Selenium IDE, Record/Playback, and Migration to Maintainable Code: Diagnostics, Failure Modes, and Production Practices

1. Failure taxonomy: preserve the first failure

Use the course diagnostic sequence unchanged: preserve first-failure evidence → confirm Selenium/IDE/binding/browser/driver/Grid versions → confirm target/environment/test data → inspect session/capabilities/context → inspect locator/element/synchronization state → inspect AUT/browser/network evidence → inspect Grid/CI/resource state if remote → apply the least destructive correction → rerun the smallest controlled scenario.

The recording adds one evidence source: the exact command, target, value, and recorded project version at the failed step.

2. Intentionally broken locator: interpret before repair

Run the broken WebDriver translation against /v2 only after the loopback fixture is healthy:

# Deliberately brittle: it preserves a generated-looking recorder locator.
from selenium import webdriver
from selenium.webdriver.common.by import By

browser = webdriver.Chrome()
try:
    browser.get("http://127.0.0.1:8826/v2")
    browser.find_element(By.ID, "display-name").send_keys("Ada")
    browser.find_element(By.ID, "save-profile-427").click()  # v2 changed this id
finally:
    browser.quit()

Expected symptom: NoSuchElementException (or an equivalent IDE “target not found” message) at id=save-profile-427. Evidence shows v2 contains data-testid="save-profile" and a different generated-looking ID. Corrective action: change the locator contract. A retry, browser restart, or giant timeout would preserve the wrong selector and obscure the cause.

3. Fixed waits and transient assertions

A recording may include a pause because the recorder cannot infer what “ready” means. Do not normalize that pause into production code. The fixture status is observable, so the migrated test waits for saved: or error:. This keeps the timeout bounded while allowing fast runs to proceed immediately.

If a recorded assertion checks saving, ask whether that transient state is actually the scenario contract. Assertions on intermediate states can create artificial flakes even when the product is correct.

4. Secrets, profiles, and captured identity

IDE recordings can persist typed values. Exported code can persist them again. Screenshots and project files can therefore become credential/PII artifacts. Use synthetic values in this chapter. In enterprise environments, obtain dedicated test tenants/accounts and secrets from the approved CI secret store after migration; do not record MFA codes, anti-abuse workarounds, production SSO sessions, or personal browser profiles.

5. Tooling limitation is not automatically a Selenium failure

Older Selenium IDE material commonly assumes a browser-extension workflow. The current SeleniumHQ repository describes an Electron application and npm/release installations. Organizations can also restrict extensions, downloads, native apps, clipboard access, or browser profiles. If recording cannot attach or an export option is absent, first verify the installed IDE build and organizational policy.

Similarly, do not assume every command/plugin/export from historical documentation exists unchanged in the installed build. Treat missing parity as a tooling/version constraint. The mandatory manual migration path remains available because WebDriver code is the durable interface for the course.

6. Duplicated command lists create blast radius

Suppose 25 recordings contain id=save-profile-427. One markup change produces 25 edits. The migrated design uses one ProfilePage.SAVE selector; the same change becomes one controlled edit plus targeted regression verification. This is the maintenance-cost argument for Page Objects/components—not a claim that every selector must be globally centralized.

7. Failure evidence helper after migration

Once the scenario is code, preserve the first failure in a unique directory before teardown. Do not let a later rerun overwrite it.

from pathlib import Path
from datetime import datetime, timezone
import json


def preserve_failure(driver, test_id, exc):
    stamp = datetime.now(timezone.utc).strftime("%Y%m%dT%H%M%S.%fZ")
    out = Path("evidence") / f"{test_id}-{stamp}"
    out.mkdir(parents=True, exist_ok=False)
    driver.save_screenshot(str(out / "failure.png"))
    (out / "state.json").write_text(json.dumps({
        "exception": type(exc).__name__,
        "message": str(exc),
        "session": driver.session_id,
        "url": driver.current_url,
        "title": driver.title,
        "browser": driver.capabilities.get("browserName"),
        "browserVersion": driver.capabilities.get("browserVersion"),
    }, indent=2), encoding="utf-8")
    return out

Redact secrets before writing logs or page fragments. Screenshots cannot always be safely text-redacted after capture, so prevent real sensitive values from appearing in the disposable AUT in the first place.

8. Production troubleshooting boundaries

Do not respond to a flaky recorded test with blanket playback retries, long pauses, JavaScript clicks, TLS disablement, browser/Grid restarts, or a switch to production for “more realistic” reproduction. Isolate the smallest failing layer. Measure runner/browser/Grid performance only when evidence shows resource pressure is causal; recorder slowness, browser startup, Grid queueing, AUT latency, and artifact IO are different costs.

Knowledge checks

Answer from the operating model, then reveal the explanation.

Summary and next bridge

• Classify the first failure before changing timing or retries.
• Locator failures require locator evidence and causal repair.
• Fixed waits are not a migration strategy.
• IDE projects/exports can carry secrets and must use synthetic data.
• Tooling/policy/export limitations are separate from WebDriver semantics.

Lesson 5 combines the complete record → inspect → break → repair → migrate workflow and requires an evidence-backed migration rationale.

Primary references and version notes

• Selenium downloads — current stable WebDriver client/Grid release baseline.
• SeleniumHQ/selenium-ide — current Selenium IDE repository, installation model, and release history.
• Selenium IDE releases — verify the exact IDE build before relying on UI or export behavior.
• Selenium IDE Code Export — official export workflow and documented target frameworks.
• Selenium IDE Commands — documented command semantics including element waits and assertions.
• WebDriver waits — synchronization principles used after migration.
• Page Object Models — service-oriented abstraction and component composition used in the migrated design.