Skip to content

Prepare release 1.2.0 #369

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
g-despot merged 1 commit into from
Aug 24, 2026
Merged

Prepare release 1.2.0 #369

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
105 changes: 91 additions & 14 deletions CHANGELOG.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,30 +9,105 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0

## [Unreleased]

---

## [1.2.0] — 2026-08-21

### Highlights

- **Boost**: a query-time soft-ranking rescorer available on every search and generate method (Weaviate ≥ 1.38)
- **Diversity selection (MMR)**: on hybrid search (Weaviate ≥ 1.38.6) and on every `near*` search (Weaviate ≥ 1.37.0)
- **`BM25Operator.AndCross`**: cross-property keyword AND, with a client-side server-version guard
- New modules: `multi2vec-twelvelabs` vectorizer and `generative-deepseek`
- New `Namespaces` RBAC permission scope for the 1.38 `manage_namespaces` action
- Incremental backups, plus new `text2vec-aws` / `text2vec-google` / `text2vec-openai` / `text2vec-morph` / `generative-google` settings
- Multimodal vectorizer weights now actually reach the wire — and the two Google multimodal factories were transposing video and audio fields
- **Not binary-compatible with 1.1.x.** 125 public signatures gained trailing optional parameters; recompile against 1.2.0 (see *Changed*)

### Added

- **`generative-deepseek`** — New `GenerativeConfigFactory.Deepseek(...)` collection config and `GenerativeProviderFactory.Deepseek(...)` runtime provider for the `generative-deepseek` module. Requires Weaviate ≥ 1.36.19.
- **Incremental Backups** — `BackupCreateRequest.IncrementalBaseBackupId` names an existing backup to build on, so files unchanged since that backup are not copied again; `Backup.IncrementalBaseBackupId` surfaces it on `List()` and `GetStatus()`. `Create` throws `WeaviateVersionMismatchException` below 1.37.0, the documented feature floor. The server has accepted the field on create since 1.34.18, but only returns it on read from 1.37.6.
- **`apiEndpoint` on `Multi2VecGoogle`** — Both `VectorizerFactory.Multi2VecGoogle(...)` overloads take an optional `apiEndpoint`, which selects the Gemini API (`generativelanguage.googleapis.com`) instead of Vertex AI. Both `Multi2VecGoogleGemini(...)` overloads also gained an optional `dimensions`.
#### Query and Generate

- **Boost — query-time soft ranking** ([#355](https://github.com/weaviate/weaviate-csharp-client/pull/355)) — New `Models.Boost` and an optional `boost` parameter on every BM25, Hybrid and `near*` search and generate method. A boost re-scores rather than filters: non-matching objects rank lower but are never dropped. Build one with the `Boost.*` factories. gRPC only. Requires Weaviate ≥ 1.38.
- **Diversity selection via MMR** ([#366](https://github.com/weaviate/weaviate-csharp-client/pull/366)) — New `Models.Diversity` with `Diversity.MMR(uint? Limit = null, float? Balance = null)`, passed as the optional `diversitySelection` parameter; `Balance` is the trade-off in `[0.0, 1.0]` (`1.0` pure relevance, `0.0` pure diversity). Not offered on aggregation. **Floors differ by surface**: `near*` searches require Weaviate ≥ **1.37.0**, hybrid search ≥ **1.38.6**.
- **`BM25Operator.AndCross`** ([#365](https://github.com/weaviate/weaviate-csharp-client/pull/365)) — New keyword operator requiring every query token to appear across the searched properties taken together rather than within one; all of them must share tokenization and analyzer settings. Landed in Weaviate **1.39.0**, backported to **1.37.15** and **1.38.8**, so the guard is per minor branch: **1.37.15** passes, **1.38.7** is refused with `WeaviateVersionMismatchException`.
- **`searchOperator` on `Generate.BM25`** ([#365](https://github.com/weaviate/weaviate-csharp-client/pull/365)) — All four `GenerateClient` / `TypedGenerateClient<T>` BM25 overloads now take `searchOperator`, which was previously accepted everywhere except here, leaving `AndCross` unreachable from generative keyword search. Same position and version guard as on `Query.BM25`.

