- template formatting screwed up
- add link to gmaps for individual photo
- add admin screen to view upload processing errors.
- migrate views into a separate set of modules.
- almost all processing takes at least one second before completion, so this avoids unnecessary checks
- improve logging to include source image as well as key
- add script to upload image via api
- expose endpoint for confirming upload processing success.
- add function to validate whether or not an image exists, as a proxy for whether or not processing has completed.
- remove lint warning about wildcard import
- when this flag is toggled "on" it attempts to connect to the remote db from local, causing this to fail.
- reenable db migration task
- avoid key error when getting options.
- new column:
focal_lengthshows formatted view of value, e.g. "f/5.6" - integration of new metadata-processor, along with HTTP interface for it
- switch to use new version of metadata-processor
- get all the var names straight, conforming them to the other services.
- separate out iam config for different services to facilitate working with them.
- update component versions.
- remove automated static deploy, too many issues to bother with
- provide proper datastructure for a Doit action.
- inadvertent reference to self
- publish a new release to sentry whenever one of the services is deployed
- blend repo and account names when defining repo name
- wrong package name
- prognosticate next version.
- remove logic for publishing static files out of Makefile, as its much more reliable in deploy script
- update references to map files.
- only publish and commit if files have changed
- append newline explicitly
- include patch to js module so that deploys of static files dont generate multiple new files even when nothing has changed.
- Convert to use a versioned dep rather than a git ref.
- update references to build modules and folders
- build scripts now only do deploying
- use correct name for variable.
- improve logging and comments
- switch application service to deploy from a version rather than build it first.
- do not truncate release notes when gathering them
- include recent changes in the release notes.
- release: correct path to archive and initialization of version
- additional incorrect path for version.txt
- incorrect path for version.txt
- workaround issue with hashes on editable deps. pip does not allow hashes on "editable" deps (i.e. deps on repos like django-storages) cf.: pypa/pip#4995
- convert dependencies to
devdeps where possible
- release: add support for automated releases
- Add new build/deploy script
- Add Django Admin commnds for configuring remote system.
- Minor fixes to configure Sentry release metadata.
- Minor fixes to configure Sentry release metadata.
- Minor fix in configuration of Org name when publishing release to Sentry.
- Add additional metadata when publishing release to Sentry.
- Minor fix in format of release URL.
- Minor fix for publishing release to Sentry.
- Minor fixes in operational notifications.
- [#86] Implement security and operational fixes in prep for deployment.
- Dev module was unable to be run in a non-local env.
- [#77] Fix for bug in sharing.
- Add capability to share one or more images from a given album.
- Integrate with Github CI
- Misc cleanup and consolidation.
- Completed migration from Bitbucket to GitHub, including issues.
- Consolidated documentation from bitbucket wiki.
- Fix for 'Access Denied' error when uploading images.
- When app is released publish changelog to GitHub as a release.
- Rewrite of
elektrum-deployscript.
- Rewrite of
elektrum-releasescript.
- Introduce automated increment of pyproject version number from version.txt.
- [#63] Replace Zappa with LGW for deployment.
- [#62] Add API for uploading via CLI.
- [#57] Implement media upload via web browser.
- [#58] Automate provisioning of different environments.
- [#60] Rename project from
elektrontoelektrum. - [#61] Fix "Forbidden" error when uploading images.
- Upgrade all dependencies.
- [#50] Refactor network into separate, but identical environments.
- [#38] Multistage deployment environments
- [#43] Provide capability to reset a release to previous version.
- [#48] Minimize dependency on ELEKTRON_ENV environment variable.
- [#45] Switch from Elastic Beanstalk to use Zappa for deployment as a Lambda.
- [#49] Configuration of a NAT instance to allow for DB access from a Lambda.
- [#41] Improve robustness of photo processor deployment.
- [#36] Eliminate requirement to include create date when requesting image upload.
- [#37] Views and model for date-based views of media (collections, albums, items).
- [#10] Configure CDN, HTTPS, and bucket setup.
- [#40] Populate the date dimension table with a default range from 1/1/1970-12/31/2050.
- [#35] Upload lambda binary to S3 before deploy.
- [#34] Deploy versioned lambda function for the image processor.
- [#24] Enhance internal structure of templates to handle showing authenticated content.
- [#25] Simplify signup/login UX by putting both forms on the home page, rather than two separate pages.
- [#23] Model & test collections
- [#27] Fix syntax error in collection templates.
- [#28] Migration to pytest, add testing coverage.
- [#31] Model & upload media items.
- [#33] Remove 'collection' model.
- Testing of upload signing, upload processing.
- Switch to use Poetry for dependency management.
- [#20] Enhance user profile by adding profile image url.
- [#12] Change model to use a UUID for primary key.
- [#15]: Configure mandatory email confirmation for new accounts.
- [#16]: Default email backend will print the message to console.
- [#3]: Fix error when trying to send email when signing up.
- [#17]: Add test for email confirmation flow.
- [#18]: Make username a required field when signing up.
- [#19]: Update tests to provide username.
- Add a couple of unit tests.
- [#4]: Implement Google signup/authentication.
- [#13]: Fix issue where Google signups caused "duplicate username" issue.
- Recreate environment to test logging & https config files.
- Configure HTTPS.
- Clean test.
- Clean test.
- [#9]: Introduce multi-container config in order to address underlying issues that prevented access to staging DB.
- Configured app to use PostgreSQL across multiple environments.
- Add a single selenium-based integration test to provide basis for further tests.
- Require an email address & first/last names for login, ignoring username field; require unit tests to run and succeed before deploying.
- Ensure the app is on the tag so that
version.txthas a non-devversion number.
- Misc fixes around version number.
- [#1]: Add a library to proxy static files in-app, resolving issue where static files were 404 on beanstalk.
- [#2]: After the previous fix it turned out that files were not available to nginx, leading to 404 errors -- explaining the need for a volume LoL. Also, to better determine what version is running add some logic to rebuild the image each deploy and version it.
- Fix bug in condition check in deploy script.
- Add test to confirm static files are configured properly.
- Correct a bunch of misc errors with running container, document it better, and fail releases if container is not functioning correctly.
- Organize Django project files under a subdirectory, and expose some product information to the templates so that we know what version is being displayed.
- Run migrations before running server to collect static files and apply any db changes.
- Pause before a deploy, and rely on defaults for release bundling.
- Health checks are done direct to IP address, which is not in ALLOWED_HOSTS and causing the health check (and deploy) to fail.
- Fix stray char in ALLOWED_HOSTS that caused a bad request.
- Fix typo in app name in Dockerfile.
- Nothing changed yet.