Patch notes
v1.0.1 - Post-launch fixes & question quality
The first update since launch. A batch of live-reported bug fixes alongside a curriculum-wide content-quality pass: questions no longer telegraph their answers, repeat encounters stop rewarding position memory, and lesson rewards now care how accurately you got there, without ever blocking progress.
🚀 New
- Cores onboarding slide explaining what Cores are for (Epic/Legendary shop tier) and how to earn them.
- Lesson crystals now scale with accuracy. Finishing a Logic First lesson at 80%+ accuracy earns the full crystal award; 60–79% earns a bit less, and below 60% earns half. Retries are still free and nothing ever blocks you from finishing. Perfect lessons keep their existing bonus on top.
- Accuracy on the completion card, color-coded the same way the Insights chapter report is.
- Practice now covers the whole curriculum. The 15 Lists, Dicts, and Strings lessons (Chapters 8–10) each gained four practice puzzles.
- Every capstone stage now has a hint (1 crystal for free users, free for Pro).
🐛 Fixed
- Practice and Capstone now correctly stop you at 0 embers, matching how the regular lesson flow already worked.
- Sign In’s “Continue with Google” now tells you when it creates a new account, instead of silently doing so.
- The streak strip now shows the real calendar week (Sunday through Saturday) instead of a rolling window that could start on an arbitrary day.
- Profile’s stat tile now correctly counts Logic First chapters instead of a legacy number most learners never see.
- The save-progress prompt works again - it’s now tied to finishing your first lesson instead of a milestone that was effectively unreachable.
- Two Big-O fill-in questions now show their blank marker correctly.
- A trace-output question no longer marks you wrong for knowing
98.0is what Python actually prints. - Several “what does this print?” questions now show Python’s real quote style in their options.
🔧 Tuning
- Distractors rebalanced across the whole question bank so the correct answer is no longer identifiable just by being the longest option.
- Multiple-choice options now reshuffle on every attempt, so replays and spaced-review encounters test recall instead of position memory.