Skip to content

Releases: ThaddeusJiang/Vmemo

Vmemo 2026.6.9

09 Jun 08:35
@ThaddeusJiang ThaddeusJiang
2026.6.9
d157f25

Choose a tag to compare

View all tags
Vmemo 2026.6.9 Latest
Latest

Changed

  • Image uploads now use a shared 50 MB limit across the web UI and REST API, accept .tiff files by normalizing them to PNG, and return clearer documented errors when uploads are too large or fail during parsing. The limit can be configured with IMAGE_UPLOAD_MAX_FILE_SIZE. (#210)
  • API request and response summaries are logged for /api/* routes without logging bearer tokens or file contents. (#210)
Assets 2
Loading

Vmemo 2026.5.17

16 May 02:46
@ThaddeusJiang ThaddeusJiang
2026.5.17
3e425da
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Choose a tag to compare

View all tags
Vmemo 2026.5.17

What's Changed

  • fix: 修复反向代理 SSL 重定向循环并完善 local-machine 的 dev 域名验证流程 by @ThaddeusJiang in #206

Full Changelog: 2026.5.16...2026.5.17

Contributors

ThaddeusJiang
Loading

Vmemo 2026.5.16

16 May 00:57
@github-actions github-actions
2026.5.16
bbe5bc8
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Choose a tag to compare

View all tags
Vmemo 2026.5.16

[Vmemo - 2026.5.16] - 2026-05-16

End Users

Added

  • Added Vmemo MCP image management tools for searching, creating, reading, updating, and deleting images from MCP-capable AI clients.
  • Added MCP image resources so clients can lazily read image URLs, HTML previews, or base64 image data after search results are returned.
  • Added image tag management and tag-based classification support.
  • Added tag pages so users can browse and review images by tag.
  • Added image jobs page and notification integration for background job progress visibility.

Changed

  • Changed MCP image search to return lightweight image details and resource references instead of embedding image data in every search result.
  • Changed image cards to use a unified dropdown action menu and support delete directly from list view.
  • Changed image base64 reading to a unified path for more consistent AI and API behavior.
  • Changed background async processing to a top-level Jobs domain/resource model.

Fixed

  • Fixed Vmemo MCP image tools returning a server error when requests reached AshAi without an authenticated actor.
  • Fixed thumbnail and image file fallback behavior to improve load reliability when optimized variants are unavailable.
  • Fixed browser clipboard image upload compatibility across multiple payload formats.
  • Fixed REST API clipboard-image upload failures and reduced coupling between upload and AI flow.
  • Fixed upload fallback behavior when OpenRouter is unavailable or misconfigured.
  • Fixed image file validation and AI-unavailable warning paths to reduce silent failures.

Maintainers

Added

  • Added database migrations for tags and jobs, plus a backfill migration to align historical image-processing states with job statuses.
  • Added Typesense migration for memo_images.tags to support tag faceting.

Fixed

  • Fixed Typesense migration tracking behavior to avoid drift between expected and applied schema changes.

What's Changed

Full Changelog: 2026.5.9...2026.5.16

Contributors

ThaddeusJiang
Loading

Vmemo 2026.5.9

09 May 13:45
@ThaddeusJiang ThaddeusJiang
2026.5.9
7bee298

Choose a tag to compare

View all tags
Vmemo 2026.5.9

Release Summary

This release introduces an improved image rotation experience and adjustments to the REST API response conventions.

End Users

Added

  • Added image rotation in the image detail dialog with instant on-screen preview, so users can quickly correct incorrect upload orientation.

Changed

  • Changed REST API image responses to use HTTP status codes for success/error states, return image detail page URLs for create/show, and include the deleted image id in delete responses for easier client-side cache updates.
  • Breaking change (REST API contract):
    • Response bodies no longer include a top-level status field for either success or error.
    • On DELETE /api/v1/images/:id, a successful response payload now returns only data.id (no success message).
    • POST/GET /api/v1/images now return image detail page URLs in data.url instead of storage file paths.

