# Personal screenshots Open a saved game's details and jump to **Media → Your screenshots → Upload** to keep a memorable moment alongside its trailers and IGDB artwork. Choose one JPEG, PNG, or WebP image, up to 8 MB. A game can be saved with any library status, including Wanted; discovery results must be added first. Saved games also show their personal gallery when opened through a discovery URL. If that saved-library lookup fails, use **Retry library lookup** in Media. Click a thumbnail to view the full image. **Delete** removes that personal image. Uploads persist across reloads and remain separate from IGDB screenshots, so refreshing game metadata does not replace them. Existing journal uploads automatically appear here; there is no migration or duplicate upload required. Images live on your Questarr server in the persistent data directory. Back it up with your database. The existing authenticated screenshot API checks game ownership, file size, MIME type, and image signatures; this change adds no new endpoint or external image service. ## Why this increment Research reviewed on 2026-11-01: - [Issue #1287](https://github.com/Doezer/Questarr/issues/1287) asks to upload personal photos in a game's gallery. The existing Journal already supported image storage, but Media only showed catalog artwork. Moving the controls to Media closes that discoverability gap and gives the collection a personal, playful element without another integration. - [Current product scope](PRD.md) includes journals and screenshots. Reusing that storage preserves existing photos and avoids a competing gallery system. - [Playnite](https://github.com/JosefNemec/Playnite) emphasizes a unified desktop library; [Gameyfin](https://github.com/gameyfin/gameyfin) emphasizes folder-based browser hosting; [ROMarr](https://github.com/BlizzHacker/romarr) emphasizes ROM acquisition and verification. Personal media is a useful fit for Questarr, rather than a claim of exclusive competitive advantage. The stronger potential differentiator remains a server-native owned-library workflow: finish the [bulk Steam import PR #1273](https://github.com/Doezer/Questarr/pull/1273), rather than starting another overlapping integration. Compared with Playnite's Windows requirement and the folder/ROM focus of the other projects, importing an existing Steam collection into Questarr's self-hosted backlog and acquisition workflow reduces setup effort without a running desktop agent. Steam import alone is not unique; the combined server workflow is the opportunity. It still depends on Steam library visibility and IGDB mapping, and is not a launcher or cross-store sync service. ## Differentiator candidates considered | Candidate | Evidence and user value | Decision for this run | | ---------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | Server-native owned Steam library import | Questarr's [PRD](PRD.md) names the library-centralizer audience and owned-source sync roadmap; [Steam GetOwnedGames](https://partner.steamgames.com/doc/webapi/IPlayerService#GetOwnedGames) supplies ownership. Playnite requires Windows while Gameyfin starts with local files. One action for a large collection can remove repeated manual cataloging. | Complete unfinished #1273 first. The bulk worker is a coherent first increment toward ownership + backlog + acquisition on a server. Competitive advantage is plausible, not validated market exclusivity. | | Webhook subscriptions | [Issue #1292](https://github.com/Doezer/Questarr/issues/1292) asks for CRUD, delivery testing, signatures and retries: useful for automation, but a broad integration surface with several security requirements. | Defer: meaningful integration work, with weaker evidence that it changes product choice. | | More reliable manual filesystem imports | [Issues #1274](https://github.com/Doezer/Questarr/issues/1274) and [#1275](https://github.com/Doezer/Questarr/issues/1275) describe long hashing/scanning and retry pain. Progress and idempotence would improve trust for large/NFS libraries. | Strong reliability follow-up; requires careful filesystem recovery design and is less distinctive against file-library products. | | Release score explanations | ROMarr already describes release scoring and verification; Questarr already has release intelligence. | Avoid another overlapping explanation panel as a weak differentiator. | | Multiplayer/co-op discovery | [PR #1295](https://github.com/Doezer/Questarr/pull/1295) already implements the filter. | Exclude active PR work. | No new differentiator PR was started alongside the unfinished Steam import: this run completes that existing increment on its original branch, and keeps the fun gallery increment on a separate branch.