Page:
Project Status
43
Project Status
Michael Mayer edited this page 2020-01-24 03:37:37 +01:00
Table of Contents
Little by little, we build our nest 🐣
Milestones
- Released working proof-of-concept for core technologies and user interface.
- Published Developer Guide containing all information necessary to get you started as contributor.
- Merged pull requests.
- Added godoc source code documentation.
- Refactored our Docker images to use multi stage building, size for production is now ~180 MB instead of ~1 GB.
- Improved continuous integration infrastructure and build scripts. Builds are 50% faster now.
- Added source code quality monitoring and coverage report.
- Added easy to understand labels for issues. Green is an idea, blue a todo/epic and red a bug. You can instantly see the status as well: in-progress, waiting, please-test or done.
- Andrea Ceroni joined our team as Scientific Adviser, see Research.
- The license of our main repository has changed from Apache 2.0 to GPL v3 after an interesting discussion on our developers list.
- Our new Patreon page is now online at https://www.patreon.com/photoprism and we've also added a SPONSORS.md file. This is a first attempt to cover our expenses and hopefully find a way to work full-time on this.
- We've spent a significant amount of time presenting PhotoPrism to potential contributors and sponsors. While we love coding, we felt this is important too. Feedback on the slides we created is welcome!
- We've set up a public demo at demo.photoprism.org, see #69.
- Samples, slides and fixtures can be downloaded from dl.photoprism.org. You are welcome to send RAW and JPEG samples from your favorite camera.
- Our 2nd Meetup in Berlin took place on 12/12/2018.
- Published slides by Andrea Ceroni on Personal Photo Management and Preservation.
- Replaced MySQL with TiDB as default database. Our updated demo already uses it: https://demo.photoprism.org/
- Some of our issues are now listed at IssueHunt. Get paid for doing open-source work! We received the first
$80$200 as donation, so don't be shy and give it a try. - Wrote a blog post for Gopher Academy: https://blog.gopheracademy.com/advent-2018/personal-photo-management/ - also published with a different screenshot on blog.liquidbytes.net
- We've been comparing various tools for RAW to JPEG conversion with Adobe Photoshop. Sips is our clear winner in terms of similarity and file size. On the downside, location data had rounding errors. See Converting RAW to JPEG.
- Apr 29, 2019: Reimplemented color indexing using a 3x3 thumbnail. It enables us to search specific parts of an image. Our previous algorithm was not deterministic, so search would have returned different results every time.
Demo: https://demo.photoprism.org/photos?q=green - May 4, 2019: Improved image classification (now based on NASNet Mobile). Added auth to our built-in TiDB server.
- May 7, 2019: Improved configuration and logging. Upgraded all JS dependencies such as VueJS, Webpack and Vuetify. Created example pages for "Calendar" and "Places". Added "make test-js" target to Makefile for running JS unit test suite.
- May 8, 2019: PhotoPrism is making big progress! Today, we've implemented a working map search based on Leaflet / OpenStreetMap. Try for yourself: https://demo.photoprism.org/places?q=San%20Diego
- June 2019: Lots of UI and label / image classification related improvements. New flat design and clipboard. Improved map, navigation and search filters. Runs as progressive web app on all modern smartphones. Added UI for upload to library, import and index.
- July 2, 2019: A read-only mode has been implemented for demo purposes and those who want to manage their files manually / using other photo management tools, see #56. Feedback welcome.
- July 3, 2019: Demo and Docker images now contain our latest improvements for performance and stability under the hood. No major UI changes. Note that the container port changed from 80 to 2342 if you update. Also you need a processor that supports AVX.
- July 16, 2019: The Jetson Nano indexed 80760 files / 471 GB in ~252.5h. That's ~11s per file. We're going to optimize that along with search performance. Our TensorFlow build config and binaries can be downloaded here: https://dl.photoprism.org/tensorflow/
- August-December, 2019: Implementation of remaining features continued after taking a break for commercial projects (and traveling to San Francisco and Crete). We now have location indexing based on S2 cell IDs, which are better to search / index than latitude and longitude. We even started to provide our own, fast Reverse Geocoding API. In addition,...
- you can create albums and add / remove photos
- labels and photo titles became much better
- there is a simple login
- our search form contains additional fields for color, year, lens & category
- settings page for selecting language and theme was implemented
- you can index, upload and import directly from our Web UI
- multiple files and albums can be downloaded as zip file
- the server can send updates to the client via websockets (e.g. server logs and photo count)
- we can find and reject potentially offensive content with our NSFW detector (also based on TensorFlow)
- January 2020:
- Many performance improvements and reduced thumbnail quality defaults to make PhotoPrism work on every computer.
- Continued working on our Geocoding API, now doesn't use any OpenStreetMap services in the background anymore.
- Released a new UI for Places, based on Mapbox GL and MapTiler Cloud: https://demo.photoprism.org/places
- Guy Sheffer made progress working on a Docker image for the Raspberry Pi.
- Dustin Oprea refactored Exif data extraction from images for additional performance.
- XMP support now working as a proof-of-concept (read-only).
- January 24, 2020:
- Image for Raspberry Pi is now frequently updated with the latest code on the
master
branch. - Implemented complete password protection (public mode without password is still available): #216
- Started working on photo edit dialog: https://github.com/photoprism/photoprism/issues/212
- Updated screenshots to document development progress: https://github.com/photoprism/photoprism/wiki/Screenshots
- Image for Raspberry Pi is now frequently updated with the latest code on the
Current Activities
- We'll continue to improve Places and do performance tests & optimizations so that large photo collections can be indexed and searched without issues.
- Nicola is having a look at TensorFlow models for scene category classification.
- File Browser (https://github.com/filebrowser/filebrowser) was archived by its maintainer last week. Maybe we can make this part of our project somehow. Added an issue for this.
- @thielepaul and @spaibs continue to work on our mobile app, see https://github.com/photoprism/photoprism-mobile
What can you do?
- Everyone is invited to contribute sample image files for testing, see https://dl.photoprism.org/samples/
- Any help with reaching out to potential sponsors and related OSS projects is highly appreciated.
- Please report bugs and test with uncommon or broken files at home.
See also MVP Milestone, Recently Updated Issues, Community, Meetups and Open Source
Quick Navigation
Issues labeled help wanted / easy can be good (first) contributions.
If you have a question or an idea, you are welcome to contact us via email, ask in our help forum or join the developers mailing list.
Our developer guide is open to editing by anyone with a GitHub account. Because renaming of pages in the wiki can break external links, please ask before renaming or removing any wiki page.