Skip to content

NEWS.md

Latest commit

 

History

History
2151 lines (1134 loc) · 55.8 KB

NEWS.md

File metadata and controls

2151 lines (1134 loc) · 55.8 KB
Raw

duckdb 1.5.5.9020

Bug fixes

rcc

  • Run the testsuite when rcc-one.sh runs outside Actions (#2669).

rcc

  • Tell a missing clang-format apart from a misformatted tree (#2671).

flavor

  • Run the flavor-rename guard where it can actually fire (#2676).

vendor

  • Refuse an upstream clone that cannot name a version (#2677).

vendor

  • Say how a retired patch reaches the vendor commit (#2679).

series

  • Report what stage 5 added, not what it set out to add (#2680).

ci

  • Backport kit fixes from rigraph, duckdb-r and dm.

duckdb 1.5.5.9019

Bug fixes

series

  • Register the ours-version driver before refusing over it (#2665).

series

  • Read every universe that publishes a green, not just duckdb (#2670).

Documentation

meta

  • Make authoring/ obey itself, name evidence in the rules, and refresh the meta/ node (#2673).

duckdb 1.5.5.9018

Documentation

meta

  • Consolidate the authoring rules, and split recorded evidence into its own leaf (#2672).

duckdb 1.5.5.9017

Bug fixes

series

  • Make the ours-version driver reach the clone that needs it (#2663).

Features

series

  • Explain the flavor surface in the convergence check (#2661).

duckdb 1.5.5.9016

Bug fixes

series

  • Refuse to replay when the ours-version merge driver is missing (#2659).

series-check

  • Classify an R CMD check gate failure instead of giving up on it (#2653).

series-advance

  • Carry the twin's work in a file the buffer also touched (#2657).

Features

series

  • Keep forwarding a base series with a live -fwd, and check the two converge (#2658).

duckdb 1.5.5.9015

Bug fixes

ci

  • Give the warning gate its own flags, and skip what a dropped Suggests cannot support (#2652).

duckdb 1.5.5.9014

Bug fixes

testing

  • Read a double-digit patch release as a release, and say that air runs here (#2638).

glue

  • Drop the semicolon after two namespace closings (#2645).

patch

  • Make the vendored tree compile without warnings (#2617).

vendoring

  • Pin the upstream clone's abbreviation, so a vendored tree is reproducible (#2489, #2631).

vendoring

  • Say when the glue check could not run, instead of blaming the glue (#2513, #2629).

scripts

  • Let BASE_SCAN_DEPTH raise the base scan in all four scanners (#2512, #2628).

rfuns

  • Unshadow two locals (#2612).

dbplyr

  • Render n_distinct() through the exported sql_glue(), and warn on an older dbplyr (#1982, #2605).

glue

  • Drop a stray semicolon after a constructor body (#2610).

glue

  • Suffix the one shadowing parameter left in a header (#2618).

glue

  • Build a string_t from R's const char* without a cast (#2619).

glue

  • Widen to R's double explicitly when copying a column out (#2611).

glue

  • Stop casting away const on R's read-only data pointers (#2609).

glue

  • Suffix constructor parameters that shadow their member (#2608).

ci

  • Read wrapped Config/gha/extra-packages values, bump auxiliary workflows to ubuntu-26.04 (#2603).

series

  • Keep stage 5 alive when the buffer is longer than the chunk (#2602).

scan

  • Materialize packed columns on R's thread, not on the scan's (#2582).

signal

  • State the interrupt handler's invariants where they are used (#2598).

Features

build

  • Gate the compiler warnings each scope answers for (#1829, #2620).

relational

  • Remove allow_materialization from rel_to_altrep() (#1052, #2643).

connections

  • Report instance settings a reused database cannot apply, and avoids treating URL-like paths like files (#2560, #2641).

build

  • Make the prebuilt-engine setup work without privileges (#22, #2622).

dbplyr

  • Translate clock::date_build() and difftime() (#2621).

series

  • Fold the base series' test-side fixes in as stage 5 consumes the buffer (#2594).

Chore

  • Format with air.

Documentation

build

  • Say that a vendored edit does not rebuild itself (#2648).

testing

  • Give the two undocumented testthat helpers a header (#2502, #2634).

branches

  • Point the invariants' deepen line at what is missing, not at notes BRANCHES.md lacks (#2516, #2633).

  • Convert the stray Rd spellings in roxygen to Markdown (#2616, #2635).

style

  • Spell out the tidyverse rules R here answers to, and the deviations (#2613).

testing

  • Say when a snapshot is the right tool, not only how to accept one (#2624).

interop

  • Say what two limits cost a caller, and pin them (#590, #1064, #2600).

authoring

  • Cut an entry to the shortest correct statement, and split the rest into a leaf (#2604).

dbplyr

  • Measure what DISTINCT ON would buy, and hand it over anyway (#384, #2601).

handbook

  • Close the last three days' issues into their leaves (#2595).

plan

  • Design a producer thread for streaming results, with a TPC-H fetch benchmark (#2583).

Code style

clang-format

  • Never reorder includes (#2642).

Refactoring

  • Raise every error through abort(), so it points at the caller (#2615, #2637).

  • Extract result data-frame allocation into a helper (#2585).

Performance

vendoring

  • Keep git's stat cache valid across a regeneration of the vendored tree (#2490, #2632).

Uncategorized

  • Docs(interactive): Give the interrupt its leaf, and answer #202 down to the client name (#2596).

duckdb 1.5.5.9013

Continuous integration

  • Skip arrow if uninstallable.

Testing

storage

  • Compare spill paths on one separator spelling (#2579).

duckdb 1.5.5.9012

Bug fixes

flavor

  • Answer rather than error when a series has no README.Rmd (#2576).

flavor

  • Catch a generated README that reached a series under the mainline name (#2575).

  • Respect the session TimeZone for TIMESTAMPTZ columns (#2401).

storage

  • Make temporary storage usable by default, with the CLI's semantics (#1604, #2562).

Continuous integration

  • Wrap the commit status update into an action.

  • Route ccache through one-word compiler wrappers on Unix.

  • Add context for cherry-picking.

  • Route ccache through one-word compiler wrappers on Unix.

  • Name every step and restore the log entry setup-pandoc swallows.

Documentation

handbook

  • Give timestamps and time zones their own leaf (#2567).

handbook

  • Prefer durable storage for hard-to-reproduce experiment results (#2564).

Refactoring

rcc

  • Cut the store's automatic writers down to the leg's own publish (#2578).

