College Board is in the final stages of rolling out changes to the AP® Computer Science A course and exam for 2025--2026. This is routine, in the sense that such revisions occur periodically to maintain the course alignment, but in this instance some of the changes are fairly substantial. Early announcements of the general plans have reliably generated much discussion and many questions, both in formal presentations and on social media.

Educators responsible for coordinating a staff to grade a large volume of exams or projects have learned, sometimes the hard way, the importance of writing good rubrics: they keep the graders on track, grading equitably, consistently, and quickly. But these benefits of a well-written rubric extend also to those of us without a grading staff, even to single classrooms of twenty or so students. Good rubrics will change your life!

This paper describes Unci, a unit-testing language with a clean and minimal interface suitable for introducing beginning programming students to the ideas of unit testing and test-driven development. We detail why CppUnit, a common C++-based unit-testing framework, is not well-suited for beginners, and present Unci and explain how it addresses the weaknesses of CppUnit. Finally, we present a comparison of CS2 student performance in the two systems, showing that moving from CppUnit to Unci resulted in an approximate doubling in the proportion of students able to write an effective test suite for a lab assignment.

Traditional processes for homework assignments are not always a good fit for the sorts of problems often seen in computer science classes. We present our experiences in implementing policies designed to encourage students to involve the instructor and fellow students in their learning process. By shifting to group assignments and permitting students a revision cycle, we improve student satisfaction and maintain or increase student outcomes while decreasing the instructor's grading load.

Selecting a first-day activity for CS1 is an important way to set the scene for an introduction to computer science. This paper presents the task of counting votes as an accessible vehicle for lessons on algorithms, data management, parallel processing, and other key computational ideas.

We demonstrate visually why doubling capacity is the better strategy when resizing arrays. The visual proof makes simple amortised analysis more accessible to a CS2 audience.

"Information and knowledge management" is the title and theme of a course outlined in CC2001 that was designed to cover a diverse range of core CS topics. In this paper we discuss our experiences implementing it: it succeeds at guaranteeing coverage, but also acts as a useful prerequisite for advanced electives, and as a survey course that engages student interest in unexpected topics.

In the field of empirical natural language processing, researchers constantly deal with large amounts of marked-up data; whether the markup is done by the researcher or someone else, human nature dictates that it will have errors in it. This paper will more fully characterise the problem and discuss whether and when (and how) to correct the errors. The discussion is illustrated with specific examples involving function tagging in the Penn treebank.

Best-first chart parsing utilises a figure of merit (FOM) to efficiently guide a parse by first attending to those edges judged better. In the past it has usually been static; this paper will show that with some extra information, a parser can compensate for FOM flaws which otherwise slow it down. Our results are faster than the prior best by a factor of 2.5; and the speedup is won with no significant decrease in parser accuracy.