already dominant market positions. For these reasons, we opted not to pursue Lean Library further.

Conclusion

Of the three tools the AUX Department explored, we believe LibKey Discovery to be the most user-friendly and effective option. If purchased, it should streamline journal browsing and article PDF downloads without adversely affecting the existing functionality of DUL’s discovery interfaces.

Smart People Who Care

It’s that time of year at the university when we’re working on our PEPs (Performance Evaluation and Planning forms) and I’m thinking about how grateful I am to have such smart staff who really care about their work, their colleagues, and the people they serve, as we advance technology across the libraries. In contrast to some corporate environments, the process here really does aim to help us improve, rather than rank us as a setup for “resource actions” (firings). This excellent article, The Feedback Fallacy by Marcus Buckingham and Ashley Goodall, reminds me to emphasize the things people do well, and encourage them to build on their strengths.

Attuned to ethical practices within organizations, I’m also excited about increasing awareness of ethics in the effects of the technologies we produce. Justin Sherman, co-founder of the Ethical Tech initiative here at Duke, did a stimulating talk at the Edge Workshop this month about ethical issues that surround technology, such as search engine bias, and AI tools that judges use to determine sentencing for crimes.  Justin recommends this podcast, with Christopher Lydon on Open Source, called Real Education About Artificial Intelligence.  Library staff are participating in the Kenan Institute for Ethics book club program (KIE), where the spring selection is Algorithms of Oppression: How Search Engines Reinforce Racism by Safiya Umoja Noble.

And, I’m pleased to exercise my hiring mantra, “smart people who care”, which has served me well for over 30 years, as we’re recruiting candidates with I/T and team leadership experience for a new position, Computing Services Supervisor.

Happy Spring!
Laura Cappelletti
Director, Information Technology Services

Sustainability Planning for a Better Tomorrow

In March of last year I wrote about efforts of the Resource Discovery Systems and Strategies team (RDSS, previously called the Discovery Strategy Team) to map Duke University Libraries’ discovery system environment in a visual way. As part of this project we created supporting documentation for each system that appeared in a visualization, including identifying functional and technical owners as well as links to supporting documentation. Gathering this information wasn’t as straightforward as it ideally should have been, however. When attempting to identify ownership, for example, we were often asked questions like, “what IS a functional owner, anyway?”, or told “I guess I’m the owner… I don’t know who else it would be”. And for many systems, local documentation was outdated, distributed across platforms, or simply nonexistent.

As a quick glance through the Networked Discovery Systems document will evince, we work with a LOT of different systems here at DUL, supporting a great breadth of processes and workflows. And we’ve been steadily adding to the list of systems we support every year, without necessarily articulating how we will manage the ever-growing list. This has led to situations of benign neglect, confusion as to roles and responsibilities and, in a few cases, we’ve hung onto systems for too long because we hadn’t defined a plan for responsible decommission.

So, to promote the healthier management of our Networked Discovery Systems, the RDSS team developed a set of best practices for sustainability planning. Originally we framed this document as best practices for maintenance planning, but in conversations with other groups in the Libraries, we realized that this didn’t quite capture our intention. While maintenance planning is often considered from a technical standpoint, we wanted to convey that the responsible management of our systems involves stakeholders beyond just those in ITS, to include the perspective and engagement of non-technical staff. So, we landed on the term sustainability, which we hope captures the full lifecycle of a system in our suite of tools, from implementation, through maintenance, to sunsetting, when necessary.

The best practices are fairly short, intended to be a high-level guide rather than overly prescriptive, recognizing that every system has unique needs. Each section of the framework is described, and key terms are defined. Functional and technical ownership are described, including the types of activities that may attend each role, and we acknowledge that ownership responsibilities may be jointly accomplished by groups or teams of stakeholders. We lay out the following suggested framework for developing a sustainability plan, which we define as “a living document that addresses the major components of a system’s life cycle”:

  • Governance:
    • Ownership
    • Stakeholders
    • Users
  • Maintenance:
    • System Updates
    • Training
    • Documentation
  • Review:
    • Assessments
    • Enhancements
    • Sunsetting

Interestingly, and perhaps tellingly, many of the conversations we had about the framework ended up focusing on the last part – sunsetting. How to responsibly decommission or sunset a system in a methodical, process-oriented way is something we haven’t really tackled yet, but we’re not alone in this, and the topic is one that is garnering some attention in project management circles.

So far, the best practices have been used to create a sustainability plan for one of our systems, Dukespace, and the feedback was positive. We hope that these guidelines will facilitate the work we do to sustain our system, and in so doing lead to better communication and understanding throughout the organization. And we didn’t forget to create a sustainability plan for the best practices themselves – the RDSS team has committed to reviewing and updating it at least annually!

