ParadeDB Files

What's Changed

• feat: documentation for indexes on expressions by @adiwajshing in https://github.com/paradedb/paradedb/pull/2764
• feat: implemented heap-based expression evaluation for non-indexed fields by @mdashti in https://github.com/paradedb/paradedb/pull/2740
• fix: Remove usage of deprecated Slack API by @philippemnoel in https://github.com/paradedb/paradedb/pull/2768
• feat: Added a GUC (enable_custom_scan_for_non_indexed_fields) for handling non-indexed fields in queries by @mdashti in https://github.com/paradedb/paradedb/pull/2767
• chore: add git commit hash to stressgres-generated graph images by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2778
• chore: Add support to the benchmark tool for emitting JSON by @stuhood in https://github.com/paradedb/paradedb/pull/2782
• ci: Remove hardcoded ICU on RHEL by @philippemnoel in https://github.com/paradedb/paradedb/pull/2786
• chore: Expand benchmarking of aggregates, and add detail to query types by @stuhood in https://github.com/paradedb/paradedb/pull/2783
• fix: "attempt to multiply with overflow" error in paradedb.aggregate by @rebasedming in https://github.com/paradedb/paradedb/pull/2766
• ci: Add generation of Stressgres results to CSV by @philippemnoel in https://github.com/paradedb/paradedb/pull/2781
• feat: Update release process to follow a singular branch by @philippemnoel in https://github.com/paradedb/paradedb/pull/2721
• ci: Make alter extension workflow properly checkout branch by @philippemnoel in https://github.com/paradedb/paradedb/pull/2791
• ci: Post "One-branch Release Model" Improvements by @philippemnoel in https://github.com/paradedb/paradedb/pull/2792
• ci: Add support for continuous benchmarking by @philippemnoel in https://github.com/paradedb/paradedb/pull/2780
• chore: Add the join dataset to continuous benchmarks, and fix alternative queries by @stuhood in https://github.com/paradedb/paradedb/pull/2794
• chore: Increase benchmark run count to reduce noise. by @stuhood in https://github.com/paradedb/paradedb/pull/2795
• chore: Fix unit for average runtime in the benchmarks by @stuhood in https://github.com/paradedb/paradedb/pull/2796
• fix: Add missing stopword filters to Jieba tokenizer by @wangkhc in https://github.com/paradedb/paradedb/pull/2790
• feat: custom fsm by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2765
• chore: make benchmarks tolerant of rogue .psqlrc files by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2798
• ci: Switch to Depot runners for benchmarks/stressgres by @philippemnoel in https://github.com/paradedb/paradedb/pull/2799
• ci: remove the cancel-in-progress by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2800
• ci: more benchmark backfill work by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2801
• ci: fix copy dest path in benchmark workflow by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2802
• ci: disable using fsm_info() in stressgres suite by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2803
• ci: fixup git ref passing to benchmark action by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2804
• ci: cleanup benchmark running by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2809
• ci: restore stressgres CSV-to-json benchmark job by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2810
• ci: separate stressgres tps metrics from others for alerting purposes by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2811
• ci: use the median for stressgres runs and do some cleanup between bench publishing jobs by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2812
• ci: sleep for random seconds before publishing to gh-pages by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2813
• ci: option (default true) for backfill to reset the data by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2814
• feat: move query benchmarks to reporting median, not average by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2815
• ci: fix backfill so it'll run when reset_data flag is false by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2816
• ci: fix stressgres generated png filename for slack by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2817
• ci: switch to a different ci provider by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2818
• ci: Manually deploy Mintlify by @philippemnoel in https://github.com/paradedb/paradedb/pull/2787
• ci: if running from backfill use input commit in slack message by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2819
• perf: Adjust benchmark job to post for perf PRs. by @stuhood in https://github.com/paradedb/paradedb/pull/2820