Testing

explain

  • Skip Windows inside each test_that() so the snapshot survives (#2577).

duckdb 1.5.5.9011

Bug fixes

readme

  • Vendor the DuckDB logos, and write two READMEs from one source (#2517, #2518, #2558).

flavor

  • Catch a renamed file that reached a series under its mainline name (#2559).

storage

  • Anchor the ~/.duckdb consent default at interactive() (#2550).

series-advance

  • Set -build-base, rather than advance it (#2540).

patch

  • Fix the re2 clang warnings instead of hiding them (#2492).

series-advance

  • Clamp the replay anchor to the merge base (#2548).

rcc

  • Fetch the record a publish replaces, so a retry's verdict can land (#2547).

rcc

  • Stop the backstop recording a pending status as a verdict (#2544).

ci

  • Let adbcdrivermanager fail to build on Windows arm64 (#2539).

Documentation

handbook

  • Record the untyped-NULL mapping as designed, declining the flip to logical NA (#155, #2561).

integrations

  • Answer the frame-library request, Arrow and non-Arrow (#642, #2557).

statements

  • Move the multi-statement prepare rule to the leaf that owns it (#179, #2555).

connections

  • State where the instance's settings bind, and both ways dbConnect() gets them wrong (#83, #171, #2553).

data-import

  • Say what a virtual filename column changes (#1733, #2556).

types

  • Give the geometry write side its route, not just its absence (#1670, #2552).

types

  • State the UTF-8 repair that works, not the one that looks like it (#12, #2551).

  • Review DuckDB API usage in the glue code and assess a C API migration (#2413).

branches

  • State the mirrors' cadence, now that krlmlr/duckdb-r is a fork (#2494, #2542).

  • An entry that wraps its subject applies everywhere, so applying proves nothing (#2546).

  • Record the August 2026 forward, and turn the cpp11 gap it walked into into a check (#2545, #2543).

Refactoring

series-loop

  • Read verdicts from the each-rcc runs, keep rcc2 as the fallback (#2549).

duckdb 1.5.5.9010

Bug fixes

tests

  • Expect the extension install to fail where DuckDB does not cover the platform (#2425, #2529).

Features

  • Mirror the fork's branches with the Pull app, and drop sync.yaml (#2534).

rcc

  • Move the verdict store to rcc2, one file per commit (#2505).

Chore

  • Increase retention to 180 days [ci skip].

Continuous integration

  • Add sharded revdep2 workflow.

Documentation

extensions

  • Close the remaining gaps the extension issues map to (#2425, #2536).

handbook

  • Record the Windows extension-coverage boundaries the issues keep re-finding (#2530).

Testing

extensions

  • Drop the duplicate platform skip that made the canary unreachable (#2537).

extensions

  • Skip the install test where DuckDB publishes no binaries, and pin the gap (#2535).

Uncategorized

  • Ci: Harden workflow_run workflows against untrusted pull requests (#106).

  • Ci: Pin third-party actions to commits and let Renovate keep them pinned (#105).

duckdb 1.5.5.9009

Bug fixes

flavor

  • Name the flavor in the README blurb, and find GNU sed (#2510).

Continuous integration

  • Remove unused pr-commands workflow.

Documentation

plan

  • Record what the closed handbook wave verified, and file what it should not have written down (#2507).

handbook

  • Weigh documenting a limitation against removing it (#2509).

handbook

  • Complete the leaves the triage closes will link (#2523).

Uncategorized

  • Ci: Pass workflow context through the environment, not into script text (#102).

  • Ci: Add a Windows arm64 (windows-11-arm) check on R-release (#99).

duckdb 1.5.5.9008

Bug fixes

series-advance

  • Restamp the vendor counter across a replay (#2525).

series-port

  • Hold main's version bumps back from a series (#2496, #2526).

windows

  • Export only the registration entry point from the DLL.

vendor-one

  • Refuse a clone HEAD off the buffer's upstream line.

series-advance

  • Drop an empty replay pick instead of aborting the extend.

Chore

  • Drop the CMake scaffolding, and the formatter branch that served it (#2511, #2520).

  • Auto-update from GitHub Actions (#2524).

Documentation

versioning

  • Record the declined max-across-prefixes merge rule (#2488, #2527).

versioning

  • Say when the version counter freezes, and why build-base lags.

series-loop

  • Send a fix to the handbook before it is written.

series-loop

  • Say where a stage-3 fix goes, and what counts as evidence.

Uncategorized

  • Feat(series-port): Detect a frozen series, and let it skip the port (#98).

duckdb 1.5.5.9007

Bug fixes

each

  • Fail the plan when green is off the branch's lineage.

Features

series-loop

  • Read r-universe builds, and fix the Windows unity-build break.

Documentation

  • Fix the clone layout, and the README links CRAN cannot follow (#2506).

handbook

  • Give the tree a way to grow into its blind spots (#2504).

handbook

  • Add usage/relational/, and give duckplyr its own section (#2495).

handbook

  • Land the MVP — every leaf written, at reference or core depth (#2493).

Testing

progress

  • Pin that a handle collected in a callback cannot re-enter.

duckdb 1.5.5.9006

Bug fixes

vendor

  • Drop the dead knobs, classify patch failures, restore the tree on a no-op run (#2491).

Documentation

vendoring

  • Fold by default, extend on pending, require a rising vendor counter.

  • Write every upward link from the repository root (#2458).

duckdb 1.5.5.9005

Bug fixes

rcc

  • Do not restore the persistent ccache under R-devel (#2455).

vendor

  • Bound the base scan and say when it comes up empty (#2434).

configure

  • Drop a stale Makevars.system-lib so a vendored build is one (#2446).

rconfigure

  • Drop bundled jemalloc from the generated source list.

Features

series-loop

  • Set the firing up, and make the tooling PR a stage (#2441).

each

  • Select work from the verdict store, not from statuses (#2440).

series

  • Make cutover a manual operation the loop only suggests (#2449).

series-loop

  • Vendor with main's script, into the buffer worktree (#2436).

Documentation

  • Propose the handbook — a strict topic hierarchy with full cover, first two levels (#2454).

scripts

  • Give every script an extractable one-line header (#2453).

readme

  • Fix the duckdb.1.4.dev badge ranges (#2452).

plan

  • Mirror the release branches from the fork, not from sync.yaml (#2450).

readme

  • Point the duckdb.dev badges at the cut-over main series (#2451).

  • Quarantine the superseded loop design and the main-dev review (#2438).

each

  • State that there is no running marker, and why (#2439).

Refactoring

series-port

  • Let the subject decide what a vendor commit is (#2435).

ci

  • Retire the legacy per-commit dispatch path (#2442).

