Access 2015 Toronto Wednesday notes

McMaster supports Access!

For the first time ever, Access took place in my backyard. Glad that it finally cycled through Toronto. The local organizers did a great job, making everyone feel welcome and putting on a great event. Kudos to them for a unique locale for the opening reception.

Continuing a trend from recent years, I noted a decrease in the ‘technically-inclined cohort,’ with many familiar faces from recent years choosing to sit out this year’s version. Countering that, it was encouraging to see more people coming to Access who would likely never have considered doing so a few years back.

This year’s Access also had some wildly creative presentations, such as Tara Robertson’s use of a DJ during her presentation, Daniel Sifton’s RPG battle of scripted headless browsers, and Thomas Guignard’s hysterical Dr. Who / Night Vale performance piece on the failings of the ILS. Great talks, all, that highlight the creativity of the community and certainly raise the bar for those of us inclined to avoid the use of humour or other “non-serious” elements.

As always, my editorial comments and notes are in italics where I remembered to do so.

Public Computing and the Revolution
Opening Keynote & Dave Binkley Memorial Lecture
Molly Sauter, McGill

Started off by asking the crowd if they know that DDOS is; most hands went up, which pleased her. She explores in her book The Coming Swarm how networks–both cellular and the Internet–create spaces and possibilities for political action.

Showed some examples of oxymoronic “privately owned public spaces,” which are parks and open spaces maintained by corporations. The problem is that these spaces have regulations placed upon them by those corporations, not by law. They are private property.

Most of us do not ‘suffer’ under these strictures; however, others do, including skateboarders, the homeless, can/bottle collectors, etc. This is a problem. While these spaces are supposed to play by rules set by municipalities, they end up typically wholly co-opted by the needs of the corportation. For example, the space in Trump Tower that should be filled with marble benches is currently filled by a Trump Store selling his campaign’s t-shirts. The developers get concessions by promising these spaces then fail to follow through and cities do not enforce the agreements, which should be no surprise.

What does this have to do with the Internet? Well, as she put it, it’s a privately owned public space. She shared an anecdote from an earlier job about ‘content moderation,’ i.e.- removal of material from a social networking site that violated the terms of server. She pointed out that people bring with them their assumptions of freedom of speech and expression to the Internet, which isn’t a flawed assumption but runs straight into the terms of service established by corporations.

Drew a connection between real-world consent agreements–e.g.- if you enter this park while a movie is being filmed, you consent to have your image used in the production–and terms of service for online services. I didn’t quite follow this, because as she put it, if you don’t like the park notice about your face, you can go around or so elsewhere. Seems like that would apply online as well, since no one is compelled to use any particular service with a complex user agreement, such as Facebook.

Spoke about how bad Twitter is at rooting out and dealing with real harassment on their platform. She made it clear that she considers it the role of the service to monitor and deal with this, but is that how our civic order works in other environments? In a bricks-and-mortar world, the responsibility ultimately falls to law enforcement.

[Had to step out of this talk somewhat early to take an urgent work call.]

AODA: Enabling Innovation
Randy Oldham and Adam Doan, U of Guelph

Randy began with an overview of WCAG and AODA. The latter stipulates that WCAG applies as of January 1, 2016, but there are rather large loopholes. He offered several other tidbits from the legislation, summarizing his points by noting that the legislation is vague.

He noted that accessibility is often viewed as being inimical to innovation, but he feels that that is a false impression and set about countering that view. He showed specific examples, such as captioning for audiovisual content, noting the relevant WCAG standards, which are not entirely clear. In essence, such content must have a transcript and captions, which addresses visual and hearing impairments, as well as helps those with learning disabilities. Their workaround for doing this is using YouTube captions, alas, as many know the automated captions are not AODA-compliant as they are not 100% accurate. They do offer a useful basis, however, that can be edited to bring it into compliance. With YouTube, the screen reader is not available to screen readers, but by pasting the transcript to the description, it can be read by such readers.

But if you want to embed this in your Website, the content must be keyboard navigable per WCAG. He used a nested Tupperware metaphor to demonstrate that these paths are not entirely straightforward for individuals with visual impairments. Embedding is of course putting a Web page within a Website, so tabbing into that field with a screen reader ‘strands’ the user in that embedded content, which violates WCAG’s intent. Solution: take a screen capture, put it on your site, and link to YouTube. As he pointed out, this is an extremely low-tech but effective solution. That said, he noted that YouTube recently changed their embed code to get around the “trapped” user issue, but it’s of course contingent on their continued focus on the issue, while his solution is fairly bombproof.

He took an excursion into tables, demonstrating how they are not optimal for screen readers. The solution is to use a CMS and create alternate views, e.g.- for a calendar which in tabular form is not easy to parse with a screen reader. Also spoke about the transcription problem that arises with digital manuscript collections and suggested using crowdsourced transcriptions.

