Many people are first exposed to code through web development, yet little is known about the barriers beginners face in these formative experiences. In this article, we describe a study of undergraduate students enrolled in an introductory web development course taken by both computing majors and general education students. Using data collected during the initial weeks of the course, we investigate the nature of the syntax errors they make when learning HTML and CSS, and how they resolve them. This is accomplished through the deployment of openHTML, a lightweight web-based code editor that logs user activity. Our analysis reveals that nearly all students made syntax errors that remained unresolved in their assessments, and that these errors continued weeks into the course. Approximately 20% of these errors related to the relatively complex system of rules that dictates "when" it is valid for HTML elements to be nested in one another. On the other hand, 35% of errors related to the relatively simple tag syntax determining "how" HTML elements are nested. We also find that validation played a key role in resolving errors: While the majority of unresolved errors were present in untested code, nearly all of the errors that were detected through validation were eventually corrected. We conclude with a discussion of our findings and their implications for computing education.