Uncategorized

  • Feat(rcc-one): Bound each stage with its own timeout (#97).

duckdb 1.5.5.9004

Continuous integration

each

  • Queue the shards oldest first (#2445).

duckdb 1.5.5.9003

Bug fixes

series-advance

  • Register the version merge driver before the replay.

series

  • Compute the buffer anchor safely, in both scripts.

series

  • Anchor the buffer replay on -dev's newest vendor commit.

Features

series

  • Bring -dev level with main as a loop stage — identity over curation.

Documentation

  • Add the vendoring simplification plan, doc routing, and lag badges.

  • Document cutover.

  • Fix badge.

readme

  • Flesh out the Flavors table, document it in series-open (#2432).

readme

  • Add a Flavors section with lag badges and a Documentation root.

Refactoring

series

  • Read the vendored SHA one way, in one helper — and say what decides.

Uncategorized

  • Test(rcc): Stop git gc racing the harness's own clones, and say why a commit checks itself (#93).

duckdb 1.5.5.9002

Bug fixes

rcc

  • Make a verdict's newest writer win, not its slowest.

  • Let series-cutover.sh create a base series that does not exist yet (#2428).

rcc

  • Make writes to the rcc branch survive a lost race.

rcc-one

  • Restore the check action's CRAN-incoming default.

vendor-one

  • Make the glue gate work again.

Features

vendoring

  • Raise the vendoring limit from 25 to 100 commits.

rcc

  • Publish each commit's result as its own file, and consolidate the branch by hand.

each

  • Number shards along the history, and quote failing stages into the run summary (#2427).

each

  • Rerun one commit with a retry pair, not an amend.

each

  • Build every commit as a sharded matrix instead of one run per commit.

Continuous integration

  • Replace workflow vendoring with the routine-driven series loop.

  • Harvest rcc logs every 30 minutes.

Documentation

  • Space.

skills

  • Reset -fwd-green to the seed on every rebase.

skills

  • Split rebasing a forward series out of forwarding one.

  • A WIP forward series can always be rebased onto current mainline.

  • Agentic-loop plan refinements + main-dev review.

  • Describe the forbidden forms in AGENTS.md instead of spelling them.

vendoring

  • Dev-branch invariants, the fork-point rule, and a corrected VENDORING.md.

Refactoring

each

  • Anchor a retry on the series' green, not a ref of its own.

series-forward

  • Replay by cherry-pick instead of rebuilding trees.

  • Rename the flavor rename from "lts" to "flavor".

Performance

each

  • Rebalance shards across waves instead of capping at max-parallel.

each

  • Split large shards to trade compute for wall clock.

Testing

  • Derive the expected storage home from the package name.

  • Make storage and extension snapshots stable across flavors.

  • Skip the httpfs e2e test on development versions.

duckdb 1.5.5.9001

Features

  • New tbl_file() and tbl_query() to explicitly access tables and queries as dbplyr lazy tables (#96).

Continuous integration

  • Sync R-CMD-check workflow; drop matrix-env plumbing from custom action (#2422).

Testing

  • Guard the flavor rename against hard-coded package names (#2423).

Uncategorized

  • Ci: Run all smoke-test checks even when one fails (#97).

duckdb 1.5.5.9000

Continuous integration

  • Lock down format-suggest egress (audit → block).

fledge

  • CRAN release v1.5.5 (#2420).

duckdb 1.5.5

Features

  • Update to DuckDB v1.5.5, see https://github.com/duckdb/duckdb/releases/tag/v1.5.5 for details.

  • Disable extensions on libc++ Linux builds, with startup message and load error (#1107, #2414).

  • Rework how extensions and secrets choose their on-disk location, replacing the package-library extension cache attempted across 1.5.4.1 through 1.5.4.3. Both now share a single "home" root, resolved fresh for every duckdb() driver. An interactive prompt in duckdb() and messages in non-interactive sessions guide the process. See ?duckdb_storage for details. (#2396, #2398).

Bug fixes

  • Support OR and IN predicates pushed down to registered Arrow tables (#2410).

  • Error instead of crashing when calling duckdb_fetch_arrow() on an Arrow result that has been consumed with duckdb_fetch_record_batch() (#2406).

  • Fix edge case in computation of r_base::min/r_base::max/r_base::sum aggregates (#2404, #2405).

  • Honor the na.rm argument of r_base aggregates in SQL queries (#2407).

Documentation

  • Document database-instance caching and driver reuse (#2399).

duckdb 1.5.4.3

Bug fixes

  • Remove the package-library extension storage option. The duckdb_extension_storage() function no longer accepts "library" (#2390).

duckdb 1.5.4.2

Bug fixes

  • Fix shared on-disk storage path on Windows (#2385).

duckdb 1.5.4.1

Features

  • DuckDB's on-disk storage locations now follow a unified policy. By default nothing is written outside the R session's temporary directory, with one exception: the extension cache is placed in the package library when it is writable and falls back to the temporary directory otherwise. Each location can be redirected through the config argument of duckdb(), an R option, or an environment variable. Configure the location for extensions and secrets with the new duckdb_extension_storage() and duckdb_secret_storage(), inspect the resolved locations with duckdb_storage_status(), and see ?duckdb_storage for the full resolution policy (#2370, #2372, #2377).

    These functions replace the experimental duckdb_consolidate_secrets() introduced in 1.5.4.

Continuous integration

  • Create ~/.duckdb so CI exercises the shared storage path (#2417).

  • Install tzdata-legacy for legacy time zones like PST8PDT (#2409).

Documentation

  • Document all S4 class slots; deprecate the legacy connection slots (#2416).

  • Document database-instance caching and driver reuse (#2399).

Testing

  • Cover extensions env-var semantics and lock message wording (#2418).

  • Skip DBItest timestamp roundtrip tests when PST8PDT is unresolvable (#2408).

Uncategorized

  • Ci: Harden format-suggest against pull_request_target pwn requests (#93).

duckdb 1.5.4

Features

  • Update to DuckDB v1.5.4, see https://github.com/duckdb/duckdb/releases/tag/v1.5.4 for details.

  • Support writing MAP columns via dbAppendTable() and dbWriteTable() (#2354).

  • Add native VARIANT (@thohan88, #2313) and TIME WITH TIME ZONE (#1807, #2336) data type support.

  • Implement DBI Arrow API with dbSendQueryArrow() and streaming (#2347, #2355).

  • Store downloaded extensions inside the duckdb package install directory (#2327).

