What's Changed

• Stores any POCO/POJO with no schema definition required
• JsonSchema instead of JsonDocumentType
• Got rid of SchemaName method signatures
• Significantly cleaned up code/unit tests etc
• Added JsonQueryable
• Perf optimizations

Full Changelog: v3.0.0...v4.0

What's Changed

• Support for JsonSchema/JsonDocument untyped objects suitable for WebAPI wrapper around LottaDB
• Added Blob operations and full text indexing

Full Changelog: v1.1.0...v3.0.0

What's Changed

• Bulk operations, SaveManyAsync(), DeleteManyAsync(), etc.
• Smart index flushing, so lots of writes will end up commiting automatically once activity settles down or a search is performed.
• Vector search - You can now search .Search(a => a.Similar("some cool dude")) and get results based on vector Knn search.

LottaDB

LottaDB is a .NET library that makes it easy to story any POCO in Azure Table Storage with full Lucene search, all with the goodness of LINQ.

• One line to save
• One line to search

Overview

LottaDB gives you a document database built on Azure Table Storage with automatic full-text search via Lucene Search Engine. Each LottaDB instance is a single table wih backing lucene catalog. Objects are stored with full POCO fidelity, while selected properties are promoted into Table Storage/Lucene for efficient querying using full typed Linq expressions.

Reactive handlers (On<T>) let you build materialized views, cascading updates, and side effects that run inline after each write.

Why LottaDB?

• A lotta bang for a little buck. Table Storage is the cheapest durable storage in Azure. LottaDB adds Lucene so you get rich queries without the rich pricing.
• A lotta LINQ. Query<T>() and Search<T>(), .Where(), .OrderBy() etc.
• A lotta fidelity. Full JSON roundtrip. Lists, dictionaries, nested objects -- everything survives.
• A lotta views. On<T> triggers build materialized views with plain C#. No event buses, no eventual consistency -- just inline code.
• A lotta tenants. One instance per tenant. Natural isolation, simple backup, no noisy neighbors.
• A lotta nothing to operate. Table Storage is serverless. Lucene runs in-process. No clusters, no connection pools, no ops team required.

Sweet spot

LottaDB is ideal for per-user or per-tenant workloads -- think user profiles, settings, activity feeds, personal knowledge bases, mailboxes, or per-project data. Thousands of objects per tenant, thousands of tenants per deployment. Each tenant gets its own isolated database for pennies/month. It's not designed for billion-row analytics or high-throughput write-heavy pipelines.