Skip to content

Latest commit

 

History

History
190 lines (129 loc) · 6.66 KB

CHANGELOG.md

File metadata and controls

190 lines (129 loc) · 6.66 KB

v0.2.7

Fixes

  • Unpinned required scipy version.
  • Fixed messaging for case when there aren't 100 recent datasets to show.

v0.2.6

Fixes

  • Now correctly handles a failed sign-up request if the entered password doesn't meet the required conditions on complexity.

v0.2.5

New stuff

  • The foreground query timeout, LC zip process timeout, the maximum number of dataset rows allowed to generate an LC zip for, and the number of rows per page for a dataset are all now configurable via the site-info.json file. Here is the updated base of the site-info.json file that is copied over when the lcc-server CLI is used to init a project base directory. Add the new keys to your own site-info.json file to activate these configurable parameters.

Changes

  • Authnzerver: user login now requires email_verified AND is_active status as opposed to just is_active status previously.
  • Frontend: Some fixes to the messaging when an asychronous query is still running when the timeout expires.

Fixes

  • Fixed several test failures.
  • Authnzerver: correctly delete all user cookies and all associated session entries when a user is deleted.
  • Admin: additional guards against escalation of privilege when editing user info.

v0.2.4

Fixes

  • Fixed a possible directory traversal when looking up docs pages.

v0.2.3

Fixes

  • Remove some unused functions from authnzerver.actions.user.
  • Frontend: typo fix: noreferer -> noreferrer for generated _target=blank links.

v0.2.2

Fixes

  • Add in missing LC format JSONs to the package manifest.

v0.2.1

This is a major new release. The format of various databases used by LCC-Server has changed, so it's not backward compatible with collections being served by the v0.1 series of releases. Use the lcc-server CLI tool to re-import your light curves, checkplots, and period-finding results. The format of the output LCC-CSV light curve files produced has not changed, so if you have these lying around, you do not need to regenerate them.

New stuff

  • New user account functionality, allowing users to have private or unlisted search result datasets, per-user access privileges and rate-limits, and the ability to be emailed when long-running queries finish. API keys can be generated for each user on their user home page, allowing access to objects and datasets marked as private. This functionality requires a working email server. If one is not provided, user sign-ins and sign-ups will be disabled.
  • All search services can now accept optional limits on the rows returned, can randomly sample search result rows, and can sort result rows by any database column.
  • Imported light curve collections now add to a footprint map that is displayed on the new Collections tab of the interface. The list of LC collections on this tab now includes links to directly search each collection, a summary of each LC collection's items, and links to explore the collection using some basic search queries.
  • The full-text search service can now look up objects in SIMBAD by name using the SESAME resolver. This also enables looking up objects in LC collections that are associated with star clusters, nebulae, or other extended sources.
  • The dataset browser is now paginated for large datasets. Per-object information pop-ups now include variability index information and GAIA neighbor information.
  • Datasets can be edited by logged-in users. Their name, dataset URL, description, and citations can be changed. For staff and superuser-level users, dataset ownership can be changed.
  • LCC-Server users can be managed by superuser-level user accounts. These users can also edit LCC-Server site settings and email server settings.
  • Logged-in LCC-Server users can launch SIMBAD queries for individual objects when browsing them in the dataset view if these results of these queries weren't in the original checkplots ingested by the LCC-Server.

Changes

  • Dataset access now uses visibilities: private, unlisted, public, instead of just private and public in v0.1.
  • All search service endpoints now only accept POST requests. An API key is thus needed if you want to call these non-interactively. Per-user API keys are required used to access datasets and objects that have been marked as private.
  • New API options have been added for most search query services to enable the new result random-sample, sorting, and limiting functionality.
  • Light curve ZIP files and dataset CSVs that are associated with private datasets will no longer be visible or accessible to users other than the one that generated them (via the interface or a user-tied API key).
  • All requests are now rate-limited based on user privilege level. All search query services now check user access and privilege level and will not return result rows that are marked as private or go over the user's row limit.
  • The lcc-server CLI can now run period-finding on an imported LC collection in addition to checkplot making. Choosing this option will incorporate phased light curves into the per-object information displayed by the LCC-Server.
  • The lcc-server CLI will ask for administrative user credentials when it is used to set up a new basedir. This will generate the first superuser level user account that can be used to manage server settings, users, and edit datasets.

Fixes

Many bug fixes all over the place. See the commit log for details.

v0.1.4

New stuff

  • paths in lcformat-description.json can now point to the current collection directory and the user's home directory by using the {{collection_dir}} and {{home_dir}} shortcuts.

  • added example systemd unit files.

  • dbsearch: can now redact search arguments from search results.

Fixes

  • searchserver_handlers: better messaging if a cone search doesn't return anything in any LC collection.

  • templates: remove broken page load autofocus.

  • indexserver: fix typo in parsing cpaddr command line argument.

v0.1.3

New stuff

  • added edit-collection lcc-server CLI command.

Fixes

  • fixed typo: abs_gaiamag should have been gaia_absmag. These will now be extracted correctly from the checkplot pickles if present.

v0.1.2

Fixes

  • cli, backend: fixed operations that were actually broken, but were working because we weren't testing them right.

Changes:

  • docs: removed useless markdown version of Jupyter notebook walk-through.

v0.1.1

Fixes

  • README: use absolute links since we now use README.md at PyPi as well.

v0.1.0

  • Initial release.