News Aggregator 1.1.0

Change log

A new version has been released.
  • Feature: OPML import/export for news sources
  • Improvement: Cache the public feed's visible-item count instead of running a full COUNT(*) per pageview.
  • Improvement: Cap feed response body size to prevent runaway feeds from exhausting memory.
    • New option (Maximum feed body size) which is set by default to 5 MB.
  • Bugfix: Fix N+2 query during batch feed import.
  • Bugfix: Batch import now loads keyword filters and recent titles once per batch instead of once per source
  • Bugfix: Dedupe historical rows and promote url_canonical to a unique index so the DuplicateKeyException catch is actually load-bearing.
  • Bugfix: Clear orphaned duplicate_of_id / is_duplicate on survivors when their original is pruned.
  • Bugfix: Sanitise image URLs from feeds — reject non-http(s) schemes, resolve relative URLs, clamp length.
  • Bugfix: Cap item description length at write time to stop full-article bodies bloating the table

Release notes

A new version has been released.
  • Feature: OPML import/export for news sources
  • Improvement: Cache the public feed's visible-item count instead of running a full COUNT(*) per pageview.
  • Improvement: Cap feed response body size to prevent runaway feeds from exhausting memory.
    • New option (Maximum feed body size) which is set by default to 5 MB.
  • Bugfix: Fix N+2 query during batch feed import.
  • Bugfix: Batch import now loads keyword filters and recent titles once per batch instead of once per source
  • Bugfix: Dedupe historical rows and promote url_canonical to a unique index so the DuplicateKeyException catch is actually load-bearing.
  • Bugfix: Clear orphaned duplicate_of_id / is_duplicate on survivors when their original is pruned.
  • Bugfix: Sanitise image URLs from feeds — reject non-http(s) schemes, resolve relative URLs, clamp length.
  • Bugfix: Cap item description length at write time to stop full-article bodies bloating the table
Back
Top