  • Add secret directory configuration, package startup message, and consolidation support via new experimental duckdb_consolidate_secrets() (#2305, #2340).

Compatibility

  • Add is_distinct_from() / is_not_distinct_from() dbplyr translations for compatibility with upcoming dbplyr 2.6.0 (#2326, #2332).

  • Bump minimum R version requirement to 4.2.0 (#2233, #2334).

Testing

  • Add CRAN guards to prevent heavy C++ engine tests on CRAN (#2353, #2358).

  • Add comprehensive test coverage for MAP type reading (#2342).

duckdb 1.5.3

Features

  • Update to DuckDB v1.5.3, see https://github.com/duckdb/duckdb/releases/tag/v1.5.3 for details.

  • Add secret directory configuration, package startup message, and consolidation support via new experimental duckdb_consolidate_secrets() (#2305, #2340).

  • Add native VARIANT (@thohan88, #2313) and TIME WITH TIME ZONE (#1807, #2336) data type support.

  • Add is_distinct_from() / is_not_distinct_from() dbplyr translations for compatibility with upcoming dbplyr 2.6.0 (#2326, #2332).

Bug fixes

  • Avoid rchk error in RownamesDuplicate() (#2290, #2291).

Chore

  • Bump minimum R version requirement to 4.2.0 (#2233, #2334).

  • Store downloaded extensions inside the duckdb package install directory (#2327).

Testing

  • Add comprehensive test coverage for MAP type reading (#2342).

duckdb 1.5.2

Bug fixes

Features

  • Use TRY_CAST() instead of CAST() in dplyr SQL translation for type conversion functions (#2230, #2231).

Chore

  • Use R_getRegisteredNamespace() in R 4.6.

Documentation

  • Describe branching strategy (#2280, #2281).

Testing

  • Use explicit default duckdb connection for arrow tests (#2301).

  • Rework arrow tests, prepare for compatibility with dbplyr 2.6.0 (#2300).

duckdb 1.5.1

Bug fixes

Features

  • GEOMETRY columns can be returned, either as BLOBs (default) or as wk objects (via the wk package) using dbConnect(geometry = "wk") (#2278, #2279).

Chore

  • Fix -Wdeprecated compiler warnings (#2295, #2296) and protection buglet (#2294).

  • Use gtar when available to suppress Apple extended attribute warnings on Linux (#2227, #2228).

duckdb 1.5.0

Features

Bug fixes

  • The dbplyr translation of as.numeric() and as.double() uses DOUBLE instead of NUMERIC (#2023, #2031).

Testing

  • Update to testthat edition 3.

Internal

  • Avoid ATTRIB() for compatibility with R 4.6, materialize ALTREP row names to integer sequence with full ALTREP methods (#2034).

duckdb 1.4.4

Features

Chore

  • Bump vendored cpp11 to v0.5.3.

Documentation

  • Add alternative installation method to README (@szarnyasg, #1819).

duckdb 1.4.3

Features

Bug fixes

  • Fail with non-UTF8-encoded strings during data frame scan instead of attempting to reencode (#1795).

  • Avoid inclusion of raw error message in the output.

  • Fix translation of quantile() to use DuckDB's native QUANTILE_CONT() syntax (#1734, #1735).

Testing

  • Remove redundant R version checks from tests (#1815, #1816).

duckdb 1.4.2

duckdb 1.4.1

Features

  • Add support for wildcards in tbl_file() paths (#1614, @rplsmn).

  • Add n_distinct(..., na.rm = TRUE) support for multiple passed columns (@lschneiderbauer, #1588).

Bug fixes

  • Fix Valgrind error.

Testing

  • Ensure be able to install duckdb extensions on release version (@eitsupi, #1586).

duckdb 1.4.0

Features

  • New experimental sql_query(), sql_exec(), and default_conn() to simplify the most important operations for interactive use (#1564).

  • tbl_file() allows omitting the src argument, falling back to the default connection.

  • Full support for deep structs generated by struct_pack() for ALTREP (#1545).

Bug fixes

  • Fix progress display for fractional progress values (#1499, #1505).

  • Extensions can be installed again.

duckdb 1.3.3

Bug fixes

  • Fix timezone conversion for invalid timestamps with tz_out_convert = "force" (#1474).

  • Substitute invalid UTF-8 characters in error messages to avoid a failure when reporting the error.

  • Fix index calculation for retrieval of arrays (#1473).

  • Fix conversion for retrieval of large enums.

  • Fix compiler error in debug build (@joakimlinde, #1368).

Features

  • Add rich ErrorData-based error handling with structured error information (#1479).

  • Safeguard against deadlocks when accidentally issuing queries from the progress bar handler or other callbacks (#1475).

  • dbGetInfo() gets the version from a hard-coded value and not from a DuckDB query (#1481).

  • Package uses two cores by default for compilation (#1478).

Documentation

  • Document vendoring process and main/next branch relationship (#1488).

Testing

  • Add local_con() test fixture for cleaner DuckDB connection management (#1476).

duckdb 1.3.2

Features

duckdb 1.3.1

Features

Bug fixes

  • Correct dbplyr translations for str_starts() and str_ends() (#1182, #1247).

  • Fix multiarch build on R 4.1 for Windows.

duckdb 1.3.0

Features

Chore

  • Fix rchk (#1173).

  • Fix compiler warning (@joakimlinde, #1172).

Testing

  • Skip timing tests on CRAN.

duckdb 1.2.2

Features

  • Update to duckdb v1.2.2, see https://github.com/duckdb/duckdb/releases/tag/v1.2.2 for details.

  • Add support for duckdb arrays in R (@joakimlinde, #102, #1090). To enable, connect with dbConnect(duckdb(), array = "matrix") (@joakimlinde, #1125).

  • Support fractional seconds in TIME and INTERVAL data (#1109).

  • The autoload_known_extensions configuration option is now enabled by default (#582, #1084, #1134).

  • Mention column name for conversion errors (#1108).

Chore

  • Require R >= 4.1 (#1087, #1133).

  • Types exposed through ALTREP are the same as through DBI (#1111), including STRUCT. This enables support more types in upcoming duckplyr versions.

  • Perform optional checks for ALTREP compatibility in rel_from_df() and expr_constant() (#1117).

  • Perform time zone conversion in the C++ layer where possible, to support ALTREP (#1130).

  • Improve developer experience: pkgload::load_all() now works, source files are rebuilt if header files change, configure clangd (#1128).

  • Add dots with checks to unexported functions (#1115).

  • Clean up edge case for fetching zero rows (#1104).

  • Avoid test for timings on CRAN (#1101).

