Skip to content

AI media harness for planning, generating, and editing locally

TypeScript

Prepare 0.3.0 release notes and version

Bitti41daa7dbd985
diff --git a/RELEASE_NOTES.md b/RELEASE_NOTES.md
index af71e2a..6698075 100644
--- a/RELEASE_NOTES.md
+++ b/RELEASE_NOTES.md
@@ -1,15 +1,13 @@
-# Slopus 0.2.1
+# Slopus 0.3.0
-This alpha update adds DMAD LoRA support, resumable model downloads, and improvements to image generation, reference prompts, and the image bar.
+This alpha update adds Linux support, LAN generation workers, expanded color grading, character sheets, and shareable generator templates, alongside improvements to references and scene continuation.
## What's new
- Share URL-based generator templates using Import and Export in Generator settings. Versioned `.slop` JSON bundles include GPU download variants, LoRAs and additional safetensors, preserve existing templates on import, and leave local paths out of shared files.
-
-- Renamed Colour correction to Basic Corrections and added temperature, tint, highlights, shadows, whites and blacks alongside exposure, contrast and saturation.
+- Renamed Colour correction to Basic Corrections and added temperature, tint, brightness, highlights, shadows, whites and blacks alongside exposure, contrast and saturation.
- Added Creative with eight built-in procedural looks and intensity, faded film, sharpen and vibrance controls; RGB and hue/saturation/luma curve editors; and shadow, midtone and highlight Color Wheels with individual lightness controls.
- Vignette now includes midpoint, roundness and feather, plus negative amounts for lighter edges. Grading settings are saved with the project and applied consistently in preview and export.
-
- First and last scene frames use the searchable reference picker with image thumbnails. Reference selectors and prompt chips share configurable filters for images, audio, video, text, and RefMods.
- Configure a character sheet from the Image toolbar's Template menu, with one reference-enabled prompt for clothing and other details, a resolution selection controlling sheet height (512–2048 px), and step count and seed controls, then select Execute. A square waist-up shot and three 9:16 full-body views are combined into one child image in the image bar. The front view provides a shared clothing reference for the other views; all four views share the selected step count and seed.
- Continue scenes from a selected scene's saved latents, with adjustable overlap and a choice of beginning or end. This replaces Previous scene in First & last frame.
@@ -19,24 +17,22 @@ This alpha update adds DMAD LoRA support, resumable model downloads, and improve
- Images imported through Timeline or References, including selected video reference frames, appear in the image bar and open for editing.
- Duplicate references from their context menu, including multiple selected references with their prompts, attachments, and media settings.
- Create a video scene with an image as its first frame, or create an image reference, directly from the Image tab's image bar context menu.
+- Image canvas context menus now offer image actions, including Edit, Create Scene and Create Reference.
- Every project now includes Video and Image tabs. Older video and image projects keep their saved content, and Export supports both output types.
- Clothes and Accessories reference categories include common wearable items. Clothes offer optional color and fabric settings.
- Video references now offer a Frames mode: select still frames from a clip and use them as image references.
-- Interrupted weight and LoRA downloads retain their progress and can resume on retry, including after reopening the app.
-- Reference smart chips in image prompts and video shot descriptions, with reference artwork in the picker.
-- Queue still-image generation for individual images. Regenerated and edited images are grouped under their original in the image bar.
-- Generated still images retain their used seed, shown as a read-only field with a copy button in Image generation.
-- Higher image and video generation resolutions, plus standard image and video export sizes.
-- Downloadable DMAD 4-Step LoRA for MiniMax H3, with its re-noising sampling recipe applied automatically. Its fixed four-step schedule overrides step counts and turns off MotionCache.
+- Add an empty reference from the References toolbar and attach images from its preview header.
- Sound references: add an MP3, M4A, WAV, FLAC or Ogg file to a reference, choose a 2 to 15 second range, and cite it in a shot as `<Audio N>`, for example as a voice timbre. Sound references need a Ref2VA generator and an image or video reference in the same scene, and also work on LAN workers.
-- Bundled release notes appear once per app version.
- Slopus for Linux (x86_64): install the .deb or .rpm package, or run the AppImage, which updates itself. Local generation on Linux needs an NVIDIA GPU with CUDA 13.
