(Virtual) ICER 2021 Reflection

Kristin Stephens-Martinez
7 min readSep 25, 2021


ICER 2021 has come and gone, and it was my best online conference experience yet! Both in terms of how I handled it AND in how the organizers run it. I won’t give a detailed explanation of how ICER was run. Matthias Hauswirth, the site co-chair, made an excellent introduction video to the platform Clowdr. Not everything was perfect. But I think a lot of that was because I was new to Clowdr. After fiddling with my processes as the conference continued, I found things that worked. I’ve broken this blog post into socializing, useful processes/practices, my wishlist of changes, and highlights.

[This article is also posted on my personal blog so it is freely accessible.]


This online conference was the most social I’ve been to yet, and I loved it! I think part of it had to do with the benefit of lowering the number of tools/apps/sites that people are using and baking in the ability to socialize. SIGCSE didn’t have an easy way to socialize outside of the actual activities. ITiCSE had ohyay, but it was not in the platform they were using (Moodle). Clowdr had both the actual conference and the ability to socialize. Moreover, ICER lowered the barrier more by building social time into the program. These times were 20 minutes long, which I think was just right. Not too long to feel like you could go off and do something else for a while and leave the conference context entirely. Not too short that it doesn’t feel like a real break.

I also appreciated that the schedule was just 3 hours every day, with a 20 minute “open coffee” right before we started each day. Just like an in-person conference when everyone is slowly arriving and grabbing some refreshments! I was privileged to have the conference at a regular time of 12:40 pm — 4:00 pm, so it fit into my schedule very nicely. It also meant that I could check in on Clowdr before and after as I prepped for each day. It was interesting to see that Clowdr had very few people on it (you could kind of tell by looking at what pages people were on and willing to socialize in) before and after the official time frame.


The biggest thing I found helpful was opening multiple windows on different pages of the platform. This setup was especially handy during sessions.

During the lightning talk sessions, it was hard at first to track who was connected to which poster/paper/short video. This difficulty especially happened during the first lightning talk session as I scrambled to make sense of the new platform and get the most out of things. Eventually, I figured out I should have one tab open to stream the lightning talk session and another tab with the lightning talk’s program page to quickly find the project and open another tab if I wanted to follow up.

For talks, they streamed two videos, sent us to “break out rooms” to discuss the papers (I love that part of ICER), and then brought us back for Q/A. The videos were always available to us, but they streamed the videos at the beginning of the session, which I appreciated. It was like a “silent meeting” but as a talk! Because let’s all be honest, there is no way everyone will watch the videos in advance. Moreover, there are enough participants like that that it would be better to build things to level the playing field.

However, I actually handled the streaming part of the session a little differently. I had one window open to keep track of where everyone else was in the stream. Then I opened a window to that talk’s program page, which let me watch the video at a faster speed! I watched the video using this window, pausing to take notes and backtracking when I didn’t catch something. I usually could finish both talk’s videos around the same time the videos in the track were done streaming. So I could seamlessly join the queue to go to a round table discussion.

I love the round table discussions at ICER. It’s an opportunity to meet people and discuss papers in depth. My main critique is that the discussion time was only enough to talk about one paper, especially adding time for introductions. If I was the one to volunteer to lead a session, I often encouraged people to mention which paper they’d like to discuss after introducing themselves so that the group had a sense of which paper we’d focus on. On the other hand, time is an understandable constraint for this conference!

Change Wishlist: Discussion

There is so much I loved about how ICER was done, so I hope this section does not come off as overly critical. I honestly don’t think they could have done better, given the times and circumstances. What’s important is to ask what went well and what didn’t and how we can do better next time.

When I reflected on how things could have been better, my thoughts settled on the discussion part. I think the chat was great for synchronous, community-bonding interactions. The reaction emojis added to it by giving another way to participate and to remind everyone that they were not watching the video nor live stream by themselves. The questions sometimes got lost, and I think the “rules of engagement” on asking and getting questions answered could be improved.

The major change I think I would want to make would be to make the chat that was on each paper’s page more asynchronous, like a discussion forum. A chat is good for live, synchronous interactions. A discussion forum is better for asynchronous communication, which is more what I would be looking for when visiting a particular paper’s page.

My goal with this change is to enable the community to ask the authors questions and discuss the work. This goal does not need and in some ways should not be fulfilled only during the live, synchronous time when the collective attention of a large number of conference attendees is present. Therefore an interaction model like a discussion forum, which focuses more on asynchronous communication, would make more sense. Moreover, adding elements that enable productive discussion during the live, synchronous parts of the discussion forum would improve that experience.

So I propose that rather than a chat per paper, there is a discussion-like forum where attendees can ask a question or post a comment (both are necessary). Both post types should start a thread where others, and especially the authors, can comment. Bonus if, for question posts, the authors can mark a comment as the answer. During the live viewing of the video, both the chat and the discussion forum are available (perhaps as different tabs). If someone asks a question in the chat, the moderator could tell the asker to put it in the discussion forum or paste it in themselves. Then during the round table discussion, the discussion forum is available to the participants to write down their questions immediately or view other’s questions. During this time and the Q/A, participants could also vote on what questions they would like answered, and questions could be sorted by vote count. During the conference, the authors could answer the questions on their paper’s discussion forum. After the Q/A, they could simply say they answered the question during the Q/A if the recording is available.

I think this feature would enable extended engagement with a paper, encourage more discourse, and streamline the process of questions. However, sometimes it is hard to get people to engage in a discussion forum, so I think I’d call this change an experiment worth trying than a thing we definitely should do.

Highlights and Memories

I socialized a bunch and loved it! I met people in one-on-one, private, on-demand chats attached to the direct message chat. I also visited a bunch of different social rooms, including making a few of my own. I first made “Parenting + Conferencing,” a continuation of the text chat I asked for in last year’s Discord. Closer to the end of the conference, I made “Airport Carpool and “Airport Lobby/Line.” People preferred the carpool room to the lobby/line. One person even commented that they wouldn’t want to be stuck in an airport line after all. I love how our community doesn’t take ourselves too seriously, as evident by a “Cat Chat” social room that Amy Ko created.

I volunteered to be a Code of Conduct contact during the conference. I was not approached, which I hope means that nothing happened.

I’m grateful for only two tracks. It means less material to consume because so much stuff is fascinating!

I attended the ICER work in progress workshop, and it was helpful. I also enjoyed talking to people about my work and theirs. I got new perspectives I hadn’t even thought of. My main regret is it did make my August even crazier by taking up a weekend.


Overall, ICER was worth attending. I hope it will be hybrid going forward because travel outside of North America will be hard for me for the foreseeable future. But if it’s hybrid, I can at least virtually attend.

On the other hand, I’m tired! Even though it’s been a while since ICER. August was crazy. There was the SIGCSE deadline, the end of our CS+ undergrad research program (I had four students), ICER WiP, ICER, and the start of the semester. And the start of the semester hasn’t let up the crazy. Is there no way around it? If I only attended ICER when it’s in North America, that would at least reduce that workload.

What do you all think? Any tips on balancing workload? What did you think of ICER? Any thoughts on making virtual/hybrid conferences make sense?



Kristin Stephens-Martinez

Assistant Professor of the Practice in Computer Science at Duke University