Skip to the content.

26.10.06 14:18

FastMediaSorter v2 - Product History

Snapshot Date: September 13, 2026 Audience: Developer / product owner Purpose: A documented timeline of product lineage and major milestones, combining repository evidence with owner-supplied historical context where the current checkout is incomplete.

1. Evidence Boundary

This history is reconstructed from:

Important limitation:

2. High-Level Product Arc

FastMediaSorter as a whole evolved across three eras:

The current Android v2 product has grown into a broad media-management platform with:

3. Lineage Before Android v2

Period Milestone Confidence
~1999 Very early Windows file sorter in the Visual Basic era. Owner recalls a Visual Basic 4-based first sorter around 1999. Owner-provided
2013 FastMediaSorter appears as a Windows / VB.NET product. Owner dates this line to 2013 on .VB NET 4.5. Owner-provided
2025-05 (approx.) First Android version appears. Owner states the first Android app existed around May 2025 and was not published. Owner-provided
2025-09 (approx.) Current Android v2 development begins. Owner states this branch has been under development since September 2025. Owner-provided

4. Repo-Corroborated Early Android v2 Milestones

These do not prove the first day of v2, but they do prove that the app already existed and had substantial surface before the structured March 2026 process.

Date / Period Milestone Evidence
2025-11-09 Existing Android UI screenshots show the app was already running and visually documented by early November 2025. docs/images/Screenshot_20251109_*.png, docs/README*.md
2025-11-14 Additional player screenshot indicates the player surface already existed by mid-November 2025. docs/images/Screenshot_20251114_184930.png, docs/README*.md
2025-11-30 Privacy-policy file dates show public/legal documentation was already being maintained by late November 2025. docs/PRIVACY_POLICY-uk.md
2025-12 Versioning examples in tooling reference December 2025 builds, which supports the claim that Android development and build discipline were already active in 2025. dev/build-with-version.ps1

5. Main Milestones in the Current Checkout

Date Milestone Evidence
2026-03 Owner states this month was a process reset: memory was cleared, the ticket system was introduced, and structured engineering workflow started. Releases and development had already existed before that point. Owner-provided; consistent with the sudden appearance of structured changelog / spec / workflow artifacts in repo
2026-03-02 Earliest clearly documented v2 baseline in the current repo: technical requirements doc, release packaging work, changelog discipline dev/CHANGELOG.md earliest entries
2026-03-19 Platform baseline broadened: standard minSdk lowered to Android 8 (API 26), legacy flavor covers API 23-25; toolchain/docs updated dev/CHANGELOG.md, dev/TECH_REQUIREMENTS.md
2026-03-19 Standalone/default-player integration became a first-class product area dev/CHANGELOG.md entries around standalone player and default-player flows
2026-03-21 to 2026-03-23 Product expanded into widgets, Camera Photos virtual resource, dynamic shortcuts, and stronger public-facing docs dev/CHANGELOG.md, docs/README.md
2026-03-27 to 2026-04-02 Scheduled file operations entered the product, including DB support, dialogs, and release gating dev/CHANGELOG.md
2026-03-30 to 2026-04-01 Duplicate finder / cleanup feature added with dedicated screen and workflows dev/CHANGELOG.md
2026-04-11 to 2026-04-14 Document surface matured: EPUB / PDF / TXT search, translation, selection actions, and TTS parity dev/CHANGELOG.md
2026-04-14 Wear OS work was formalized at roadmap level with a master plan and release tooling dev/CHANGELOG.md
2026-06-22 VR stabilization and standalone-player parity work were actively landing; repository process matured with stronger prerelease and audit discipline dev/CHANGELOG.md
2026-06-27 Google Play availability explicitly confirmed for version v2.60.6270.802 .agents/MEMORY.md
2026-07-10 Windows companion strategy pivoted: instead of a separate companion app, the functionality was redirected into the existing FastMediaSorter_Lite Windows product. PLAN/spec-catalog-archive.jsonl (S0421)
2026-07-17 Launcher-mode / desktop-style surface is under active implementation; Room schema has reached version 41 dev/CHANGELOG.md, AppDatabase.kt
2026-08-17 Wear OS became a substantial standalone product surface: local browse and playback, image viewing and slideshow, album art, settings, SMB/FTP/SFTP sources, and two-way Phone Data Layer resource sync were all recorded as complete. docs/WEAR_OS_STATUS.md
2026-09-09 to 2026-09-12 Wear gained its own wear:production Play track. Version 2.60.9021.951 is recorded as completed; the following Wear submission was deliberately deferred for a package of physical-watch and listing fixes. docs/PLAY_PUBLISHING_STATE.md
2026-08 to 2026-09 The watch became an active companion rather than only a remote browser: it can listen to and record watch audio, cast eligible content through the paired phone, synchronize resources and settings, and browse phone media with paging and previews. docs/ALL_FEATURES.jsonl (S2531, S2881, S2915, S2982-S2985)
2026-09 Live broadcasting extended across devices: a phone broadcast can select/switch camera lenses, be sent to a paired watch for immediate playback, and is represented by a versioned descriptor with endpoint, codec, and latency data. Watch Streams separates Video, Audio, and phone-sent channels. docs/ALL_FEATURES.jsonl (S3038, S3051, S3062, S3063)
2026-09 OCR translation now separates spatially disconnected text before overlay rendering, avoiding one translation plate spanning unrelated regions of an image. docs/ALL_FEATURES.jsonl (S3039)
2026-09-13 A reproducible project snapshot records 653,193 non-empty code lines across Phone, Wear, and automation scripts, with 5 Gradle modules and 589 tracked images. This is a repository measurement, not a release claim. docs/PROJECT_STATISTICS.md

6. Development Phases

Phase 0 - Product Lineage Before the Current Process

Before the current structured repo process, the product had already passed through:

The current repository does not fully preserve those earliest engineering records, but it does preserve indirect proof that Android v2 was already active by November 2025.

Phase A - Process Reset and Scope Expansion (March 2026)

March 2026 appears to be the point where the project became formally structured: changelog discipline, specs, workflow documents, and more systematic planning. The product already had enough surface to justify technical requirements, release packaging automation, and public documentation maintenance. During March, the codebase broadened rapidly across:

Phase B - Feature Maturation (April 2026)

April evidence shows the product moving beyond simple media sorting toward richer reading and automation workflows:

Phase C - Specialized Surfaces and Release Hardening (June 2026)

June evidence shows stabilization and specialization:

Phase D - Product as Platform (July 2026)

By mid-July, the product is clearly operating as a platform-style Android app:

Phase E - Cross-Device Integration and Operational Maturity (July to September 2026)

The period turns Wear OS from a planned companion into a release-bearing product surface. Its own Play track exists, while the next watch submission is held for measured physical-watch and listing fixes rather than being rushed through review.

The product scope now includes:

This phase contains completed capabilities as well as ongoing broadcast and Wear release work. The distinction is intentional: a completed Play track and active development are recorded separately.

7. Current Reading

As of September 13, 2026, the repository shows a product that has moved through these stages:

  1. Windows file-sorting utility lineage,
  2. Windows FastMediaSorter product line,
  3. first Android generation,
  4. Android v2 rewrite / expansion,
  5. multi-surface Android product,
  6. platform-like shell with launcher ambitions,
  7. a standalone Wear OS companion with its own production track,
  8. cross-device media and broadcast integration.

8. Working Historical Summary

The best current historical reading is:

For code-size and effort estimation, see dev/PRODUCT_COMPLEXITY_ASSESSMENT.md.