• ci: try different cloud provider by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2824
• ci: (temporarily) disable concurrency controls on benchmark workflows by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2825
• chore: Avoid impersonating Phil by @stuhood in https://github.com/paradedb/paradedb/pull/2822
• chore: Run benchmarks on benchmark-labeled PRs. by @stuhood in https://github.com/paradedb/paradedb/pull/2826
• test: added paradedb.score to regression queries for snippets with JOIN. by @mdashti in https://github.com/paradedb/paradedb/pull/2806
• ci: Fix Rust Cache by @philippemnoel in https://github.com/paradedb/paradedb/pull/2831
• docs: Make parallel worker docs more correct by @rebasedming in https://github.com/paradedb/paradedb/pull/2833
• ci: Make benchmarks run sequentially on RunsOn metal instances by @philippemnoel in https://github.com/paradedb/paradedb/pull/2830
• fix: update backfill workflow to point to new workflow.yml files by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2835
• ci: we (likely) need to sleep before publishing benchmarks by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2836
• ci: use the custom/composite actions from @main by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2837
• ci: fix uses: line by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2838
• ci: fix typeo by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2839
• ci: fix where the custom actions are to be found by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2840
• chore: Rename datasets and add string paging queries by @stuhood in https://github.com/paradedb/paradedb/pull/2834
• ci: fix copying of current actions by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2841
• perf: Lazily load fast fields dictionaries. by @stuhood in https://github.com/paradedb/paradedb/pull/2842
• chore: Formalize stable branches by @stuhood in https://github.com/paradedb/paradedb/pull/2848
• fix: orphaned delete entries get GCed too early by @rebasedming in https://github.com/paradedb/paradedb/pull/2845
• ci: benchmark workflow cleanups by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2851
• chore: fix schema upgrade scripts by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2854
• docs: Roadmap, architecture, guarantees/tradeoffs by @rebasedming in https://github.com/paradedb/paradedb/pull/2843
• ci: publish stressgres benchmark graphs by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2852
• ci: always copy latest composite actions from main by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2856
• ci: fix benchmark-stressgres action to resolve ref to sha by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2857
• feat: Add a custom scan for aggregates by @stuhood in https://github.com/paradedb/paradedb/pull/2763
• feat: a background merger for large layers by @rebasedming in https://github.com/paradedb/paradedb/pull/2743
• chore: Fix count query ordering in benchmarks by @stuhood in https://github.com/paradedb/paradedb/pull/2859
• ci: upload stressgres graph to PR by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2858
• test: Make GUCs a tested property, and add proptests for the aggregate custom scan by @stuhood in https://github.com/paradedb/paradedb/pull/2788
• ci: don't run benchmarks when the workflow files change by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2861
• ci: fix gh-pages link in slack messages by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2865
• ci: workflow to automatically backfill by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2863
• chore: missing UDFs in 0.17.0 upgrade script by @rebasedming in https://github.com/paradedb/paradedb/pull/2862
• docs: Point README roadmap to docs by @rebasedming in https://github.com/paradedb/paradedb/pull/2864
• docs: flesh out docs around query execution by @rebasedming in https://github.com/paradedb/paradedb/pull/2866
• docs: remove pre 0.14.0 changelogs by @rebasedming in https://github.com/paradedb/paradedb/pull/2869
• ci: Mark release as latest if semVer is higher by @philippemnoel in https://github.com/paradedb/paradedb/pull/2867
• chore: Upgrade to 0.17.0 by @rebasedming in https://github.com/paradedb/paradedb/pull/2870
• ci: fix benchmark links in slack messages by @eeeebbbbrrrr in https://github.com/paradedb/paradedb/pull/2875

New Contributors

• @adiwajshing made their first contribution in https://github.com/paradedb/paradedb/pull/2764
• @wangkhc made their first contribution in https://github.com/paradedb/paradedb/pull/2790

Full Changelog: https://github.com/paradedb/paradedb/compare/v0.16.2...v0.17.0