# Changelog of YouTube Comment Scraper That Won't Quit (`thescrapelab/apify-youtube-comment-scraper-2-0`) Actor

- **URL**: https://apify.com/thescrapelab/apify-youtube-comment-scraper-2-0/changelog.md
- **Full Actor documentation**: https://apify.com/thescrapelab/apify-youtube-comment-scraper-2-0.md

## Changelog

### 2026-10-28 — Comment analysis text selection

- Question excerpts, phrase search, recurring terms and repeated-text analysis now use comment bodies even when the video's title is longer.
- Title-only rows no longer contribute comment-text analysis. Added short-comment and title-only regression checks.

### 2026-10-28 — Bounded recovery after access failures

- Final default-input verification exposed an empty YouTube access failure after earlier stress cases succeeded.
- Added automatic residential proxy recovery for empty transient source failures, with at most two fresh proxy sessions across the run.
- Recovery shares the existing deadline and per-video page allowance, preserves partial results, and records attempts without exposing credentials. Explicit page limits and dataset charging remain unchanged.
- Added regression checks for successful recovery, shared attempt limits, page and time caps, unavailable proxies, and credential-free summaries.

### 2026-10-28 — That Won't Quit branding

- Updated the display title, SEO title and README to YouTube Comment Scraper That Won't Quit.
- Updated the logo to emphasize continuing collection while preserving comments, and documented the verified sample workloads.
- Kept Comment analysis in the input form, along with the existing URL, API fields and collection limits.

### 2026-10-28 — Stress validation and clearer name

- Renamed the display title to YouTube Comments & Replies Scraper while preserving the existing Actor URL and API fields.
- Automatic page allowances now account for separate reply-thread requests; explicit API page caps remain enforced.
- Playback restrictions no longer prevent a separate check for publicly accessible comments, and failed videos retain their video ID in summaries.
- Live collection now retains short reactions, emoji-only comments and comments with unavailable author details. Existing supplied-record validation remains unchanged.
- Rejected source records now mark collection as partial and include validation reasons in the video summary.

### 2026-10-28 — Public comment collection restored

- Restored direct collection of public YouTube comments from video links or IDs without a customer API key or login.
- Simplified the form, with optional replies, top/newest ordering and phrase search. Existing record and dataset imports remain available to API clients.
- Added bounded page requests, duplicate handling, partial results and budget-aware dataset delivery. Video summaries and comment analysis are separate outputs.
- Updated customer documentation, SEO descriptions and the Store logo around the working comment workflow.