#### Vectorizers

- **`multi2vec-twelvelabs`** ([#367](https://github.com/weaviate/weaviate-csharp-client/pull/367)) — New `Models.Vectorizer.Multi2VecTwelveLabs` config record and `VectorizerFactory.Multi2VecTwelveLabs(...)` overloads taking `string[]` or `WeightedFields` field lists. Requires Weaviate ≥ 1.38.9 on the 1.38 branch, or ≥ 1.39.0.
- **`dimensions` on `text2vec-aws`** ([#354](https://github.com/weaviate/weaviate-csharp-client/pull/354)) — `Text2VecAWS.Dimensions` and an optional `dimensions` on the `Text2VecAWSBedrock(...)` and `Text2VecAWSSagemaker(...)` factories, for models with a configurable embedding size.
- **`location` on `text2vec-google`** ([#353](https://github.com/weaviate/weaviate-csharp-client/pull/353)) — `Vectorizer.Text2VecGoogle.Location` and an optional `location` parameter on `VectorizerFactory.Text2VecGoogleVertex(...)`, selecting the Vertex AI region.
- **`endpoint` on `text2vec-openai` and `text2vec-morph`** ([#359](https://github.com/weaviate/weaviate-csharp-client/pull/359)) — `Endpoint` on both config records and an optional `endpoint` on the `Text2VecOpenAI(...)` and `Text2VecMorph(...)` factories, for proxied or self-hosted deployments.
- **`apiEndpoint` on `Multi2VecGoogle`** ([#368](https://github.com/weaviate/weaviate-csharp-client/pull/368)) — `VectorizerFactory.Multi2VecGoogle(...)` takes an optional `apiEndpoint` selecting the Gemini API (`generativelanguage.googleapis.com`) rather than Vertex AI; `Multi2VecGoogleGemini(...)` also gained an optional `dimensions`.

#### Generative

- **`generative-deepseek`** ([#368](https://github.com/weaviate/weaviate-csharp-client/pull/368)) — New `GenerativeConfigFactory.Deepseek(...)` collection config and `GenerativeProviderFactory.Deepseek(...)` runtime provider for the `generative-deepseek` module. Requires Weaviate ≥ 1.36.19.
- **`location` on `generative-google`** ([#359](https://github.com/weaviate/weaviate-csharp-client/pull/359)) — `Location` on `GenerativeConfig.GoogleVertex` and `Generative.Providers.GoogleVertex`, with an optional `location` on the matching factory, selecting the Vertex AI region.

#### RBAC

- **`Namespaces` Permission Scope** ([#356](https://github.com/weaviate/weaviate-csharp-client/pull/356)) — New `Models.Permissions.Namespaces` scope with a `Manage` flag and a `Models.NamespacesResource` filter (`Namespace`, default `"*"`, an exact name or a regex), mapping the Weaviate 1.38 RBAC action `manage_namespaces`. Requires Weaviate ≥ 1.38.0.

#### Backup

- **Incremental Backups** ([#368](https://github.com/weaviate/weaviate-csharp-client/pull/368)) — `BackupCreateRequest.IncrementalBaseBackupId` names an existing backup to build on, so unchanged files are not copied again, and is surfaced on `List()` and `GetStatus()`. `Create` throws `WeaviateVersionMismatchException` below 1.37.0; the field is accepted from 1.34.18 but only read back from 1.37.6.

