18 Months of SIGCHI Operations
In January of 2019, I stepped into the volunteer role of ACM SIGCHI VP of Operations. I must confess, I didn’t know what operations did exactly but I was told it was to run the technologies for the SIG and that my 15+ years of industry research experience would help transform the volunteer role. Ops has a sub-committee with some pretty clear roles: run the listserv, manage website content, design the conference schedule apps, and upload to the YouTube channel. Beyond these tactical roles, the vision and goal of what Operations should be was an open field.
SIGCHI is in an interesting position where we archive and consume knowledge and we don’t always consume where we archive. Beyond our papers, there are conference websites, session programs, 1000s of photos taken by SVs, membership surveys, and more. While we can rely on the ACM Digital Library for some of these things, we have to rely on other platforms too. As we set forth, I kept remembering to never trust a corporation to do a library’s job; we can’t just rely on a commercial platform alone. For operations overall, I wanted to scale our organizational archiving efforts, optimize the publication pipeline, and reduce the tech debt we carried.
A New Beginning
The first thing I noticed were the email requests flowing in. Lets say, someone has an issue with the website. Was it filed before? If so was it fixed and how? Did anyone else have this issue? Can we follow up? A lot of organizational stuff was just missing and the first set of issues that I saw related to font-sizes on https://sigchi.org; there was no way to really investigate.
Two things were clear. First, we needed some organization to track and triage problems because I had no idea how many outstanding issues were on the table. Second, we had nobody who could help with accessibility on the sub-committee. Lucky, we had a GitHub account in place which we could use for filing issues, and there was a volunteer (Harsh Patel) in the pool from the last call who joined ops and had access experience. Immediately we set up an email that takes a request and files a ticket to our GitHub. I moved some people from SIGCHI EC Communications and Membership team members onto the GitHub repo, so issues related to content could be assigned to them. Other Ops related issues could be similarly tasked.
Here’s a real example of how it works. Somebody emails us or uses the “Feedback” mechanism on https://sigchi.org/ asking for conference totals to appear on the graphs on the website. A ticket is made in Github which I actually assigned that to myself. There was a discussion with some of the ops team on Github. I then modified the code, pushed it live, and closed the ticket. This one edit took me months to figure out, but that’s the benefit of having it in the system: we can see a request, discuss with the team, and most importantly make sure we fix, close it, and follow up with the requester telling them the conference history page now has the totals displayed.
Learnings From CHI 2019
Then CHI 2019 happened, and the iOS SIGCHI App was wrought with a few major usability bugs. Fortunately, the ‘Feedback’ button in the app was tied to another issue-tracking system with our contracted engineering team. They actually fixed the bug immediately, but I was hesitant to launch a hot patch live at a large conference. Along with the bug came many requests: 1. Data-caching didn’t work for some, 2. People wanted to download their data, 3. They wanted links to DOIs or Videos, 4. They wanted calendar feeds, and 5. People wanted to use the app without login.
In the CHI 2019 town hall, I mentioned I was new here, and we were on it. Despite that, I discovered in this role you get personal hate mail from young or established members alike. Fortunately, we got more positive comments from people reaching out. A few times people stopped me in the corridor to say thanks and suggest an improvement. Following that CHI, members from the accessibility group spoke to the EC, constructively, about the venue and software issues which we took as immediate action.
One Web App to Replace Two Native Apps
Looking at a set of mobile app bugs and having no documentation on how accessibility checks were performed in the past on our software (both bespoke and contracted), we mapped a few options. There was also an android version of our native apps, so any feature would need to be built in two places. First, I contacted PCS and requested a regular practice of running access checks. We also tracked down the bug reported to me at CHI. One major screen reader error in the ‘Final Submission’ flow of PCS actually was happening with the proceedings vendor. I am working with the VP of Publications to remove the vendor from our workflows as they often strip other access from PDFs (more on that below).