Documentation

  • Tweak README.

duckdb 1.2.1

Features

Bug fixes

  • dbExecute(con, "CALL ...") no longer attempts to access the resulting data frame. Use dbGetQuery(con, "CALL ...") to access the data (#1062, #1080).

  • Fix support for the connections pane in RStudio and Positron (@dfalbel, #1063).

Internal

  • New rel_to_view() (#1075).

  • New internal AltrepDataframeRelation, used with rel_from_altrep_df(wrap = TRUE) (#949, #1072).

  • Try relational materialization only once (#1066).

Chore

  • Update vendored cpp11 to 0.5.2 (#1068).

  • Avoid calls to non-API R functions.

duckdb 1.2.0

Breaking changes

  • Breaking change: Remove substrait API: duckdb_get_substrait(), duckdb_get_substrait_json(), duckdb_prepare_substrait(), duckdb_prepare_substrait_json() (@pdet, #1021).

Features

  • Update to duckdb v1.2.0, see https://github.com/duckdb/duckdb/releases/tag/v1.2.0 for details.

  • Progress is shown for slow operation. This is on by default in interactive mode and can be controlled by setting the "duckdb.progress_display" option to a logical scalar (#199, #951, @meztez).

  • Add translation for median() (@toppyy, #993, #1011).

  • Floor sub-day precision date before casting to int (@toppyy, #517, #981).

  • Set value returned by PRAGMA user_agent to r-dbi (#707, @elefeint).

Bug fixes

  • Remove unconditional use of CPPHTTPLIB_USE_POLL to support compilation with R 4.0 and R 4.1 again (@Antonov548, #1043).

  • Support reading from multiple Parquet files again (#1015, #1024).

  • Fix translation for add_days() and add_years() clock functions (#976, @IoannaNika).

duckdb 1.1.3-2

Bug fixes

  • Make cleanup truly idempotent (#612, #940).

Chore

  • Sync vendoring script with igraph (#936).

Features

  • Limit automatic materialization by number of rows or number of cells (#1017).

  • New internal rapi_rel_to_csv(),rapi_rel_to_table(), and rapi_rel_insert(); rapi_rel_to_parquet() gains options argument (#867).

Testing

  • Skip tests that are about to fail.

  • Sync tests.

duckdb 1.1.3-1

Features

  • With duckdb(environment_scan = TRUE), data frame objects are available as views in duckdb SQL queries (#140, #164).

  • Update vendored cpp11 to 0.5.1 (#636).

Bug fixes

  • Make ./cleanup script reentrant (@Antonov548, #612, #634).

  • Fix installation of extensions (#623).

  • Fix rchk and UB errors (#635).

  • Avoid loading rlang during startup (#601).

Documentation

  • Mention xz requirement in DESCRIPTION.

duckdb 1.1.3

Features

  • Update to duckdb v1.1.3, see https://github.com/duckdb/duckdb/releases/tag/v1.1.3 for details.

  • New duckdb.materialize_callback option, supersedes get_last_rel() (#589).

  • New rel_explain_df() and rel_tostring() (#587).

  • Handle empty child values for list constants (#186, @romainfrancois).

Chore

  • Undef TRUE and FALSE (#595).

  • Remove enable_materialization argument to rel_from_altrep_df() in favor of creating a new data frame when needed (#588).

  • Flip argument order for expr_comparison() (#585).

  • Keep cleanup files to accommodate different build scenarios (#536).

duckdb 1.1.2

Features

Features

  • Long-running queries can now be canceled immediately with Ctrl + C (terminal) or Escape (RStudio IDE and Workbench) (#514, #515).

  • Add col.types argument to duckdb_read_csv() (#445, @eli-daniels).

  • Rethrow errors with rlang if installed (#522).

  • Improve error message for parsing erros during statement extraction (tidyverse/duckplyr#219, #521).

Bug fixes

  • Avoid RStudio IDE crashes when ending session with open objects (#520).

  • rfuns extension: %in% works correctly as part of a & conjunction (#528).

Internal

  • New interal APIs: rapi_get_last_rel_mat(), rapi_rel_to_altrep(allow_materialization = TRUE), rapi_rel_from_altrep_df(enable_materialization) (#526).

  • xz-compress duckdb sources in the tarball (#530).

  • rfuns extension: Fix signedness.

duckdb 1.1.1

Features

  • Update to duckdb v1.1.1, see https://github.com/duckdb/duckdb/releases/tag/v1.1.1 for details.

  • Add comparison expression to relational API (@toppyy, #457).

  • Temporarily change max_expression_depth during ALTREP evaluation (#101, #460).

  • Add temporary argument to duckdb_read_csv() (@ThomasSoeiro, #223).

Chore

  • Update vendored extension sources to hannes/duckdb-rfuns@20cde009b51b9355e6041b72b87105c6b45793fe.

  • Remove warnings for uninitialized variables.

duckdb 1.1.0

Features

duckdb 1.0.0-2

Features

  • Reduce the package installation size on macOS (#185).

duckdb 1.0.0-1

Bug fixes

  • Upgrade vendored cpp11 to 0.4.7 to fix compilation with R-devel.

  • Support dplyr::tbl(conn, I(...)).

duckdb 1.0.0

Bug fixes

duckdb 0.10.3

Features

  • Update to duckdb v0.10.3, see https://github.com/duckdb/duckdb/releases/tag/v0.10.3 for details.
  • Support fetching MAP type (#61, #165).
  • Add dbplyr translations for clock::date_count_between() (@edward-burn, #163, #166).
  • round() duckdb translation uses ROUND_EVEN() instead of ROUND() (@lschneiderbauer, #146, #157).
  • New sort argument to rel_order() (@toppyy, #168).
  • Add dbplyr translations for clock::add_days(), clock::add_years(), clock::get_day(), clock::get_month(), and clock::get_year() (@edward-burn, #153).

Bug fixes

  • Correct usage of win_current_group() instead of win_current_order() in SQL translation (@lschneiderbauer, #173, #175).

duckdb 0.10.2

Features

Bug fixes

  • rel_sql(rel, "{{sql}}") works even on a read-only database (@Tmonster, #138).
  • Avoid R CMD check warning regarding SETLENGTH() and SET_TRUELENGTH() (#145).

duckdb 0.10.1

Features

  • Update to duckdb v0.10.1, see https://github.com/duckdb/duckdb/releases/tag/v0.10.1 for details.
  • Fix shutdown semantics for the driver object created by duckdb(). A database file is closed (and available to be opened from another session) after the last connection that uses this file calls dbDisconnect() . The shutdown argument to dbDisconnect() or the duckdb_shutdown() functions are no longer necessary. Two database connections from the same R session can access the same file concurrently in read-write mode (#124).