- LAN workers: run the Slopus worker package on another Windows or Linux computer and pick it in the new **Settings → Workers** tab to generate there. Slopus finds workers on the local network automatically, or by address. Workers download generator weights from their download links themselves and use CUDA when available, with a Vulkan fallback; local-only files are sent from this computer, and finished videos, images and latents are saved in your project as usual.
## Fixes and improvements
- New empty images created while browsing child images stay in the same image family.
-- Colour correction includes a Brightness control for preview and export.
+- The Save button keeps its label while saving, preventing the top bar tabs from shifting.
+- The Image toolbar spans the workspace, with canvas tools and view controls grouped in the center.
+- Agent chat uses message bubbles with formatted replies and avoids repeating completed responses.
+- Export notification buttons appear below the output path so long filenames remain readable.
- Continue preserves the full joined video and audio from Slopfab. Adjacent source and continuation scenes use that shared decode in preview and export, preserving context across the cut. Regenerate older Continue scenes to apply the fix.
- The timeline playhead follows playback without waiting for editor updates.
- Copy and paste effect settings between matching effects from their **…** menus, including bypass state and LUT data.
@@ -46,12 +42,5 @@ This alpha update adds DMAD LoRA support, resumable model downloads, and improve
- The Video scene list scrollbar stays clear of Generate and Cancel, while scene header backgrounds extend to the edge.
- Timeline media thumbnails stay inside their preview area, keeping filenames clear in grid and list layouts.
- Portable builds show available updates and open GitHub releases for manual updating instead of running an installer.
-
-- Simplified image edits and added references for individual edits. Still-image generation receives only the references mentioned in its prompt, with clear image and RefMod identifiers.
-- Removed box coordinates and palettes from image prompts, and limited image controls to supported generator model types.
-- Improved image family deletion so removing a variant preserves its original. Selected image boxes can also be deleted from the canvas.
-- Long prompts scroll inside their fields. Smart chips align with surrounding text and no longer overlap on adjacent lines.
-- Moved the Add reference scrollbar clear of reference cards, and made the image bar scrollbar appear while scrolling.
-- Agent-generated shots use full language names, such as English.
-- Removed the opening-frame helper caption and made the release-note introduction render as normal text.
-- Updated the bundled SlopFab runtime to API 1.16 and adapted LoRA preparation to preserve companion-grid downloads for older adapters.
+- Improved reference picker scrollbar visibility and initial keyboard focus in the release notes dialog.
+- Updated the bundled SlopFab runtime to API 1.18 for refmod export and support for version-5 refmod bundles.
diff --git a/package-lock.json b/package-lock.json
index 5ec76b5..d5f3b8b 100644
--- a/package-lock.json
+++ b/package-lock.json
@@ -1,12 +1,12 @@
{
"name": "slopus",
- "version": "0.2.1",
+ "version": "0.3.0",
"lockfileVersion": 3,
"requires": true,
"packages": {
"": {
"name": "slopus",
- "version": "0.2.1",
+ "version": "0.3.0",
"dependencies": {
"@fluentui/react-icons": "^2.0.343",
"@tauri-apps/api": "^2.8.0",
diff --git a/package.json b/package.json
index fd0c269..43d4640 100644
--- a/package.json
+++ b/package.json
@@ -1,7 +1,7 @@
{
"name": "slopus",
"private": true,
- "version": "0.2.1",
+ "version": "0.3.0",
"description": "An AI media harness for planning, generating, and editing media locally.",
"type": "module",
"scripts": {
diff --git a/src-tauri/Cargo.lock b/src-tauri/Cargo.lock
index cf8d7f2..3e56118 100644
--- a/src-tauri/Cargo.lock
+++ b/src-tauri/Cargo.lock
@@ -3415,7 +3415,7 @@ checksum = "0c790de23124f9ab44544d7ac05d60440adc586479ce501c1d6d7da3cd8c9cf5"
[[package]]
name = "slopus"
-version = "0.2.1"
+version = "0.3.0"
dependencies = [
"getrandom 0.3.4",
"image",
@@ -3443,7 +3443,7 @@ dependencies = [
[[package]]
name = "slopus-core"
-version = "0.2.1"
+version = "0.3.0"
dependencies = [
"getrandom 0.3.4",
"image",
@@ -3462,7 +3462,7 @@ dependencies = [
[[package]]
name = "slopus-worker"
-version = "0.2.1"
+version = "0.3.0"
dependencies = [
"slopus-core",
"tauri-winres",
diff --git a/src-tauri/Cargo.toml b/src-tauri/Cargo.toml
index c44dec0..4dd23b0 100644
--- a/src-tauri/Cargo.toml
+++ b/src-tauri/Cargo.toml
@@ -1,7 +1,7 @@
# One version for the app and the worker. scripts/updater-manifest.mjs checks
# it against package.json and tauri.conf.json.
[workspace.package]
-version = "0.2.1"
+version = "0.3.0"
[workspace]
members = ["core", "worker"]
diff --git a/src-tauri/tauri.conf.json b/src-tauri/tauri.conf.json
index 69c0175..826e44c 100644
--- a/src-tauri/tauri.conf.json
+++ b/src-tauri/tauri.conf.json
@@ -1,7 +1,7 @@
{
"$schema": "https://schema.tauri.app/config/2",
"productName": "Slopus",
- "version": "0.2.1",
+ "version": "0.3.0",
"identifier": "com.slopus.desktop",
"build": {
"beforeDevCommand": "npm run dev",