Roman Rädle on the Ops committee suggested ditching all the native mobile apps for a progressive web app. A web app would work across devices and platforms, as well as, can be audited more easily for accessibility. This started as an experiment in June 2019 and was beta launched for November’s CSCW 2019. This web-based program is an archive for the conference program, and it means chairs no longer have to try to make one themselves. Having a web-based application allows us to use modern features like local caching and easily add modules like the SIGACCESS Conference FAQ so they can be found by attendees easily. This also lets the EC audit which conference chairs are in compliance. The PWA also gets regular automated accessibility checks; the release notes show more detail on everything that gets updated. Both PCS and the PWA are undergoing a full access audit this summer (not just an automated checker). This was done with the instruction to feature freeze until they pass within a minimum of A-grade compliance. This includes stopping all requested PCS and PWA features from conference chairs (if you’re a chair reading this: surprise!); nothing new gets added until we make the grade.
Learning From A Global Pandemic
COVID — heard of it? — showed up next. Conferences had to go virtual. The Human-Robot Interaction conference was the first to approach us and ask for video support for the paper talks. Here the PWA came to our rescue as it’s an extensible web app. We were able to work with conference chairs, PCS, our video team, and the ACM to get a data flow. For conferences, DOIs and video links are now in our web program. This was a massive job for the Ops Video Team. We documented most of it, including instructions on how to record and caption video. We are working on adding live stream and Q&A links and rethinking how to better schedule virtual events. We’ve also worked with the ACM on a plan for virtual conferences in which we are leading an effort for sustainability and accessibility well beyond our SIG.
There’s a good chance you’ve heard of TAPS: The ACM Publishing System. TAPS makes beautiful, accessible proceedings in PDF and HTML5 and circumvents vendors causing issues (see above). Together with Julie Williamson (VP Publications) we created an overall workflow to link PCS, TAPS, videos, and our PWA system. ACM IMX 2020 used TAPS end to end, and it simplified everything from author workload to chair duties. It wasn’t without bumps in the road but was easier than the conventional route. The workflow helps a lot when everything from publishing to running the conference is happening over distance.
New Infrastructure for Sustainable Growth
In hand with documenting and communicating our video plan, we worked with Neha Kumar (SIGCHI VP-at-Large) to reach our members through a new Official SIGCHI Medium Publication. The overall effort has been very successful with about 10k views in the past 3 months. Remember that line about trusting a corporation? Ops helped design the process where each post is marked outside the Medium paywall, each post is mirrored on our own (low traffic) Wordpress blog for archiving, and each Medium post links back to our blog as the canonical source. I borrowed this process from the Ops Video Team, where we don’t monetize our official YouTube channel and we cross post archival content to the ACM DL.
And we had some tech debt issues. We removed the native apps, which saved a lot of engineering. I worked on consolidating a lot of systems to save costs and add organization. This included a shared secured password vault (with about 25 passwords in it that I recovered or discovered), an enterprise SurveyMonkey account (to ensure data privacy and GDPR compliance), and some other utilities we are adding to support our global chapters. We also sunset the communities portion of the website as it was, in effect, a costly email list engine. The ACM provides a listserv, and we will use SurveyMonkey for elections where needed.
Finally, I want to thank Minsuk, Carla, David, Kash, Scooter, Max, Harsh, Lindsay, & Madeline on the SIGCHI OPS team, Pejman Interim VP Communications, Julie VP Publications, Neha VP-at-Large, the people at the ACM who often look toward SIGCHI for insights, and, most importantly, anyone who filed a bug, had an issue, or dealt with me getting requirements to improve things. We still have a ways to go but happy to see a lot of good progress being made with the help of our community and volunteers. The priorities we set forth remain: systems need to be accessible to all, systems need to have a global reach, systems need to scale to our users and the ACM, and systems need to reduce tech debt and feature creep to ensure sustainability. We are making progress towards that future. Operations for SIGCHI has plenty on the road ahead while we look toward hosting hybrid virtual/IRL conferences, archiving more content (like conference programs) in our digital library, and empowering our community with tools.