Bug fixes

  • Don't run tests that invoke re2 by default (#121, #127).

  • Fix compilation for R 4.0 and R 4.1, regression introduced in v0.10.0. Using librstrtmgr.a from UCRT build of rtools40 (#130).

Internal

  • The C++ core is now vendored commit by commit, once every five minutes. Vendoring stops if R CMD check fails or if a previously unreleased tag is reached.

  • New maintainer: Kirill Müller.

Continuous integration

  • Add rhub2 workflow.

duckdb 0.10.0

Bug fixes

  • dplyr::tbl() works again when a Parquet or CSV file is passed instead of a table name (#38, #91).

  • DBI::dbQuoteIdentifier() correctly quotes identifiers that start with a digit (#67, #92).

  • Align the argument order of dbWriteTable() with the DBI specs (@eitsupi, #43, #49).

Features

  • New tbl_file() and tbl_query() to explicitly access tables and queries as dbplyr lazy tables (#96). The cache argument to tbl() and to the new functions must be named.

  • Initial ALTREP support for LIST logical type (@romainfrancois, #77).

  • Update core to duckdb v0.10.0 (#90).

  • New private rel_to_parquet() to write a relation to parquet (@Tmonster, #46).

Chore

  • Change directory location for extensions and secrets for v.0.10.0 release (@Tmonster, #73).

  • Remove last instance of default_connection() (#50).

Documentation

  • Add list of contributors (#2, #94).

  • Use pkgdown BS5 (@maelle, #31, #70) with DuckDB logo (#76, @romainfrancois).

  • Link to R documentation page.

  • Include NEWS.md on CRAN (#48, @olivroy).

Testing

  • Add csv reading test for duckdb_read_csv(na.strings = ) (@Tmonster, #10).

  • Fix snapshot tests.

  • Tweak tests for compatibility with v0.10.0 (#84).

duckdb 0.9.2-1

  • Fix compiler warning on R-devel (#45).

duckdb 0.9.2

duckdb 0.9.1-1

  • Fix LTO checks on CRAN.

duckdb 0.9.1

duckdb 0.8.1-3

  • Internal changes to support the duckplyr package.

duckdb 0.8.1-2

  • Compatibility with dbplyr.

  • Internal changes to support the duckplyr package.

duckdb 0.8.1-1

  • Fix CRAN checks.

duckdb 0.8.1

duckdb 0.8.0

duckdb 0.7.1-1

duckdb 0.7.0

duckdb 0.6.2

  • New duckdb_prepare_substrait_json().

duckdb 0.6.1

duckdb 0.6.0

duckdb 0.5.1

duckdb 0.5.0

duckdb 0.4.0

duckdb 0.3.4-1

  • Minor changes for CRAN compatibility.

duckdb 0.3.4

duckdb 0.3.3

duckdb 0.3.2

duckdb 0.3.1

duckdb 0.3.0

duckdb 0.2.8

This preview release of DuckDB is named "Ceruttii" after a long-extinct relative of the present-day Harleqin Duck (Histrionicus Ceruttii). Binary builds are listed below. Feedback is very welcome.

Note: Again, this release introduces a backwards-incompatible change to the on-disk storage format. We suggest you use the EXPORT DATABASE command with the old version followed by IMPORT DATABASE with the new version to migrate your data. See the documentation for details.

SQL

  • #2064: RANGE/GENERATE_SERIES for timestamp + interval
  • #1905: Add PARQUET_METADATA and PARQUET_SCHEMA functions
  • #2059, #1995, #2020 & #1960: Window RANGE framing, NTH_VALUE and other improvements

APIs

  • Many Arrow integration improvements
  • Many ODBC driver improvements
  • #1815: Initial version: SQLite UDF API
  • #2001: Support DBConfig in C API

Engine

  • #1975, #1876 & #2009: Unified row layout for sorting, aggregate & joins
  • #1930 & #1904: List Storage
  • #2050: CSV Reader/Casting Framework Refactor & add support for TRY_CAST
  • #1950: Add Constant Segment Compression to Storage
  • #1957: Add pipe/stream file system

duckdb 0.2.7

This preview release of DuckDB is named "Mollissima" after the Common Eider (Somateria mollissima). Binary builds are listed below. Feedback is very welcome.

Note: This release introduces a backwards-incompatible change to the on-disk storage format. We suggest you use the EXPORT DATABASE command with the old version followed by IMPORT DATABASE with the new version to migrate your data. See the documentation for details.

Major changes:

SQL

  • #1847: Unify catalog access functions, and provide views for common PostgreSQL catalog functions
  • #1822: Python/JSON-Style Struct & List Syntax
  • #1862: #1584 Implementing NEXTAFTER for float and double
  • #1860: FIRST implementation for nested types
  • #1858: UNNEST table function & array syntax in parser
  • #1761: Issue #1746: Moving QUANTILE

APIs

  • #1852, #1840, #1831, #1819 and #1779: Improvements to Arrow Integration
  • #1843: First iteration of ODBC driver
  • #1832: Add visualizer extension
  • #1803: Converting Nested Types to native python
  • #1773: Add support for key/value style configuration, and expose this in the Python API

Engine

  • #1808: Row-Group Based Storage
  • #1842: Add (Persistent) Struct Storage Support
  • #1859: Read and write atomically with offsets
  • #1851: Internal Type Rework
  • #1845: Nested join payloads
  • #1813: Aggregate Row Layout
  • #1836: Join Row Layout
  • #1804: Use Allocator class in buffer manager and add a test for a custom allocator usage

duckdb 0.2.6

This preview release of DuckDB is named "Jamaicensis" after the blue-billed Ruddy Duck (Oxyura jamaicensis). Binary builds are listed below. Feedback is very welcome.

Note: This release introduces a backwards-incompatible change to the on-disk storage format. We suggest you use the EXPORT DATABASE command with the old version followed by IMPORT DATABASE with the new version to migrate your data. See the documentation for details.

Also note: Due to changes in the internal storage (#1530), databases created with this release wil require somewhat more disk space. This is transient as we are working hard to finalise the on-disk storage format.