Ad-Blocking … and a Mea Culpa

It’s all but impossible to use the internet and not be aware of the sheer quantity of advertising out there. Some estimates suggest Google alone serves out nearly 30 Billion ads per day; other estimates suggest 300-700 ads are shown per person per day. In trying to get more eyeballs on their images, advertisers have resorted to more and more intrusive ad displays — pop-over ads (so you can’t see content until you close the ad), pop-under ads (so even after you’ve left the site, you’ll see one more ad), animated gifs (the motion naturally causes your eye to look at it), auto-playing (and loud) videos. More recently, ads have even been implicated in malware attacks.

So it’s no surprise that ad-blocking technology is now mainstream. All major browsers have multiple ad-blocker plug-ins, and any list of “Best Add-Ons for <browser>” will likely include at least one. By blocking ads, these plugins reduce the annoyance of the ads while also helping protect your privacy by reducing sites’ ability to track you.

As an additional bonus, they can also accelerate your web browsing — by not downloading all that ad content, you’re able to see the “real” content faster. A New York Times article showed that between 33% and 50% of the time to download a page was due to advertising — and in extreme cases, advertising could be 80% or more of the time to download and view a webpage.

In IT-Core Services, we had intended to deploy an ad-blocker to all of the public computers in order to allow our patrons and users to block ads while they were doing work, collecting papers, or doing research on Library computers.  The plugin is called “uBlock Origin” and is one of the leading open-source ad-blockers around.

But … Oops.

We accidentally pushed it to all public AND staff computers several months ago.

Given the very few number of tickets we’ve seen about it, we’re guessing people either didn’t notice, or else welcomed the arrival of uBlock.  We’re now planning on keeping uBlock deployed on all staff and public computers.  We feel that the privacy, performance, and security benefits of uBlock outweigh the desire for a “ad-full” web experience — and you can easily un-block any websites you want to, if you find that the blocker is somehow interfering with that site.

How to Un-Block a Website:

To unblock the websitethat you’re visiting — that is, to show the ads on the page — look for the uBlock logo (a brick-red shield) at the top of the main browser window.  Clicking that logo will pop up a dialog box like this:By clicking on the power-button symbol (circle with the line at the top), you’ll tell uBlock to NOT block ads on that webpage in the future.  You should then reload the page to get the ad-full experience (by default, un-blocking a website does NOT reload or re-display the ads, you must explicitly reload the page).  Note:  if the uBlock logo is grey, or the power-button icon is grey, then the current website is already un-blocked (and the browser is showing you the ads).

How to Un-Block All/Many Websites:

To unblock a lot of websites at once, you have to go to the uBlock “Dashboard” or settings menu.  Again, click on the uBlock shield logo at the top of the browser window, then look for the 3-sliders icon (immediately below the power-button, to the right of the dialog box).  Clicking that will bring up a new virtual webpage with a variety of settings on it:

Click on the “Filter Lists” tab and you’ll see a set of “Filter” checkboxes.  Each checkbox represents a set of websites that are to be blocked (checked) or unblocked (unchecked).  To unblock all websites — to essentially deactivate uBlock Origin altogether — just uncheck all of the Filter sets.  FWIW, most of the filter-sets have Home icons where you can find more info on what that filter-set does (e.g. “Malware Domains” links to a website at malwaredomains.com, which is run and maintained by the company RiskAnalytics).

If you have any questions, please just submit a ticket and someone will get back to you.

(title image is the “Million Dollar Webpage“, I’ll link you to the Wikipedia page rather than the ad-full page!)


UPDATE – 15 Feb 2019 – uBlock and DUL Newsletters:

While a great many websites will work fine with uBlock Origin installed, it turns out the Library’s own newsletter system does NOT!  If you are experiencing problems with the newsletter, go into the uBlock settings (process described above) and go to the “Filter Lists” tab.  One of the filter-sets at the bottom is named “Peter Lowe’s Ad and tracking server list” — this is the one that seems to catch the iContact server used by our newsletter.  If you disable that (the box will be un-checked), then reload the page, you should be back in operation.  Sorry!

New Duke Libraries catalog to go live January 16

The wait is almost over!  After over two years of hard work by library staff at Duke, NCCU, NCSU, and UNC, we’ll be launching a new catalog for researchers to use to locate and access books, DVDs, music, archival materials, and other items held here at Duke and across the other Triangle Research Libraries member libraries (NCSU, NCCU, UNC).  The new collaboratively developed, open-source library catalog will replace the decade-old Duke Libraries catalog and Search TRLN catalog.    

