Michael J. Sullivan
Michael J. Sullivan
When EDGEDB_TEST_REPEATS is set, the test suite duplicates transactional query tests, and the re-run version will typically use the function cache. There are some actual failures when I ran this...
There are a few different things we can do here, in difficulty order: 1. Have the CLI indicate when parse is finish and execute has started 2. Submit migration DDL...
In our test database, if you do `ext::ai::to_context(Astronomy)`, it will always grab `.content`, even though for the subtype `Stuff` it should do `.content ++ .content2`
Currently, while we allow in-place upgrades between minor versions, upgrading between major versions requires a dump and restore. This is slow, and requires downtime. Currently we use the patch system...
We have an LSP "language server now", but should support more features. - [x] read gel.toml and `project.schema-dir` instead of using `dbschema/` - [x] run schema delta on .esdl files...
Currently the value of negative constants are omitted. Only omit *0* constants. An earlier attempt (#244) caused a mismatch with the protobuf json output, since that wants to omit values...
I was trying to log in on a remote machine, and had to source dive and then use `RUST_LOG=debug` to make it work.
It might get dropped after the databases have been enumerated
Draft an RFC proposing the elimination of path factoring. Any feedback about what to expand?