Show HN: Every problem and solution in Beyond Cracking the Coding Interview
Why the problems are free / goals of the project
- Creators say the main goals are:
- Get people to read the book (large portions are free).
- Drive usage of interviewing.io.
- They argue practice problems themselves aren’t a competitive advantage; there are already many free ones.
- They dislike paywalls and see high‑quality free content as the best marketing to engineers.
Perceived value of the book and platform
- Several commenters praise the book for:
- Practical guidance on resumes, outreach, and breaking into companies.
- A more structured way of thinking about problems (e.g., “boundary thinking,” triggers).
- The AI interviewer is viewed as useful for simulating real interviews, though some users would prefer an option to just submit code and see if it passes.
Debate: usefulness and fairness of LeetCode-style interviews
Pro‑side:
- Coding tests are seen as essential to filter out candidates who simply cannot code, even with strong‑looking resumes.
- “Easy/medium” questions are defended as checking basic competence and foundational CS knowledge (arrays vs linked lists, complexity, etc.).
- Some claim these interviews are effective in aggregate, citing the success of large tech firms and arguing they have low false‑positive rates.
- Others say LeetCode‑type skills apply more than critics admit, especially beyond simple CRUD work.
Critical side:
- Many argue real jobs rarely need “fancy” algorithms; production‑quality code, design, and communication matter far more.
- They see an arms race: as candidates train, companies raise difficulty, pushing interviews toward memorization and grind.
- Strong concern about live‑coding anxiety and performance under observation, especially for senior/principal roles where architecture and leadership are more relevant.
- Suggestions include simpler tasks, collaborative problem‑solving, pair programming, PR/code review, or small take‑homes, with candidates choosing the format.
Are LeetCode-style interviews dying?
- One view: signal is degrading due to AI and cheating; these puzzles may fade.
- Counterview (including from the author): companies are conservative; DS&A interviews aren’t going away, though verbatim LeetCode questions should. The focus should shift toward teaching and evaluating how candidates think.
Software vs other disciplines, credentials, and “grind”
- Some compare software unfavorably to other engineering fields that have licensing bodies and standardized credentials, arguing this forces companies to over‑screen.
- Others note that most engineering disciplines can also be self‑taught; what differs is tooling cost and mentoring pathways.
- There is disagreement over whether willingness to “grind” LeetCode is an important, job‑relevant signal or an arbitrary hoop.
Interview as performance; analogies to other fields
- Commenters compare live coding to:
- Auditions for actors, “staging” for chefs, hands‑on tests for trades, and case studies for managers/analysts.
- Others argue the best test of engineering is real work, not high‑pressure performance in an artificial setting; the NFL combine is used as an analogy for imperfect proxies.
Meta: Show HN etiquette and tone
- A substantial subthread debates whether harsh criticism of the premise (technical interview prep itself) is appropriate in a Show HN.
- Participants cite HN guidelines: avoid fulmination, shallow dismissals, and generic tangents; be substantive, measured, and civil when critiquing someone’s work.
Other points
- Privacy: the site uses Clearbit to enrich emails with names, but interviews are anonymous unless both sides opt in.
- A user in India hits a country restriction; the team calls this unintended and routes them to support.
- One commenter notes that “interesting, contained” problems are a finite personal resource: once you’ve seen a solution, you can’t unsee it, so you lose the chance to solve it fresh in the future.