What to expect from the new catalog

While the basic functionality of the catalog remains the same – researchers will be able to search across the 15 million+ items at Duke and area libraries – we think you’ll enjoy some nice enhancements and updates to the catalog:  

  • a more modern search interface
  • prominent filters to modify search results, including an option to narrow your search to items available online
  • a more intuitive button to expand search results to include items held at UNC, NCCU, and NCSU
  • updated item pages with easier access to request and export materials
  • improved access to an item’s availability status and its location  
  • more user-friendly options to email, text, or export citation information
  • improved display of multi-part items (click “View Online” to access individual episodes)
  • more robust Advanced Search with options to search by Publisher and Resource type (e.g., book, video, archival material)
Screenshot of full item page.
Item pages have been updated and include a sidebar with easy access to Request and Email/Text options.

You might notice some differences in the way the new catalog works.  Learn more with this handy new vs. old catalog comparison chart.  (Note that we plan to implement some of the features that are not currently available in the new catalog this spring – stay tuned for more info, and let us know if there are aspects of the old catalog that you miss.)  And if you run into trouble or have more questions about using the new catalog, check out these library catalog search tips, or contact a librarian for assistance.   

Screenshot of Search Tips webpage
Get tips for using the new catalog.

We welcome your feedback

While the new catalog is fully functional and boasts a number of enhancements, we know there is still work to be done, and we welcome your input.  We invite you to explore the new catalog at https://find.library.duke.edu/ and report problems or provide general feedback through this online form.  We’ll continue to conduct user testing this spring and make improvements based on what we learn – look for the “Free Coffee” sign in the Perkins Library lobby, and stop by to tell us what you think.  

Want more info about this project?  Learn about the vision for developing the new catalog and the work that’s been completed to date.

Highlighting Duke Scholars Alongside Their Research

2018 has featured several monumental changes in the library’s technology platforms. One of the most impactful shifts this year was revitalizing DukeSpace, our DSpace-based institutional repository (IR) software, home to over 16,000 open-access articles, theses, and dissertations from Duke scholars.

Back in March, we celebrated a successful multi-version upgrade for DSpace, and along with it, a major upgrade to the integral Symplectic Elements Research Information Management platform.  On the heels of that project, we decided to capitalize on the project team’s momentum and invest two more months of focused attention (four “sprints” in developer-speak).

The goals for that period? First, tie up the loose ends from the upgrade. Then, seize some clear opportunities to build upon our freshly-rearchitected metadata, creating innovative features in the UI to benefit scholars. By scholars, we mean — in part — the global audience of researchers openly discovering and using the articles in DukeSpace. But we especially mean the scholars at Duke who created them in the first place.

We are excited to share the results of our work with the Duke community and beyond. Here are the noteworthy additions:

Scholars@Duke Author Profiles

Item pages now display a brief embedded profile for each Duke author, featuring their preferred name, a photo, position title, and brief description of their research interests. This information comes from the scholars themselves, who manage their profiles via Scholars@Duke (powered by the open-source VIVO platform).

Scholars@Duke & DukeSpace
A brief author profile in a DukeSpace item  leads to the full profile in Scholars@Duke.

Scholars@Duke provides a handy SEO-friendly profile page (example) for each scholar. It aggregates their full list of publications, courses taught, news articles in which they’re mentioned, and much more. It also has useful APIs for building widgets to dynamically repurpose the data for display elsewhere. Departments throughout Duke (example) use these APIs to display current faculty information on their web sites without requiring anyone to manually duplicate or update it. And now, the library does, too.

Scholars@Duke profile in DukeSpace
Scholars’ own descriptions of their research interests entice a visitor to learn more about their work.

Featuring researchers in this manner adds several other benefits, including:

  • Uses a scholar’s own preferred current version of their name and title; that may not be identical to what is stored in the item’s author metadata.
  • Puts users one click away from the author’s full profile at Scholars@Duke, where one can discover the entirety of the author’s publications, open-access or not.
  • Helps search engines make stronger semantic connections between an author’s profile information and their works available online.
  • Introduces a unique value-add feature for our open-access copy of an article that’s unlikely to ever be possible to replicate for the published version on the academic journal’s website.
  • Makes the DukeSpace item pages look better, warmer, and more inviting.
Scholars@Duke multiple author profiles in DukeSpace
Profiles for multi-author articles

With this feature, we are truly pushing beyond the boundaries of what an institutional repository traditionally does. And likewise, we feel we’re raising the bar for how academic research libraries can showcase the members of their communities alongside their collected works.

