Timestamps, contours and context are some pretty general terms that are often conflated or used somewhat interchangeably in the literature. This post will try to disambiguate the three concepts and distill each into their origin, intended purpose, and nuances.
Query-by-Synthesis (QBS) [@cheung2013optimizing] is a system developed to optimize database applications via static analysis. It is a very powerful system that, in the end, converts sections of code that would otherwise perform relational-query-like operations on the application server to a semantically equivalent SQL query that can be run on the DBMS.
A lot of developer work on database applications goes into making sure the interaction between the application layer and the database layer is efficient. In my experience in industry, this work is mostly done manually, and in an ad-hoc fashion. [@ramachandra2014database] give an account about some ways this can be done manually.
A General Fine-Grained Reduction Theory for Effect Handlers (Sieczkowski et al. 2023) presents a theory of effect handlers that can be used for term rewriting systems, and provides a few different abstract machines for effect handlers.
A significant contribution of this paper I believe is that it provides a top down small-step semantics for effect handlers, which is a bit unusual.
Refocusing in Reduction Semantics (Danvy and Nielsen Nov. 2004) presents a way to go from a reduction semantics to an refocused pre-abstract machine, and eventually to an abstract machine. The paper is a good read after being introduced to reduction semantics and abstract machines, and has a few simple walkthroughs of how to actually apply the theory.
There are (at least) three different "sensitivities" an abstract analysis can have:
This article briefly describes the difference between them.
Abstract interpretation is a very common technique for analyzing programs. This is a quick primer on the topic.
The core idea of an abstract interpretation is to "run the program", but instead of typical program states/values/etc (sometimes referred to collectively as configurations), we use some replacement that means the execution is decidable (the program always terminates).
Galois Transformers and Modular Abstract Interpreters (Darais et al. 2015) provides a way to do abstract interpretation in general, without having to specify a particular language or analysis. This paper is fairly mathematically dense, so I'll do my best to break down what I understood from it. This paper is also a good starting point if you want to understand Dr. Germane's paper on Full Control-Flow Sensitivity.
The Abstracting Abstract Machines (Van Horn and Might 2010) paper is frequently cited due to its presentation of a methodological process of abstracting a small-step abstract machine for analysis.
This is a Blog for the BYU Static Analysis Lab
Dr. Kimball Germane is our advisor at BYU. https://kimball.germane.net/
He has a lot of experience with static analysis, and especially with making control flow analysis practical for real world use.
In this blog post we will introduce static analysis from a few different perspectives.
Click (Read More) below to read the rest of the post.