Special Snowflakes: Applying New UX Tools to Library Systems
Shelley Gullikson, Carleton; Susanna Galbraith, McMaster; Krista Godfrey, Memorial

Described the results of an ethnographic project that each carried out within their home institution. The three met at a library UX conference in the UK earlier this year and were eager to try out this methodology with their users. We know inherently that our students are uniquely constituted, but the extent to which the differences dictate their behaviour and use of library resources is what they sought to investigate.

As Krista noted, we all know by now that “we are not our patrons,” but ethnography is about learning the motivations of our students as they inhabit our spaces. Used an image of a designed path and an adjacent desire path to illustrate that there is often a gap between design and user need.

Observation as a method has its limitations; the act of observing can modify behaviour. They used three methods: interviews, mapping (i.e.- literally mapping spaces on paper, but also process mapping, e.g.- how they write a paper), and love/breakup letters. At Memorial, the mapping results were more fruitful than the letters.

At McMaster, students expressed appreciation for the institution and its resources, but offered some useful feedback on their experiences. It was noted that none of them indicated seeking out help from a person in the library. Most of the letter writers were proud to share their letters (n=11). She noted that the letter content often contradicted the mapping exercise (they blame tools, rather than the library’s collection).

Carleton’s results were similar, but of the letters she got, only one was a love letter, while the rest were breakups. Would be interesting to know the ratio at the other schools (I may have just missed this). The love letter was very positive about their Summon interface, but this is, after all, just an n of one.

At the end of the process, they compared notes, maps, and letters and searched for commonalities. One thread: librarians are not present as people who can provide help or instruction. Users would go “past the point of frustration” but still would not seek out help.

Building on their insights, they conducted a usability test in an attempt to direct people toward help, renaming the help function in relevant interfaces in all three libraries. They conjecture that an interface change isn’t the answer, perhaps, but that other factors discourage people from seeking help from library staff.

Ask What DH Can Do for You
Lisa Goddard, U of Victoria

I was looking forward to her talk anyway, but I nearly cheered when she used Samuel L. Jackson to poke fun at the libraries and DH discourse.

We have naturally complementary aspects that bring libraries and DH into the realm of partnerships, for example funding cycles. In libraries, we have a merciless one-year scope, while funded research often stretches over the years.

We also have complementary labour pools. In libraries, we have stable, longterm staff with deep expertise, but we can’t “reposition” people quickly. In DH, it’s much more fluid, made up of students who have skills and flexibility, but cycle through very quickly.

We also have common research goals, among them transcription. Images only get us so far, we need text, and OCR isn’t an option for many manuscripts. Enhanced description is another area where we overlap. We may call them finding aids, but they are essentially the application of human knowledge as a guide to collections, whatever we call them. Similar is what she called interpretation, i.e.- making our collections accessible (intellectually) to non-expert users such as the K-12 audience. Another is XML markup, which resonates with me based on my experience at DH conferences. Lots of DH scholars are really into this, and if we can interest scholars in our collections, we may be able to make progress (students?). Last, but not least, there is linked data. Some of the major LOD projects are taking place outside of libraries in the DH realm, so we should connect and collaborate.

The list went on, to discovery environments, which move beyond string searching to more semantically driven tools. Also, collaborative interpretation, which I took as a way of recasting crowdsourcing as a way to create annotated texts with large numbers of contributors.

She noted that the theme of the Digital Humanities 2017 conference (in Montreal) is access, an area where we are also active and engaged. Digital preservation is also on that list; humanists are concerned about these issues which is an area where we could help by offering our systems and expertise.

Key for UVic is that they have a grants librarian as a bridge builder for this kind of work.

Overall, we should find ways to strategically align ourselves with the digital humanities.

Changing the Way We Do Things
Peter Binkley, Tricia Jenkins, Kenton Good, U of Alberta; Alan Harnum, OCAD U

Kenton began by outlining how they did things in earlier times, which he defined as 1999-2013. They weren’t quite living up to their own standards in terms of being open and sharing. In their current era, they are moving more clearly in the direction of open data, open code, and open content.

What changes have they made? They embraced Github, went “all in” on open source solutions, put Creative Commons licenses on their Web content, and offer free OJS journal hosting. To be more responsive, they adopted Scrum and revised their server practice to achieve better integration, and try to foster better inter-team cooperation (between programmers, sys admins, metadata librarians, etc.). They also wanted to have sustainable solutions, so they look for community solutions rather than lone wolf products. They also have a digital asset management strategy (across the various nodes of IR, collections, etc.).

Peter delved into their working methods, mainly Scrum, which they implemented in Github, which means working in public. He outlined the basic elements of Scrum, which involve user stories, sprint kickoffs, etc. They use the daily standups, as well, which he noted has been the biggest change and has brought about the most change. As he concluded: “we’re doing a whole lot more and a lot better than we were a year ago.”

Tricia took up the issue of working as a team, noting that they started using Vagrant to achieve this. This allows them to share workspaces and share them around. In general, this is about creating a development environment that is similar to production, so that there is now less “mystery.” Similarly, they use Ansible to standardize how roles and services roll out. They take an hour on Fridays to do a Birds of a Feather session where they talk about the tools they use and share practices and ideas. It helps with onboarding, she noted.