### Fixed

- **Backup `Size` Dropped on List** — `BackupClient.List()` discarded the size and incremental base id it had already parsed from the response, so `Backup.Size` was null for every listed backup. Both fields are now mapped.
- **Aggregate Zeros for Absent Values** — Int, Number, Boolean, Text and Date results, `Count` included, reported `0`, `0.0` or `false` where the server had sent no value at all. Every scalar is now presence-checked and left null when unset, so an empty aggregation is distinguishable from a genuine zero.
- **`Multi2VecGoogleGemini` Emitted a Non-Existent Module** — The vectorizer declared the module name `multi2vec-google-gemini`, which no Weaviate server provides, so a collection configured with it could never be created. `VectorizerFactory.Multi2VecGoogleGemini(...)` now emits `multi2vec-google` with the Gemini API endpoint.
- **Multimodal Vectorizer Weights Never Reached the Wire** ([#359](https://github.com/weaviate/weaviate-csharp-client/pull/359)) — Weighted `multi2vec` factories assigned `VectorizerWeights`, but it never serialized, so per-modality weights were silently dropped and collections were created with uniform weighting. Weights now reach the wire, and are omitted when no modality carries any.
- **Google Multimodal Factories Transposed Video and Audio** ([#359](https://github.com/weaviate/weaviate-csharp-client/pull/359)) — `Multi2VecGoogle` and `Multi2VecGoogleGemini` passed their video and audio field lists into the `audioFields` and `depthFields` parameters, so video weights were labelled as audio and audio weights landed under a modality the module does not have, which can fail collection creation.
- **Empty Multimodal Modalities Sent as Empty Arrays** ([#359](https://github.com/weaviate/weaviate-csharp-client/pull/359)) — An empty `WeightedFields` or `string[]` serialized as e.g. `"textFields": []`, which the server rejects, so an image-only config failed collection creation and `Multi2VecBind` was unusable with a subset of its seven modalities. Empty is now sent as null.
- **`generative-google` Sent an Empty `location`** ([#359](https://github.com/weaviate/weaviate-csharp-client/pull/359)) — An empty `location` was sent as present-but-empty, suppressing the server-side default. Vertex now sets it only when non-empty; Gemini leaves it unset.
- **`multi2vec-twelvelabs` Sent `vectorizeCollectionName`** ([#367](https://github.com/weaviate/weaviate-csharp-client/pull/367)) — No `multi2vec` module reads the setting, but the client sent it, so it read back as though it had taken effect. Removed from `Multi2VecTwelveLabs` and its factories; the other `multi2vec` records keep theirs for now.
- **`AndCross` Version Guard Threw a Server-Side Exception Type** ([#365](https://github.com/weaviate/weaviate-csharp-client/pull/365)) — The client-side guard threw `WeaviateFeatureNotSupportedException`, a `WeaviateServerException`. It now throws `WeaviateVersionMismatchException`, matching every other client-side version gate; update catch blocks written for the old type.
- **Backup `Size` Dropped on List** ([#368](https://github.com/weaviate/weaviate-csharp-client/pull/368)) — `BackupClient.List()` discarded the size and incremental base id it had already parsed from the response, so `Backup.Size` was null for every listed backup. Both fields are now mapped.
- **Aggregate Zeros for Absent Values** ([#368](https://github.com/weaviate/weaviate-csharp-client/pull/368)) — Aggregation scalars, `Count` included, reported `0`, `0.0` or `false` where the server had sent no value. They are now null when unset, so absence is distinguishable from a genuine zero.
- **`Multi2VecGoogleGemini` Emitted a Non-Existent Module** ([#368](https://github.com/weaviate/weaviate-csharp-client/pull/368)) — The vectorizer declared the module name `multi2vec-google-gemini`, which no server provides, so such a collection could never be created. It now emits `multi2vec-google` with the Gemini API endpoint.

### Changed

- **`Multi2VecGoogleGemini` is now an `[Obsolete]` shim over `Multi2VecGoogle`** — This breaks at runtime, not only at compile time: the factory now returns a `Multi2VecGoogle`, so `is`/`as`/pattern matches on `Multi2VecGoogleGemini` silently stop matching, and a `switch` with arms for both types no longer compiles (`CS8120: The switch case is unreachable`). Bind the result as `Multi2VecGoogle` or `VectorizerConfig`.
- **`Multi2VecGoogle.ProjectId` and `.Location` are no longer `required`** — Both getters changed from `string!` to `string?` so the Gemini endpoint, which is scoped to neither a project nor a region, can omit them. Callers with nullable reference types enabled may see CS8600/CS8601.
- **`Aggregate.Boolean` counts are nullable** — `PercentageTrue`, `PercentageFalse`, `TotalTrue` and `TotalFalse` changed from `double`/`long` to `double?`/`long?`.
- **`BackupCreateRequest` gained a trailing optional parameter** — Its primary constructor and `Deconstruct` went from six parameters to seven, so six-element positional deconstruction (`var (id, backend, inc, exc, cpu, comp) = req;`) no longer compiles; deconstruct seven elements instead.
- **Binary compatibility** — The signature changes above are source-compatible for callers using named arguments, but not binary-compatible: assemblies compiled against 1.1.0 throw `MissingMethodException` until recompiled.
- **`Multi2VecGoogleGemini` is now an `[Obsolete]` shim over `Multi2VecGoogle`** ([#368](https://github.com/weaviate/weaviate-csharp-client/pull/368)) — The factory now returns a `Multi2VecGoogle`, so `is`/`as`/pattern matches on `Multi2VecGoogleGemini` silently stop matching and a `switch` with arms for both types no longer compiles (`CS8120`). Bind the result as `Multi2VecGoogle` or `VectorizerConfig`.
- **`Multi2VecGoogle.ProjectId` and `.Location` are no longer `required`** ([#368](https://github.com/weaviate/weaviate-csharp-client/pull/368)) — Both are now `string?`, so a Gemini config can omit them. Callers with nullable reference types enabled may see CS8600/CS8601.
- **`Aggregate.Boolean` counts are nullable** ([#368](https://github.com/weaviate/weaviate-csharp-client/pull/368)) — `PercentageTrue`, `PercentageFalse`, `TotalTrue` and `TotalFalse` changed from `double`/`long` to `double?`/`long?`.
- **`BackupCreateRequest` gained a trailing optional parameter** ([#368](https://github.com/weaviate/weaviate-csharp-client/pull/368)) — Its primary constructor and `Deconstruct` went from six parameters to seven, so six-element positional deconstruction no longer compiles — deconstruct seven.
- **Binary compatibility** — Source-compatible for callers using named arguments, but not binary-compatible: 125 public members gained a trailing optional parameter or a nullability change, so assemblies compiled against 1.1.0 or 1.1.1 throw `MissingMethodException` until recompiled against 1.2.0. **Migration: recompile.** No source change is needed unless you used positional arguments past the insertion point, deconstructed `BackupCreateRequest` positionally, or pattern-matched on `Multi2VecGoogleGemini` (see above).
- **`Weaviate.Client.VectorData` republished at 1.2.0** — Ships under the same tag and version as the core client but has **no API or behavior changes**: its public surface is byte-identical to 1.1.0 and 1.1.1. The 1.2.0 build references the 1.2.0 core client, so upgrade the two together.

