# /.#.tech - Slash Dot Hash Dot Tech Local-first software for image archives and historical weather. Synapic and Weather Stat by Slash Dot Hash Tech. - Site: https://slashdothash.tech/ ## Synapic beta on track for release - Link: https://slashdothash.tech/2026/10/13/synapic-beta-on-track-for-release/ - Published: 2026-10-13T20:05:15+00:00 - Author: Dean Kruger After months of coding and getting to know the Daminion API inside and out, I’m finally on the verge of releasing a beta version of Synapic — the tool I’ve been building to take some of the more tedious, repetitive work out of Daminion. If you’ve ever spent an afternoon on the same catalog chore for the hundredth time, that’s exactly the kind of thing this is meant to remove. The first thing it goes after is the AI tagging engine, and specifically the deduplication problem that comes with tagging at scale. Once you start cataloging in volume you quickly accumulate near-identical images — re-exports, resized copies, the same shot saved twice under different names — and the last thing you want is the engine burning through time re-writing metadata that already exists. So I’ve built a deduplication algorithm on perceptual hashing that flags those near-matches up front, lets you compare them side by side, and then tag, move, or clear them out in bulk instead of one painful item at a time. Then there’s upscaling — something I’ve been looking at for a long time and have always wanted to get right. A lot of older catalogs are full of images that were perfectly fine when they were captured but now look soft on a modern display, and being able to bring them back up without leaving the catalog closes a real gap. That combination — smarter tagging, cleaner deduplication, and native upscaling — is what this beta is about, and I’m close to getting it into people’s hands. ## Weather Stat – Historical Weather Analysis & Charts - Link: https://slashdothash.tech/2026/09/27/weather-stat-aggregator-3/ - Published: 2026-09-27T00:24:31+00:00 - Author: admin ## Overview Weather Stat transforms decades of daily weather data from Open-Meteo into actionable insights. It computes averages, typical ranges, extremes, and year-over-year climate trends, then generates both static and interactive visualizations. ## Key Features * **Historical analysis** – aggregates decades of daily weather data into averages, typical ranges, extremes, and year-over-year climate trends. * **Multiple output formats** – exports to CSV for further analysis, generates static charts for documentation, and produces Datawrapper-ready interactive charts for publishing. * **Dual interface** – available as both a command-line tool for scripting and automation, and a desktop GUI for interactive exploration. * **Open data source** – pulls data from Open-Meteo, providing access to global historical weather records without API keys or subscriptions. ## Use Cases * Climate research and trend analysis * Agricultural planning based on historical patterns * Weather-dependent project scheduling * Educational demonstrations of climate data * Integration with data pipelines via CLI ## Technical Details The tool connects to Open-Meteo’s historical weather API, fetches daily observations for specified locations and date ranges, then processes the data to compute statistical summaries and generate visualizations. Output options include raw CSV data, static chart images, and interactive HTML charts compatible with Datawrapper’s embedding format. ## Availability Weather Stat enters beta on December 1, 2026, available on request. The alpha release follows on January 1, 2027, sold with a 7-day money-back guarantee. Source code will be available on GitHub for review and contribution. ## Synapic – Automated Image Tagging for Daminion - Link: https://slashdothash.tech/2026/09/27/synapic-3/ - Published: 2026-09-27T00:24:31+00:00 - Author: admin ## Overview Synapic automates image catalog tagging using vision-language models running entirely on your local machine. It generates categories, keywords, and plain-language descriptions without relying on cloud services. ## How It Works Synapic consists of two processes that communicate over a local HTTP connection: * **Avalonia UI (C#)** – handles the wizard interface, Daminion server integration, and the processing loop that writes metadata back to files or the catalog. * **Python sidecar** – responsible solely for model loading and inference. It never touches the filesystem except to read the image it is asked to tag. The sidecar launches automatically with the application, waits until the model is ready, and tags images on demand. When the app closes, the sidecar shuts down cleanly— nothing lingers in the background. ## Key Features * **Local models only** – runs on Hugging Face vision-language models (LFM2.5-VL- 1.6B and similar). CPU or CUDA supported, with automatic fallback. * **Two data sources** – tag a local folder recursively, or connect directly to a Daminion server to tag items within a catalog or scope. * **Probability scoring** – tier-annotated confidence scores accompany each plain- language description, allowing you to trust or re-tag results as needed. * **Bulk processing** – includes a progress view with ETA, live colorized logging, pause, abort, and per-item retry capabilities. * **Metadata writing** – writes EXIF/IPTC for JPEG, tEXt/iTXt for PNG and TIFF, plus direct writes back to Daminion. * **Deduplication** – perceptual hashing (pHash/dHash/aHash/color-moment) identifies near-duplicates, with side-by-side comparison and bulk tag/move/delete actions. * **Results & export** – review results in a grid and export to CSV for further analysis. ## Cross-Platform Support Windows builds are available now. macOS and Linux are in development, planned for 2027 if there is demand. Windows installers ship now, and model weights can either be baked into the installer or downloaded on first run. A lite installer is available if you prefer to download the model yourself. ## Pricing Synapic is sold as a paid product with no free trial. A 7-day money-back guarantee lets you evaluate it against your real workload with no risk. Detailed product demonstrations are available so you can see it working before you buy. Downloads and licensing are handled through our store. Source code is available on GitHub for review and contribution.