My talk "How to Compare Fuzzers" at SAPLING'24 yesterday is available on line here #SoftwareEnginnering #MutationTesting #Fuzzing

I will be giving a talk at SAPLING'24 at ANU Canberra on comparing fuzzers. If you are at Canberra, please do drop in, and say Hi.

How is the participation in Activitpub vs Bluesky these days? Has Bluesky [actually won]? Any one has an insight into what is the status now?

@prophet asked this on X.

"i'm going to regret asking, but does anyone have a *rigorous* definition of strong/weak typing as orthogonal to static/dynamic? specifically, if your explanation contains the term "type conversions", please explain what that means, especially in dynamically typed languages"

I wonder if a software testers perspective can help here. For me, a strong test suite is the one that finds most bugs. Assuming that the most significant source of error is simply mistyping the intended code while programming, we can leverage mutation analysis to estimate the strength of a test suite.  That is, the strongest test suite is the one that identifies a difference between the original program, and the most number of mutants (one token variants), and the weakest one is the one that is unable to distinguish any variant from the original.

Transplanting that to type systems, I wonder if I can say: The strongest type is the type that eliminates (makes it impossible to represent) the most number of variants to the original program. 

I don't know how much I can trust this, but this sure was fun to read!
Claude interacting with ChatGPT on its own volition, without being prompted!

The chat log from here.

Mentioned by this reddit user.

This figure is so accurate! (Originally from here by Matt)

There was site that once showed the change in participation in platforms such as Mastodon and Bluesky in response to Twitter/X's change in management. Does anyone know where to find this? One of my students is investigating ActivityPub, and it would be nice to be able to show the change in participation in Fediverse over time.

Notebook LLM is awesome, but very much a work in progress. For example, while I can provide multiple sources, there is no way to preview PDFs or even download them once I have added one of these as source material.  Secondly, the source guide extraction of text leaves a lot to be desired. While the RAG is very useful, when I click on the citations, I wish it took me to the actual document rather than an extracted summary of it. I also wish that there was a way to query individual documents separately.

Danushka Liyanage, postdoctoral researcher at University of Sydney Software Engineering group at his convocation.

Congratulations Dr. Liyanage!

Google has just published their scholar ranking for publication venues here (The link is for software systems).