26.12.1.dev2

Pre-release of 26.12.1: a preview built from the engine release named below, for trying the next release early.

What changed since 26.12.1.dev1

Other

• Dev release: Maven resolves the pinned timestamped snapshot; the facets example prints no varying count (#310)

Install

uv add "arcadedb==26.12.1.dev2"      # native build
uv add "arcadedb-jvm==26.12.1.dev2"  # JVM build

A pre-release is installed only when it is named exactly, as above, or with --prerelease allow.

Engine

ArcadeDB engine 26.11.1.
This is the engine the native arcadedb wheels link. The arcadedb-jvm preview wheels take the jars of the upstream image tag latest.

Packages

• arcadedb 26.12.1.dev2
• arcadedb-jvm 26.12.1.dev2

Full change list: 26.12.1.dev1...26.12.1.dev2

26.11.1

What changed in the milestone 26.11.1

Bug fixes

• native: save() refuses to overwrite a record changed after it was read (#194)
• Vertex.new_edge: resolve both ends by record id on the JVM build (#216)
• Native insert_many: date-times, Decimals, bytes, numpy scalars and pandas NA stay on the JSON bulk path; an uncarriable value moves its own row, not the chunk (#218)
• JVM async executor: define each JPype proxy class once per process, not once per callback (#241)
• Native: an RID built from an RID keeps the text, so a find_nearest hit can be saved (#266)

Performance

• Native find_nearest: send the query vector as one float32 array, not a Python list (#203)
• JVM read path: fewer crossings per call (get_raw, one, count, lookup_by_key, transaction mark) (#214)
• Native to_json_list: skip the walk for untagged results; RIDs and map keys JSON-safe (#223)
• Native GraphBatch: float32 vectors in a binary side buffer (arcadedb_n_gb_bin) (#224)
• Native transactions: no JSON decode for a status the caller does not read (#226)
• Native point operations: cut the repeated Python work around the one crossing (#230)
• JVM readers: a record or row to dict, the property names and result-set iteration in one crossing each (#233)
• Native append_samples: numeric columns cross by address, not as .tolist() and JSON text (#234)
• insert_many (JVM build): pause the cyclic garbage collector for a list or tuple input (#236)
• Native iter_json_batches reads the cursor one batch at a time (#242)
• JVM GraphBatch: list, dict and date-time properties stay on the JSON bulk path (#244)

Features

• Native find_nearest: include_vector=False returns hits without the indexed vector (explicit opt-in) (#204)

Documentation

• Copyright line of the bindings: Arcade Data Ltd and HumemAI (#105)
• Docs: native find_nearest sends the query as one float32 buffer (#203) (#250)
• Docs: add SECURITY.md and CONTRIBUTING.md (#256)
• After the 26.11.1 engine bump: example 29 on native again, NOTICE refreshed first (#262)
• Docs for 26.11.1: query language, find_nearest width and distance, read-only open, single-pass results, server mode, and fixes from the audit (#264)

Other

• CI: resemble ArcadeData/arcadedb automation (labels, milestones, in-progress, PR template, Dependabot merge) (#252)
• Dev engine bump: link ArcadeDB 26.11.1, dev line moves to 26.12.1.dev0 (#258)
• CI: assign-milestone sets a pull request's milestone (pull-requests: write) (#260)
• CI: a JVM release and a manual dispatch test all 16 platform x Python legs (#269)

Install

uv add arcadedb      # native build
uv add arcadedb-jvm  # JVM build

Engine

ArcadeDB engine 26.11.1.

Packages

• arcadedb 26.11.1
• arcadedb-jvm 26.11.1