### Removed

- **`Multi2VecGoogleGemini`'s own members** ([#368](https://github.com/weaviate/weaviate-csharp-client/pull/368)) — Now that the type is a shim over `Multi2VecGoogle`, its seven own properties are gone, replaced by the base type's equivalents. Migrate `.Model` → `.ModelId`; the rest keep their names.

### Minimum Supported Weaviate Version

| Feature | Minimum Weaviate Version |
|------------------------------------------------------------------------------------------|------------------------------------------|
| Core client | 1.32.0 |
| Diversity selection on `near-vector` / `near-object` / `near-text` / `near-media` | 1.37.0 |
| Incremental backup create (`IncrementalBaseBackupId`) | 1.37.0 (read-back from 1.37.6) |
| `generative-deepseek` | 1.36.19 |
| Boost (all search and generate methods) | 1.38.0 (silently ignored below) |
| `Namespaces` RBAC permission scope | 1.38.0 |
| Diversity selection on hybrid search | 1.38.6 |
| `BM25Operator.AndCross` | 1.39.0, backported to 1.37.15 and 1.38.8 |
| `multi2vec-twelvelabs` | 1.38.9 or 1.39.0 |

---

## [1.1.1] — 2026-05-27

### Added

- **`text2vec-digitalocean` Vectorizer** ([#339](https://github.com/weaviate/weaviate-csharp-client/issues/339)) — New `Models.Vectorizer.Text2VecDigitalOcean` config record and `VectorizerFactory.Text2VecDigitalOcean(...)`. `model` (e.g. `qwen3-embedding-0.6b`) is required and comes first; `baseURL` is optional.

### Fixed

- **Vector Index Type Defaulting** ([#341](https://github.com/weaviate/weaviate-csharp-client/pull/341)) — The client wrote `"hnsw"` into every empty `vectorIndexType`, defeating `DEFAULT_VECTOR_INDEX`, added in Weaviate 1.37.5. The field is now left unset on servers ≥ 1.37.5, and still set to `"hnsw"` on servers below 1.37.5 or of undetermined version.

### Removed

- **`ReplicationAsyncConfig.MaxWorkers` and `ReplicationAsyncConfig.AliveNodesCheckingFrequency`** — Both fields have been no-ops on the server since Weaviate 1.37.3 and are now removed from the user-facing model, the OpenAPI spec, and the generated DTO. Existing code that sets these properties will not compile after upgrading; no behavioral change results from the removal.
- **`Multi2VecGoogleGemini.Model`** — Removed along with the rest of the type's own members; the base `Multi2VecGoogle` declares the same setting as `ModelId`. Migrate `.Model` → `.ModelId`.
- **`ReplicationAsyncConfig.MaxWorkers` and `ReplicationAsyncConfig.AliveNodesCheckingFrequency`** — Both have been no-ops on the server since Weaviate 1.37.3 and are now removed. Code that sets them will not compile after upgrading; behavior is unchanged.

---

Expand Down Expand Up @@ -217,7 +292,9 @@ Initial stable release of the Weaviate C# client.

---

[Unreleased]: https://github.com/weaviate/weaviate-csharp-client/compare/1.1.0...HEAD
[Unreleased]: https://github.com/weaviate/weaviate-csharp-client/compare/1.2.0...HEAD
[1.2.0]: https://github.com/weaviate/weaviate-csharp-client/compare/1.1.1...1.2.0
[1.1.1]: https://github.com/weaviate/weaviate-csharp-client/compare/1.1.0...1.1.1
[1.1.0]: https://github.com/weaviate/weaviate-csharp-client/compare/1.0.1...1.1.0
[1.0.1]: https://github.com/weaviate/weaviate-csharp-client/compare/1.0.0...1.0.1
[1.0.0]: https://github.com/weaviate/weaviate-csharp-client/releases/tag/1.0.0
Loading
Loading