Other New Features

Beyond these new author profiles, we managed to fit in a few more enhancements around citations, the homepage, and site navigation. Here’s a quick rundown:

Citations

We now present an easily copyable citation, composed from the various metadata available for each item. This includes the item’s permalink.

DukeSpace citation for a dissertation
DukeSpace citation for a dissertation

In cases when there’s a published version of an article available, we direct users to review and use that citation instead.

DukeSpace citation with DOI
DukeSpace article with a DOI for a published version.

 

Item pages also now display a “Citation Stats” badge in the sidebar, powered by Digital Science’s Dimensions tool.  Click it to explore other scholarly work that has cited the current item.

Homepage

Finally, we topped off this project phase by redesigning DukeSpace’s homepage. Notable among the changes: a clearer indication of the number of items (broken down by type), a dynamic list of trending items, and streamlined menu navigation in the sidebar.

 

Final Thoughts

Duke Libraries’ current strategic plan emphasizes the mission-critical importance of our open-access publishing and repository efforts, and also demands that we “highlight and promote the scholarly activities of our faculty, students, and staff.”  This two-month DukeSpace enhancements project was a great opportunity for us to think outside the box about our technology platforms, and consider how those goals relate.


Many thanks to several people whose work enabled these features to come to life, especially Maggie Dickson, Hugh Cayless, Paolo Mangiafico, and the Scholars@Duke team.

A New(-ish) Look for Public Computing

Photo of library public computing terminals

Over the past year, you’ve probably noticed a change in the public computing environments in Duke University Libraries. Besides new patron-facing hardware, we’ve made even larger changes behind the scenes — the majority of our public computing “computers” have been converted to a Virtual Desktop Infrastructure (VDI).

The physical hardware that you sit down at looks a little different, with larger monitors and no “CPU tower”:

Close-up photo of a public terminal

What isn’t apparent is that these “computers” actually have NO computational power at all! They’re essentially just a remote keyboard and monitor that connects to a VDI-server sitting in a data-center.
Photo of the VDI serverThe end-result is really that you sit down at what looks like a regular computer, and you have an experience that “feels” like a regular computer.  The VDI-terminal and VDI-server work together to make that appear seamless.

All of the same software is installed on the new “computers” — really, virtual desktop connections back to the server — and we’ve purchased a fairly “beefy” VDI-server so that each terminal should feel very responsive and fast.  The goal has been to provide as good an experience on VDI as you would get on “real” computers.

But there are also some great benefits …

Additional Security:
When a patron sits down at a terminal, they are given a new, clean installation of a standard Windows environment. When they’re done with their work, the system will automatically delete that now-unused virtual desktop session, and then create a brand-new one for the next patron. From a security standpoint, this means there is no “leakage” of any credentials from one user to another — passwords, website cookies, access tokens, etc. are all wiped clean when the user logs out.

Reduced Staff Effort:
It also offers some back-end efficiency for the Specialized Computing team. First off, since the VDI-terminal hardware is less complex (it’s not a full computer), the devices themselves have been seen to last 7 to 10 years (vs. 4 years for a standard PC). There have also been reports that they can take quite a beating and remain operational (and while I don’t want to jinx it, there are reports of them being fully submerged in water and, once dried out, being fully functional).

Beyond that, when we need to update the operating system or software, we make the change on one “golden image” and that image is copied to each new virtual desktop session. So despite having 50 or more public computing terminals, we don’t spend 50-times as much effort in maintaining them.

It is worth noting that we can also make these updates transparent to our patrons. After logging in, that VDI session will remain as-is until the person logs out — we will not reboot the system from under them.  Once they logout, the system deletes the old, now-outdated image and replaces it with a new image. There is no downtime for the next user, they just automatically get the new image, and no one’s work gets disrupted by a reboot.

Flexibility:
We can, in fact, define multiple “golden images”, each with a different suite of software on it. And rather than having to individually update each machine or each image, the system understands common packages — if we update the OS, then all images referring to that OS automatically get updated. Again, this leads to a great reduction in staff effort needed to support these more-standardized environments.

We have deployed SAP and Envisionware images on VDI, as well as some more customized images (e.g. Divinity-specific software).  For managers who don’t otherwise have access to SAP, please contact Core Services and we can get you set up to use the VDI-image with SAP installed.

Future Expansion:
We recently upgraded the storage system that is attached to the VDI-server, and with that, we are able to add even more VDI-terminals to our public computing environment. Over the next few months, we’ll be working with stakeholders to identify where those new systems might go.

As the original hardware is nearing it’s end-of-life, we will also be looking at a server upgrade near the end of this year. Of note: the server upgrade should provide an immediate “speed up” to all public computing terminals, without us having to touch any of those 50+ devices.