I want to learn this properly over the next three months. Course first, or project first?

Option A is the better fit for learning properly over three months because the explicit goal calls for a reliable foundation before independent problem-solving. Four weekends of structured material first provide a mental map of the domain, reduce unknown unknowns, fill baseline gaps, and prevent the fragmented, bad-habit-prone knowledge that often comes from purely reactive learning. The remaining time is then spent building, so this is not tutorial hell: most of the three months is still active practice, now guided by a framework and followed by retrieval and application that consolidate understanding. For a true beginner or foundation-heavy field, project-first is often frustrating and inefficient because the learner lacks the vocabulary, architecture, and fluency to navigate documentation or troubleshoot effectively; documentation lookup becomes ad hoc, gaps stay invisible until they cause failures, and shipping can crowd out proper learning. Option B has genuine strengths: it can be motivating, relevant, and efficient just-in-time learning for someone who already has partial foundations or a specific deliverable. But for the stated aim of learning properly in a new area, its reactive patching risks shallow understanding and stalled progress. The main caveat is that a course cannot literally fill all obvious gaps, since real gaps surface while building; however, course-first followed by project work is still the more concrete, time-boxed, and balanced path, with project-first remaining defensible mainly for experienced learners in hands-on crafts.