JSE-40-01 deep review — September 29, 2026

The study guide maps all 30 canonical scope bullets, grouped 3/5/5/5/7/5 across six unweighted blocks. The separate practice kit's 8/20/18/21/21/12 percentages describe practice content, not canonical exam weights or item counts. This same-context AI review retains human review as pending.

Source evidence

The full exam/scope page, TestNow policy and course landing were read via the public web reader after direct failures. The objective monitor failed; the retained snapshot was compared manually, excluding the later holder profile from objective counts. Active JSE-40-01 remains 30 single/multiple-select questions, 40 minutes plus five for tutorial, 70%, English/Spanish and TestNow. Current booking was not inspected. Policy text has conflicting entry-level-specific and generic proctoring defaults; the associated JSA title also contains an unrelated Python certification copy error. Neither is silently converted into an exam requirement.

There are 31 registered sources, with direct health outcomes {'error': 4, 'ok': 25, 'blocked': 2}. HTTP200 does not establish a rendered curriculum or complete reading. Selected Mozilla reference sections support language and host semantics. The Chrome tutorial was read completely; the protocol's static URL was a redirect shell, so a fresh public browser rendered its current viewer and selected method contracts were read. Every source's actual reading boundary appears in the operation record.

Teaching and execution

All original teaching and 40 readiness answers were reviewed and improved. Repairs distinguish classic-script global properties, lexical bindings and modules; TDZ typeof from unresolved names; Number precision from exact BigInt text conversion; UTF16 units from code points; array results/mutation/shallow sharing/holes; precedence from operand order; inherited keys from own properties; parse failures from catchable parsing operations; and synchronous try scope from later callback failures.

Eight exact public JavaScript blocks ran in Chrome 154.0.8037.58. The original core workbook passed 114 checks in both Chrome and Node v24.18.1. The three integrated programs passed 109 scenario checks, including 19 actual dialog flows, timer completion/cancellation/timeout/restart and inventory boundaries. The introductory input block separately passed 16 actual prompt cases. Existing alias, loop and function examples ran, with four explicit calls supplied to the definitions-only function example. Separate classic/module probes confirm global-binding behavior. A bounded console.time/timeEnd probe produced the correct sum 5050 and an observed timing message; its duration is not a benchmark.

The order demo uses integer cents, a bounded decimal quantity grammar and confirmation before an in-memory append; no payment or remote order exists. Valid zero, cancel, empty text, invalid grammar/tier, declined confirmation and upper limits have separate results. The timer demo uses bounded settings and clears all identifiers on terminal paths. Observed callback counts and order do not promise exact timing, especially in busy/background pages. A separate original callback threw a recorded page error that the scheduling try did not catch. Inventory checks empty/first/last/missing records, duplicate codes, stock/quantity bounds, insufficient-stock preservation and alias-visible mutation.

An original debugger fixture actually paused, stepped over and exposed its local variable changing from 4 to 5. A second invocation changed the local to 2 before the faulty increment and returned 3. The unchanged source still returned 5 on a later call; correcting the source to decrement returned 3. These observations use Chrome DevTools Protocol, not visual DevTools clicks or Google's external demo. Exact original/fixed sources and observations are preserved. Local code execution is recorded as offline-only; public source browsing is separate. No installed runtime changed. Eight broader learner activities remain proposed.

Learning resources and remaining limits

OpenEDG's six-module course lists 40 hours; its broader Symbol material does not add an exam objective. The practice product is USD29, multiple-launch and practice-only, with 12-month redemption validity; current exact launch count and account UI were not verified. The Microsoft series page lacks its prior 51-part count, while the supporting repository is archived June 15, 2026 and assumes another programming language. These are selection boundaries, not proof of lesson quality.

Pluralsight's public headline now lists 28 courses/38 labs/85 hours. Selected Fundamentals 6h08m plus Debugging 1h35m total 463 minutes (7h43m); three selected guided labs total 100 + 40 + 64 = 204 minutes (3h24m). Only selected listing sections were read; no subscription lesson or lab was entered. OReilly/Udemy returned HTTP403 and old runtimes remain unverified. Cisco/YouTube yielded shells; no transcript/playback review occurred. Planning estimates are labeled.

The operation record contains objective mappings, source receipts, exact public code hashes and execution evidence. Full unit, repository, catalog and strict generated-site gates are recorded after completion. The result is reviewed-with-blockers for inaccessible contents, conflicting public copy, unverified account/booking details and human review. No account, enrollment, purchase, exam attempt or external message occurred.