Release v1.7.2

New projects scaffolded with a SQL database or with cratis-aspire now run end to end, and every template ships its .gitignore. Projects you have already created are unchanged; the fixes below list what to copy if you want them.

Fixed

• cratis --Database PostgreSQL, MsSql, or SQLite: the sample listing now shows registered items. Previously the query failed because the client still sent read models to a MongoDB sink the kernel did not have, and the app connected to a database and user that did not exist. The generated appsettings.json now sets Cratis:Chronicle:DefaultSinkTypeId to SQL and points ConnectionStrings:Cratis at chronicle+<event store>, the database Chronicle projects into
• cratis --Database SQLite now keeps Chronicle's data in a chronicle-data folder in the project, so the backend can read the read models; it is listed in .gitignore
• cratis-aspire with the default MongoDB database no longer stops at startup, and the backend now connects to the Chronicle container Aspire starts instead of localhost:35000
• cratis-aspire with PostgreSQL, MsSql, or SQLite now starts Chronicle with the development client credentials the backend uses, waits for the database before starting Chronicle, and gives the backend its read-model database
• cratis-aspire SQL variants no longer add Cratis.Arc.MongoDB, and pin Cratis.Arc.EntityFrameworkCore to the current release like the other Cratis packages
• The sample Listing read model now records the event source id it is keyed by, which the listing page uses as its row key
• Every template now produces a .gitignore; NuGet had been dropping it from the packages
• yarn build no longer fails with Yarn 2 and later: the frontend templates include a .yarnrc.yml that sets nodeLinker: node-modules