About (non-)mandatory attendance at PODC 2025

PODC 2025, the flagship conference of (the theory of) distributed computing will take place in Huatulco, a seaside resort in southern Mexico.

This location choice has faced criticism due to its high carbon footprint. For instance, the carbon-equivalent emissions associated with travel will be more than double those for a conference in Paris. (See the report we wrote with Tijn de Vos. At the time we only knew we only knew that the conference would be in Mexico, so we assumed Mexico City.) Additionally, it is unclear whether the conference will significantly benefit the local research community, given that only a small number of researchers in Mexico specialize in this area. No discussions, as far as I know, have taken place about organizing events to massively engage local students, such as satellite workshops. Furthermore, the location was not initially suggested by local researchers but by individuals from other countries who have fond memories of previous conferences in Mexico.

Anyway, this choice is made. Now, a way to mitigate the carbon impact is to relax mandatory in-person attendance (which I blogged about earlier). I proposed to the steering committee that the call for papers explicitly allow the program committee chair to permit remote presentations on a case-by-case basis. While this suggestion represents only a small step—since the chair could deny all requests—it would have been a move in the right direction. The recently published call for papers actually does not mention mandatory attendance at all. While most people likely expect papers to be presented on-site, there is no formal basis for removing a paper from the proceedings if the author opts to present online.

I therefore encourage those hesitant to travel to Mexico to request online presentations.

One might ask why one would submit to a conference if not to participate in that conference. I agree it makes little sense a priori, but our academic system makes it so that having a PODC paper can change your career. Some countries, for recruiting academics, simply compute a weighted average of the publications, and PODC usually has a high weight, even compared to DISC or Distributed Computing (respectively, the second flagship conference of the area and the main journal). Since it is impossible to change this in a few month, we have to adapt to it.

A second issue is that online presentation can be less engaging than on-site talks, because of technical issues, less dynamic talks, and lack of interaction at coffee breaks. My bet is that if one does things right, an online talk can overall be better than an average on-site talk. I would urge anyone planning to present online to make big efforts to make the experience of the participants as good as possible. In particular, having a great talk, prepared well in advance, rehearsed with diverse audiences, and incorporating feedback. Also having recorded versions of different formats (teaser, 20 minutes, extended version) would be great. Finally, being very responsive with questions: being connected to the PODC-DISC zulip channel, promptly answering questions via email, offering video calls to discuss specific aspects. Of course this is more work than sketching half-baked slides a few hours before your talk, but think about all the time you will not spend in planes and airports! (From Lyon, where I am, the round trip is almost 40 hours, see skyscanner)

As for my plans, I decided not to submit to PODC this year (being in the PC, and having nothing that would be ready without rushing). However, I am happy to assist anyone making the decision to present online both on the diplomatic side and for preparing a great presentation.