PCEP-30-02 deep review — September 29, 2026

The study guide now maps 84 listed syllabus bullets across 15 numbered objectives, with counts 27/9/25/23 under the four weighted blocks. The active baseline remains PCEP-30-02. The public credential table still calls PCEP-30-03 in development with its announced Q3 2026 target; the date alone is not evidence of release. This is an AI review in the same context, with human review pending.

Evidence and source reconciliation

All original guide text was read. The current HTML syllabus, credential page and testing policies were read through the web reader after direct requests timed out. All five pages of extracted syllabus PDF text were read; no local PDF byte hash or visual-table review is claimed. Manual comparison preserves the existing snapshots, without claiming a successful automated objective-monitor comparison.

The canonical weights remain 18/29/25/28, with 7/8/7/8 items. The HTML block 4 introduction says seven objectives/19 sub-objectives, but its actual list has four numbered objectives/23 bullets. The PDF agrees with the actual list. The MQC narrative groups topics differently. The official practice store still uses 18/28/26/28 and gives a different practice redemption role than the credential page. Those inconsistencies are disclosed; no hidden objectives, averaged weights, raw-question pass count or verified account workflow is invented.

The source set contains 25 registered entries: direct health outcomes are {'error': 5, 'ok': 18, 'blocked': 2}. HTTP success is not equivalent to a rendered lesson or a full-document reading. Versioned Python reference pages displayed 3.13.15, the general tutorial 3.14.7, and execution used existing CPython 3.13.14. Selected source sections and reading limits are recorded in the operation evidence. No runtime installation or upgrade occurred.

Teaching and executed examples

The guide repairs blanket claims about misspellings, Boolean results and string methods. It explains signed floor/modulo versus truncation, power grouping, short-circuit operands, loop exits, copy depth, repeated matrix-row aliases, tuple-contained mutables, dynamic dictionary views, mutable defaults, local/global names, generator timing and exception propagation. PEP 8 style recommendations are distinguished from parsing requirements. The existing 40 answers were retained and improved; they were not newly invented as exam items.

All four public Python blocks executed successfully with 136 checks: 82 named checks in the trace workbook and 54 additional console/scenario checks. The original console block used real input with simulated stdin and captured prompts. Normal, zero, negative, fractional, empty and invalid cases were checked. The existing loop prints 9. The complete original capstone program prints the quote, conversion recovery, score summaries, isolated score adjustment, zero-versus-missing search and low-stock list shown in the guide.

Shipping boundary tests cover both zones around 1000/5000 grams and the 1/10000 endpoints, plus invalid values. Score checks distinguish missing from zero, the 70 cutoff, type/range failures and copy isolation. Inventory checks cover empty/first/last/missing searches, explicit duplicate rejection, negative/fractional quantity rejection and generator threshold behavior. Rates and decision rules are fictional. Small recursion assumes integers and remains subject to depth limits. No production application or exam platform was tested; eight broader learner activities remain proposed.

Learning resources

The complete Edube public landing lists 42 hours and four sections; lessons were not entered. The public practice listing advertises five launches per test and a 12-month voucher redemption window; paid questions were not read. Microsoft lists four modules and no captured total duration, so its old 3h12m estimate is qualified. Pluralsight lists 17 courses/25 labs/47 hours; Foundations 161 minutes, Data Structures 116 and Functions/Modules 98 sum to 375 minutes/6h15m. These are public catalog facts, not lesson-quality or complete exam-coverage findings.

OReilly and Udemy returned HTTP 403; prior durations are explicitly historical. Cisco returned an application shell. YouTube only supplied title-level text and no playback. Harvard's landing was read through the web reader and confirms its ten-week structure and first four topic headings; lectures/problems were not completed. Planning estimates are labeled as author judgment.

Validation and remaining limits

The operation record records objective mappings, direct fetch receipts, selected-source boundaries, exact public code and execution evidence. Required repository, unit and site gates are recorded there after completion. This review is reviewed-with-blockers for inaccessible/paid contents, unresolved current vendor-page discrepancies, future-version/booking verification and human review. No account, cart, checkout, enrollment, exam attempt, external message or infrastructure action occurred.