news
Free and Open Source Software
-
PocketSphinx - speaker-independent continuous speech recognition engine
PocketSphinx is a compact, speaker-independent continuous speech recognition engine originating from Carnegie Mellon University’s CMU Sphinx work.
Unlike newer neural speech recognition systems that depend on large models and substantial computational resources, PocketSphinx uses established statistical speech recognition techniques whose relatively modest resource requirements remain useful for embedded applications, offline command recognition, constrained hardware and specialised recognition tasks.
The project provides both a standalone command-line program and programming interfaces for applications needing direct control of decoding. It can recognise complete utterances, process live audio in speech segments and perform forced alignment against known text. Current PocketSphinx also incorporates functionality that was previously supplied separately by SphinxBase.
This is free and open source software.
Omnom - self-hosted web content preservation service
When you save a web page, Omnom can capture the page as rendered by a browser rather than simply storing its URL. This means dynamic content can be retained along with locally saved multimedia. Multiple snapshots of the same URL can be kept, with summary and comparison views helping you see how a page has changed over time.
Omnom also functions as an RSS reader and can follow ActivityPub streams. Its web interface supports multiple users, while flexible filtering can narrow stored content by date, text, tags, users, domains and URLs.
Browser extensions are available for Firefox and Chromium-based browsers. Omnom can be installed from a single binary or deployed using Docker. Authentication options include one-time login tokens, OAuth and trusted remote-user headers, while administrators also have access to command-line management functions.
This is free and open source software.
Barcoder – barcode and QR code generator
Barcoder is a graphical utility for generating one-dimensional barcodes and two-dimensional codes from text. Its workflow is deliberately straightforward: enter a value, select the required symbology and the application immediately renders a preview that can be saved for printing, labelling or use in other documents.
The interface is built with GTK 4 and libadwaita and is designed to remain practical on smaller screens. Barcode generation is performed locally with the rxing library. Input is checked against the requirements of the selected code type, with generation failures presented in the interface rather than silently producing an unusable result.
This is free and open source software.
jsonrepair - repair invalid JSON documents
jsonrepair repairs invalid JSON documents. It can correct many common syntax problems automatically, making malformed or incomplete JSON usable without manually locating and fixing every error.
The software is available both as a library and a command-line utility. It also provides a streaming implementation for processing very large documents without loading the entire input into memory.
This is free and open source software.
Aliza MS – medical imaging and DICOM viewer
Aliza MS is a medical imaging and DICOM viewer designed for examining clinical imaging studies on the desktop.
Its workflow centres on a DICOM scanner: users can open a directory containing DICOM files or a DICOMDIR file, select one or more series, and load them into the viewer. Directories can also be opened with drag-and-drop. It supports both conventional and time-resolved datasets, with tools for navigating and analysing studies in multiple dimensions. The project provides native Linux builds and can also be built from source on Linux.
This is free and open source software.
whisper.cpp - implementation of OpenAI's Whisper automatic speech recognition model
The project is designed for efficient local inference without requiring the original Python and PyTorch software stack, making Whisper practical on systems ranging from conventional Linux desktops and servers to Raspberry Pi boards, mobile devices and WebAssembly environments. Its implementation builds on ggml and places considerable emphasis on portability, low runtime overhead and hardware-specific acceleration.
Users can run transcription entirely on the CPU or take advantage of several GPU and accelerator backends. Models are converted to ggml format, with quantization available when reduced memory consumption and storage requirements are important.
This is free and open source software.
Kodaskanna – multi-format 1D/2D code scanner
The application focuses on image-based input and is not tied to a camera. It can run as a standalone program or be invoked through KDE desktop integrations. Code recognition is handled by ZXing-C++, while the current codebase uses Qt 6 and KDE Frameworks 6.
Kodaskanna’s broader design goal is to act as a reusable component in workflows that need machine-readable data supplied by the user. The project envisages extensible input sources and reusable validation or preview components, with the scanner usable either in-process or out-of-process.
This is free and open source software.
Lap - local-first desktop photo manager
Lap includes tools for finding duplicate and visually similar photographs, comparing images in a four-pane viewer, and culling large collections. It recognises Apple Live Photos and Android Motion Photos and can keep RAW files paired with matching JPEG or HEIC companions.
Basic editing includes cropping, rotation, resizing and image adjustments. There is broad support for photographic, RAW and video formats, while local AI facilities provide text-based image search, similarity matching, automatic subjects and face clustering. These operations run locally rather than uploading photographs to an external service.
Lap uses SQLite for its catalogue and is built with Tauri, Rust and Vue.
This is free and open source software.
Readeck - self-hosted read-it-later application
Bookmarks can be organised with labels, favourites, archives and collections. Full-text search helps locate saved material, while highlighting and annotation let you mark important passages. Readeck handles conventional articles as well as pages focused on images or videos.
A browser extension is available for quickly saving pages from Firefox and Chromium-based browsers. Readeck also supports EPUB export and provides an OPDS catalogue, making saved material accessible from compatible e-readers and reading applications.
The software uses SQLite by default and is designed to be straightforward to self-host. It is distributed as a single binary and is also available as a container image for Docker or Podman.
This is free and open source software.
Color Picker for COSMIC - color utility for COSMIC desktop
It follows the desktop’s native visual language through libcosmic and provides a focused workspace for examining colours without relying on a browser-based tool. The application is a third-party project rather than an official COSMIC component.
Its interface can accommodate several colour representations at the same time, which makes it useful when translating colours between design, graphics, and development workflows. Screen colour selection uses the desktop portal infrastructure, so the application works with modern Wayland security boundaries rather than requiring unrestricted access to the display.
This is free and open source software.
FunASR - end-to-end speech recognition toolkit
FunASR is an end-to-end speech recognition toolkit. It provides tools for building, training and deploying speech-processing systems for offline, streaming and edge use.
The project brings together automatic speech recognition, voice activity detection, punctuation restoration, speaker diarization, emotion recognition and audio-event detection within a common framework. It supports a range of model families, including Fun-ASR-Nano, SenseVoice, Paraformer and integrations for other speech models, with deployment options spanning local CPU or GPU inference, WebSocket streaming and API-based serving.
This is free and open source software.
OTree - terminal user interface for viewing structured data
OTree is a terminal user interface for viewing structured data as an interactive tree. It supports JSON, YAML, TOML and XML files, making deeply nested documents easier to explore from the terminal.
The interface combines a tree overview with a data view and provides navigation, filtering, syntax highlighting and mouse support. Its colours and key bindings can be customised through a configuration file.
This is free and open source software.