Fixed

  • Fixed the delayed rotation feedback issue where images appeared unchanged until a full page refresh.

Maintainers

Changed

  • Updated image file response cache negotiation to use stronger validators (ETag/Last-Modified) for images updated in storage.

Migration Guide (Client)

  • Use the HTTP status code as the single source of truth for success/error.
  • Replace checks like response.status === "success" with response.ok (or a status-range check).
  • For failed requests, read error.code and error.message from the response body.
  • For delete cache updates, read response.data.id and invalidate/remove that image id locally.

What's Changed

Full Changelog: 2026.4.29...2026.5.9

Contributors

ThaddeusJiang
Loading

Vmemo 2026.4.29

29 Apr 13:23
@github-actions github-actions
2026.4.29
9eee540
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Choose a tag to compare

View all tags
Vmemo 2026.4.29

End Users

Added

  • Post-login pages now support English, Chinese, and Japanese.
  • Added a global AI Drawer entry so users can open AI chat from anywhere.
  • Unified image caption/query flows through the OpenRouter path for more consistent AI behavior.
  • Unified light-theme semantic components (Alert/Badge/Toast) for more consistent feedback UI.
  • Added independent profile fields (name, avatar, language, appearance).
  • Extracted notifications dropdown into a reusable component for more consistent interaction.

Changed

  • AI image requests now preprocess large images before calling external vision services, reducing transfer size while keeping original uploads intact in storage.

Fixed

  • Aligned visual and interaction details across landing/auth/app pages.

Maintainers

  • Change: Runtime URL settings for dev/test are now centralized in config/runtime.exs.
    Migration:
    1. Ensure required variables are set in runtime environment.
    2. Restart services and verify runtime checks.
      Example:
DATABASE_URL=<value>
TYPESENSE_URL=<value>
MOONDREAM_URL=<value>

Changed

  • Refined worktree workflow: trigger only on explicit request, with standardized create/cleanup steps.
  • Docker runtime image now includes ImageMagick so vision preprocessing is always available in production containers.

Release Gate Notes

  • Config/env changes were detected against base ref 2026.4.19.
  • Release proceeded with explicit confirmation from user.
  • GitHub Release/tag publishing is handled by CI after merge.

What's Changed

Full Changelog: 2026.4.19...2026.4.29

Contributors

ThaddeusJiang
Loading

Vmemo 2026.4.19

19 Apr 06:38
@ThaddeusJiang ThaddeusJiang
2026.4.19
3e56458

Choose a tag to compare

View all tags
Vmemo 2026.4.19

What's Changed

Full Changelog: 2024.12.25...2026.4.19

Contributors

ThaddeusJiang and devin-ai-integration
Loading

Vmemo 2024.12.25

19 Apr 05:39
@ThaddeusJiang ThaddeusJiang
2024.12.25
738f1bb

Choose a tag to compare

View all tags
Vmemo 2024.12.25

Overview

Vmemo is a visual memory application that helps users record life in a visual way rather than with boring, easily forgotten text.

Why Vmemo?

Our brains are naturally inclined to remember visual content better than text. Visual memory is more intuitive, impactful, and easier to review, making it a powerful tool for capturing and retaining life’s moments.

Core Features

  • Account and authentication flows: sign up, sign in, password reset, and user settings (Phoenix LiveView).
  • Photo upload and browsing: upload entry points, photo detail pages, and list/waterfall browsing.
  • Photo deletion and note updates: delete photos from detail pages and maintain note content linked to photos.
  • Typesense-powered search: document write/read and hybrid search support for photos and notes.
  • File access and web delivery: storage file routes and baseline page experience.

PRs

  • #1 feat: upload images then insert typesense
  • #2 chore: deploy to zeabur
  • #3 feat: delete photo
  • #4 fix: Dockerfile npm install
Loading