I wanted to write something in the blog again. Unfortunately, the website no longer builds with my original theme. After spending some time to debug and fix the theme, I decided to give up and asked Claude to migrate my blog to a new theme.

Fortunate enough, the advancement of AI technology means that I can watch my blog changing theme automagically from a distance; and now my blog has a new theme and I can write about stuff again.

A friend of mine found out something interesting with the date command in coreutils:

Are these different for you too?

$ date -d 'a week ago'
$ date -d '1 week ago'

Indeed, after trying it on my terminal, the two dates differ by one hour.

$ date -d 'a week ago'
Mon  6 Nov 21:51:16 GMT 2023
$ date -d '1 week ago'
Mon  6 Nov 22:51:18 GMT 2023

However, the results are different for my friend:

Conferences have different submission requirements for papers. Submissions are usually subject to a page limit, but supplementary materials (or appendices) are permitted. Here is a guide to produce various versions of a paper in LaTeX.

Short and Long Versions

The easiest way is probably to produce a short version that satisfies the page limit, and a long version containing omitted details (usually in an appendix). This way has the benefit that one can simply read the long version, as it usually contains the short version in its entirely.

My colleague Martin Vassor and I worked on a paper with title Neo-Classical Logic for SIGBOVIK 2022¹.

You can read the paper here (with reviews).

In this April Fool’s paper, we explore a possible logic that encapsulates seemingly contradictory observations in the real life. We investigate how such contradictions can be represented in a logic, and show its decidability. Moreover, we prove standard theorems such as “Law of Excluded Middle”, but also novel theorems such as “Law of Contradiction”.

Many researchers write paper in $\LaTeX$ because it provides a convenient macro system that makes typesetting maths equations easy (or difficult, when it doesn’t work).

One way to write papers in $\LaTeX$ is to use as few “advanced features” as possible, for a vanilla experience: no fancy packages, just use the minimal setup. Gradually, as the paper starts to grow long, you begin to think of adding some macros for things that are too long to type, for example:

I presented my paper Statically Verified Refinements for Multiparty Protocols at OOPSLA 2020 this year, held virtually. This paper is co-authored by Francisco Ferreira, Raymond Hu, Rumyana Neykova, and Nobuko Yoshida.

You can find the paper here from the publisher (Open Access). The full version of the paper is available on arXiv at 2009.06541.

The video of the talk is available on YouTube (full version, 15 min), and a short version (5 min) is also available.

Our paper has an accompanying artifact (as a Docker image). Source codes are available on GitHub.