Kenton closed the Alberta portion noting that they don’t have 100% adoption. Still some lone wolves about. For now, this is being used around their Hydra project, so it remains to be seen how it scales out to their many other projects. They had a few train wrecks with configuration management, too.

Alan noted that although he now works at OCAD, he spoke about work he did while still at Toronto Public Library. Started his talk in the early days of 2015 in the e-services group at TPL (which largely = Web services), which last redesigned their site in 2009.

Opted to go with a different solution when faced with the challenge of launching a new site. They wanted to have a flexible stack, i.e.- access to servers and tools when they wanted them. They also had a small budget for this.

He noted that moving out into the cloud requires cultural change, for instance, one must increase their risk tolerance. This requires management backing, of course. For technology, they used DigitalOcean for infrastructure (he spoke highly of it; cheaper and easier than S3/AWS), Ansible for automation, and Ruby for “fast product development.”

Noted that while organization culture matters, often the tools do as well and can themselves be drivers of change.

DIS-INTE-GRATE!
Thomas Guignard, Ontario Colleges Library Service

First Access presentation I’ve seen with sound effects. It got a round of applause. Like the creativity, particularly of the slides, even though the font was way too small for my aged eyes.

He riffed comically on the failings of the integrated library system, in particular the failings of the catalogue. Really mocked the fact, straight on, that we have two expense accounting systems: our ILS and our institutional accounting system. The lack of integration is what makes this necessary, and the lack of integration largely stems from the fact that these systems were developed long before such things were possible.

To truly appreciate the humour and style, one clearly needed to be a Dr. Who and/or Night Vale Radio podcast fan. I belong to the .02% minority of librarians who are not one or the other, so much of it went smack over my head. Still, I got the point of using humour to deliver a pointed criticism of the absurdity of our ILS.

Maker Culture on the Go: Portable Maker Kits at Vaughan Public Libraries
Alison Neal and Dina Stevens, Vaughan Public Library

Interestingly, they avoid putting out “crafting” items in their maker kits, so sewing, knitting, etc. are out of bounds. Other than that limitation, they have a fairly broad definition, ranging from vinyl cutters, to musical instruments, Makey Makeys, etc. They have multiple types of kits, such as a moviemaking kit with phone grips.

Some steps they took that seem wise include identifying “kit experts” and holding all-staff training sessions, encouraging their staff to practice and work with the kits before launching the service.

Kits are checked out by appointment, and they also use them in various promotional activities and events. They have had 2402 people attend scheduled programs, which does not include drop-in use or scheduled checkouts. They have added new kits based on their initial success.

Contest of Champions: Headless Battle!
Daniel Sifton, Vancouver Island U

Hard to take notes on lightning talks at the best of times. In this case, nearly impossible. Essentially an RPG-esque talk about the problems of using various Ruby and Python scripting options/libraries to achieve the same task (bad summary; my fault). Best part: his 10-yo son drew the characters.

Ensmartening the Web
Michel Castagné, U of Ottawa

Introduced the old and new of smart watches and then posed his first big question: can we send overdue book notices? Offered some interesting thoughts on the technology as it exists, as well as some design principles for apps to exploit the technology. Second big question: is anyone using these things? The sales figures are well below what analysts predicted.

In Our Own Words: Libraries, Makerspaces and Community Engagement
Lydia Zvyagintseva, U of Toronto Scarborough

Described a project that started with recording studios that the Edmonton Public Library opened as part of their makerspace array. Lydia and her collaborator decided to use them for an oral history project. Captured a number of stories, which are posted online. There is also an ‘information layer’ that maps the locations that are mentioned in the histories, which required doing some metadata work as well as connecting the mention with an actual timestamp. As she put it, in her dream world, these kinds of projects happen all over the place, and there’s a “platform and a place to preserve them,” as well as a mechanism to use and reuse them. She thinks both academic and public libraries can do this, but first one must define the community, as well as have recording equipment, a plan (ethics and metadata), a marketing strategy (promotion and outreach), and a platform.

Héritage: Metadata Challenges and Opportunities
Julienne Pascoe, Canadiana

They have a metadata problem: many, many objects with poor metadata, which makes access a challenge. Most of their information is at the collection level; item-level data is sparse. Also, many documents are handwritten, so OCR isn’t an option for gathering information and creating access points. Much of it exists on film, as well.

Creating this metadata in house for 60 million images is costly, to say the least. Not surprisingly, one option they have considered is crowdsourcing. Another option would be to use transcription services to capture form data, which then via scripts can be massaged into structured metadata. The third option is linked open data.

Linked Open Data: the Value-Add to a Postcard Collection
Robert Warren, Carleton; Sharon Farnel, U of Alberta

Showed a graphic of the place names that had red dots for places represented on Alberta’s prairie postcard collection. Interestingly, some went beyond the region, which could be either bad data or some outlier postcards in the collection.

Rob pointed out that people make mistakes, but the technology we have now can take these instances and improve the data. One interesting way he expressed the value of linked data is that one “inherits the work that others have done” by virtue of linking to it.