Sqlmesh

:beetle: Bug fixes
* Databricks properly handle missing is_serverless env var (3101) eakmanrq
* Allow overriding blocking setting of builtin audits (3106) georgesittas
* Make sure signals are loaded when running models (3107) izeigerman
* Allow macros in set operations (3103) VaggelisD
* Support dbt sqlserver 1.8.0 (3112) eakmanrq

:broom: Chore
* Use external github image links in README (3111) treysp

:sparkles: Features
* Support plan builder class overriding in context (3078) georgesittas
* Dry run flag alias (3091) sungchun12
* Improve cold start of snapshot cache with multi processing (3084) tobymao
* Support BigQuery nested column additions without dropping (3082) Themiscodes
* Format check (3073) etonlels

:beetle: Bug fixes
* Set columns attribute for python models that generate sql (3087) ravenac95
* Macro function comments closes 3088 (3093) tobymao
* Disable airflow multiprocessing (3097) tobymao

:broom: Chore
* Add retries on engine adapter integration tests (3090) erindru
* Bump sqlglot to v25.20.1 (3100) georgesittas

:sparkles: Features
* Use multiprocessing to speed up loading (3077) tobymao

:beetle: Bug fixes
* Always provide feedback during plan execution (3071) plaflamme
* Support for unique key expressions for SCD2 by column models (3075) izeigerman
* Prepopulate the model's depends on in-memory cache for converted dbt models (3074) izeigerman
* Properly nested update gateway configs (3085) eakmanrq

:broom: Chore
* Isolate web tests (3076) vchan
* Update SQLMesh and Tobiko logos to the new style in header and favicon (3079) caiters
* Bump sqlglot to v25.19.0 (3083) georgesittas

:sparkles: Features
* Improve table diff sample output in CLI (3017) Themiscodes
* Improve environment loading performance by lazy evaluating snapshot and snapshot id fields (3038) crericha
* Allow creating environments without changes when using selector (3060) izeigerman
* Add clickhouse engine adapter (3019) treysp

:beetle: Bug fixes
* Skip loading context on run closes 1511 (3037) tobymao
* Rename renderer cache so it's not serialized (3039) tobymao
* Don't cache snapshot intervals (3041) izeigerman
* User-provided plan end must take precedence over the max interval end per model (3042) izeigerman
* Add path to Invalid model definition error, fixes 3040 (3045) akv-akv
* Heuristics that determine the default start date of the forward-only preview plan (3043) izeigerman
* Allow omitting the MODEL (...) block when name inference is activated (3049) georgesittas
* Test failure caused by a new version of rich (3053) izeigerman
* Don't process models that have already been processed (3052) tobymao
* Seed and model column mismatch due to normalization (3036) Themiscodes
* Handle missing source target aliases in sample data diff (3046) Themiscodes
* **(bigquery)** Do not use execution_project as default catalog in partitions query (3056) treysp
* Handle metaprogramming inspect.getfile TypeError (3058) georgesittas
* Cache inclusive exclusive for performance (3057) tobymao
* Ignore expired environments in the selector (3059) izeigerman
* Retain indexes when creating state table backups (3061) izeigerman

:broom: Chore
* Document batch_concurrency for unique_by_key closes 3031 (3034) tobymao
* Bump sqlglot to v25.17.0 (3048) georgesittas
* Treat the target environment as missing during planning if it's expired (3051) izeigerman

:alien: Other
* Document unavailable schema placeholder for sqlmesh macros (3035) georgesittas
* Beautify README (3068) sungchun12
* Move README images to docs/readme/ (3069) treysp

:sparkles: Features
* Use multiprocessing to speed up loading (3077) tobymao
* Improve cold start of snapshot cache with multi processing (3084) tobymao

:beetle: Bug fixes
* Disable airflow multiprocessing (3097) tobymao

:beetle: Bug fixes
* Support for unique key expressions for SCD2 by column models (3075) izeigerman
* Prepopulate the model's depends on in-memory cache for converted dbt models (3074) izeigerman