Major changes:

Engine

  • #1666: External merge sort, #1580: Parallel scan of ordered result and #1561: Rework physical ORDER BY
  • #1520 & #1574: Window function computation parallelism
  • #1540: Add table functions that take a subquery parameter
  • #1533: Using vectors, instead of column chunks as lists
  • #1530: Store null values separate from main data in a Validity Segment

SQL

  • #1568: Positional Reference Operator #1 etc.
  • #1671: QUANTILE variants and #1685: Temporal quantiles
  • #1695: New Timestamp Types TIMESTAMP_NS, TIMESTAMP_MS and TIMESTAMP_NS
  • #1647: Add support for UTC offset timestamp parsing to regular timestamp conversion
  • #1659: Add support for USING keyword in DELETE statement
  • #1638, #1663, #1621 & #1484: Many changes arount ARRAY syntax
  • #1610: Add support for CURRVAL
  • #1544: Add SKIP as an option to READ_CSV and COPY

APIs

  • #1525: Add loadable extensions support
  • #1711: Parallel Arrow Scans
  • #1569: Map-style UDFs for Python API
  • #1534: Extensible Replacement Scans & Automatic Pandas Scans and #1487: Automatically use parquet or CSV scan when using a table name that ends in .parquet or .csv
  • #1649: Add a QueryRelation object that can be used to convert a query directly into a relation object, #1665: Adding from_query to python api
  • #1550: Shell: Add support for Ctrl + arrow keys to linenoise, and make Ctrl+C terminate the current query instead of the process
  • #1514: Using ALTREP to speed up string column transfer to R
  • #1502: R: implementation of Rstudio connection-contract tab

duckdb 0.2.5

This preview release of DuckDB is named "Falcata" after the Falcated Duck (Mareca falcata). Binary builds are listed below. Feedback is very welcome.

Note: This release introduces a backwards-incompatible change to the on-disk storage format. We suggest you use the EXPORT DATABASE command with the old version followed by IMPORT DATABASE with the new version to migrate your data. See the documentation for details.

Major Changes:

Engine

  • #1356: Incremental Checkpointing
  • #1422: Optimize Top N Implementation

SQL

  • #1406, #1372, #1387: Many, many new aggregate functions
  • #1460: QUANTILE aggregate variant that takes a list of quantiles & #1346: Approximate Quantiles
  • #1461: JACCARD, #1441 LEVENSHTEIN & HAMMING distance scalar function
  • #1370: FACTORIAL scalar function and ! postfix operator
  • #1363: IS (NOT) DISTINCT FROM
  • #1385: LIST_EXTRACT to get a single element from a list
  • #1361: Aliases in the HAVING clause (fixes issue #1358)
  • #1355: Limit clause with non constant values

APIs:

  • #1430 & #1424: DuckDB WASM builds
  • #1419: Exporting the appender api to C
  • #1408: Add blob support to C API
  • #1432, #1459 & #1456: Progress Bar
  • #1440: Detailed profiler.

duckdb 0.2.4

This preview release of DuckDB is named "Jubata" after the Australian Wood Duck (Chenonetta jubata). Binary builds are listed below. Feedback is very welcome.

Note: This release introduces a backwards-incompatible change to the on-disk storage format. We suggest you use the EXPORT DATABASE command with the old version followed by IMPORT DATABASE with the new version to migrate your data. See the documentation for details.

Major changes: SQL

  • #1231: Full Text Search extension
  • #1309: Filter Clause for Aggregates
  • #1195: SAMPLE Operator
  • #1244: SHOW select queries
  • #1301: CHR and ASCII functions & #1252: Add GAMMA and LGAMMA functions

Engine

  • #1211: (Mostly) Lock-Free Buffer Manager
  • #1325: Unsigned Integer Types Support
  • #1229: Filter Pull Up Optimizer
  • #1296: Optimizer that removes redundant DELIM_GET and DELIM_JOIN operators
  • #1219: DATE, TIME and TIMESTAMP rework: move to epoch format & microsecond support

Clients

  • #1287 and #1275: Improving JDBC compatibility
  • #1260: Rework client API and prepared statements, and improve DuckDB -> Pandas conversion
  • #1230: Add support for parallel scanning of pandas data frames
  • #1256: JNI appender
  • #1209: Write shell history to file when added to allow crash recovery, and fix crash when .importing file with invalid
  • #1204: Add support for blobs to the R API and #1202: Add blob support to the python api

Parquet

  • #1314: Refactor and nested types support for Parquet Reader

duckdb 0.2.3

This preview release of DuckDB is named "Serrator" after the Red-breasted merganser (Mergus serrator). Binary builds are listed below. Feedback is very welcome.

Note: This release introduces a backwards-incompatible change to the on-disk storage format. We suggest you use the EXPORT DATABASE command with the old version followed by IMPORT DATABASE with the new version to migrate your data. See the documentation for details.

Major changes:

SQL:

  • #1179: Interval Cleanup & Extended INTERVAL Syntax
  • #1147: Add exact MEDIAN and QUANTILE functions
  • #1129: Support scalar functions with CREATE FUNCTION
  • #1137: Add support for (NOT) ILIKE, and optimize certain types of LIKE expressions

Engine

  • #1160: Perfect Aggregate Hash Tables
  • #1133: Statistics Rework & Statistics Propagation
  • #1144: Common Aggregate Optimizer, #1143: CSE Optimizer and #1135: Optimizing expressions in grouping keys
  • #1138: Use predication in filters
  • #1071: Removing string null termination requirement

Clients

  • #1112: Add DuckDB node.js API
  • #1168: Add support for Pandas category types
  • #1181: Extend DuckDB::LibraryVersion() to output dev version in format 0.2.3-devXXX & #1176: Python binding: Add module attributes for introspecting DuckDB version

Parquet Reader:

  • #1183: Filter pushdown for Parquet reader
  • #1167: Exporting Parquet statistics to DuckDB
  • #1162: Add support for compression codec in Parquet writer & #1163: Add ZSTD Compression Code and add ZSTD codec as option for Parquet export
  • #1103: Add object cache and Parquet metadata cache

duckdb 0.2.2

This is a preview release of DuckDB. Starting from this release, releases get named as well. Names are chosen from species of ducks (of course). We start with "Clypeata".

Note: This release introduces a backwards-incompatible change to the on-disk storage format. We suggest you use the EXPORT DATABASE command with the old version followed by IMPORT DATABASE with the new version to migrate your data. See the documentation for details.

Binary builds are listed below. Feedback is very welcome. Major changes:

SQL

  • #1057: Add PRAGMA for enabling/disabling optimizer & extend output for query graph
  • #1048: Allow CTEs in subqueries (including CTEs themselves) and #987: Allow CTEs in CREATE VIEW statements
  • #1046: Prettify Explain/Query Profiler output
  • #1037: Support FROM clauses in UPDATE statements
  • #1006: STRING_SPLIT and STRING_SPLIT_REGEX SQL functions
  • #1000: Implement MD5 function
  • #936: Add GLOB support to Parquet & CSV readers
  • #899: Table functions information_schema_schemata() and information_schema_tables() and #903: Add table function information_schema_columns()

Engine

  • #984: Parallel grouped aggregations and #1045: Some performance fixes for aggregate hash table
  • #1008: Index Join
  • #991: Local Storage Rework: Per-morsel version info and flush intermediate chunks to the base tables
  • #906: Parallel scanning of single Parquet files and #982: ZSTD Support in Parquet library
  • #883: Unify Table Scans with Table Functions
  • #873: TPC-H Extension
  • #884: Remove NFC-normalization requirement for all data and add COLLATE NFC

Client

  • #1001: Dynamic Syntax Highlighting in Shell
  • #933: Upgrade shell.c to 3330000
  • #918: Add in support for Python datetime types in bindings
  • #950: Support dates and times output into arrow
  • #893: Support for Arrow NULL columns

duckdb 0.2.1

This is a preview release of DuckDB. Binary builds are listed below. Feedback is very welcome. Major changes:

Engine

  • #770: Enable Inter-Pipeline Parallelism
  • #835: Type system updates with #779: INTERVAL Type, #858: Fixed-precision DECIMAL types & #819: HUGEINT type
  • #790: Parquet write support

API

  • #866: Initial Arrow support
  • #809: Aggregate UDF support with #843: Generic CreateAggregateFunction() & #752: CreateVectorizedFunction() using only template parameters

SQL

  • #824: strftime and strptime
  • #858: EXPORT DATABASE and IMPORT DATABASE
  • #832: read_csv(_auto) improvements: optional parameters, configurable sample size, line number info

duckdb 0.2.0

This is a preview release of DuckDB. Binary builds are listed below. Feedback is very welcome.

SQL:

  • #730: FULL OUTER JOIN Support
  • #732: Support for NULLS FIRST/NULLS LAST
  • #698: Add implementation of the LEAST/GREATEST functions
  • #772: Implement TRIM function and add optional second parameter to RTRIM/LTRIM/TRIM
  • #771: Extended Regex Options

Clients:

  • Python: #720: Making Pandas optional and add support for PyPy
  • C++: #712: C++ UDF API

duckdb 0.1.9

This is a preview release of DuckDB. Binary are listed below. Feedback is very welcome. Major changes: New website woo-ho!

Engine

  • #653: Parquet reader integration

SQL

  • #685: Case insensitive binding of column names
  • #662: add EPOCH_MS function and test cases

Clients

  • #681: JDBC Read-only mode for and #677 duplicate()` method to allow multiple connections to same database

duckdb 0.1.8

This is a preview release of DuckDB. Feedback is very welcome.

SQL

  • SQL functions IF and IFNULL #644
  • SQL string functions LEFT #620 and RIGHT #631
  • #641: BLOB type support
  • #640: LIKE escape support

Clients

  • #627: Insertion support for Python relation API

duckdb 0.1.7

This is the sixth preview release of DuckDB. Feedback is very welcome. Binary builds are available as well.

SQL

Engine

  • LIKE optimization rules #559
  • Adaptive filters in table scans #574
  • ICU Extension for extended Collations & Extension Support #594
  • Extended zone map support in scans #551
  • Disallow NaN/INF in the system #541
  • Use UTF Grapheme Cluster Breakers in Reverse and Shell #570

Clients

  • Relation API for C++ #509 and Python #598
  • Java (TM) JDBC (R) Client for DuckDB #492 #520 #550

duckdb 0.1.6

This is the fifth preview release of DuckDB. Feedback is very welcome. Binary builds can be found here: http://download.duckdb.org/alias/v0.1.6/

SQL

  • #455 Table renames ALTER TABLE tbl RENAME TO tbl2
  • #457 Nested list type can be created using LIST aggregation and unpacked with the new UNNEST operator
  • #463 INSTR string function, #477 PREFIX string function, #480 SUFFIX string function

Engine

  • #442 Optimized casting performance to strings
  • #444 Variable return types for table-producing functions
  • #453 Rework aggregate function interface
  • #474 Selection vector rework
  • #478 UTF8 NFC normalization of all incoming strings
  • #482 Skipping table segments during scan based on min/max indices

Python client

  • #451 date / datetime support
  • #467 description field for cursor
  • #473 Adding read_only flag to connect
  • #481 Rewrite of Python API using pybind11

R client

  • #468 Support for prepared statements in R client
  • #479 Adding automatic CSV to table function read_csv_duckdb
  • #483 Direct scan operator for R data.frame objects

duckdb 0.1.5

This is the fourth preview release of DuckDB. Feedback is very welcome. Note: The v0.1.4 version was skipped because of a Python packaging issue.

Binary builds can be found here: http://download.duckdb.org/rev/59f8907b5d89268c158ae1774d77d6314a5c075f/

Major changes:

  • #409 Vector Overhaul
  • #423 Remove individual vector cardinalities
  • #418 DATE_TRUNC SQL function
  • #424 REVERSE SQL function
  • #416 Support for SELECT table.* FROM table
  • #414 STRUCT types in query execution
  • #431 Changed internal string representation
  • #433 Rename internal type index_t to idx_t
  • #439 Support for temporary structures in read-only mode
  • #440 Builds on Solaris & OpenBSD

Note: This release contains a bug in the Python API that leads to crashes when fetching strings to NumPy/Pandas #447

duckdb 0.1.3

This is the third preview release of DuckDB. Feedback is very welcome. Binary builds can be found here: http://download.duckdb.org/rev/59f8907b5d89268c158ae1774d77d6314a5c075f/

Major changes:

  • #388 Major updates to shell
  • #390 Unused Column & Column Lifetime Optimizers
  • #402 String and compound keys in indices/primary keys
  • #406 Adaptive reordering of filter expressions

duckdb 0.1.2

This is the third preview release of DuckDB. Feedback is very welcome. Binary builds can be found here: http://download.duckdb.org/rev/6fcb5ef8e91dcb3c9b2c4ca86dab3b1037446b24/

duckdb 0.1.1

This is the second preview release of DuckDB. Feedback is very welcome. Binary builds can be found here: http://download.duckdb.org/rev/2e51e9bae7699853420851d3d2237f232fc2a9a8/

duckdb 0.1.0

This is the first preview release of DuckDB. Feedback is very welcome.

Binary builds can be found here: http://download.duckdb.org/rev/c1cbc9d0b5f98a425bfb7edb5e6c59b5d10550e4/