Michael J. Sullivan
Michael J. Sullivan
The only way to produce json objects is by casting from objects and named tuples and converting a string from `to_json`. I think we want something like `function json_object_pack(els: array)...
``` select [(1,'foo'), (2,'bar'), (3,'baz')][1:]; ``` gives ``` edgedb error: InternalServerError: return type record[] is not supported for SQL functions ``` We will need to essentially inline the array slicing...
It would be nice to list out what was causing one side of the operation to be volatile and maybe even what causes the other side to be multi. This...
We strive for backwards compatibility, but a huge swath of clear bugfixes can cause migrations scripts to fail to apply, since they can contain basically arbitrary expressions. (Additionally, we can...
Another instance of a classic edgedb bug. Slicing is a regular (non-optional) operation, and so empty set inputs should result in an empty set output, like such: ``` edgedb> select...
``` edgedb> select ([])[0:]; {} ``` Correct behavior would be to produce `{[]}`. Strings and bytes work correct. json strings work but json arrays are broken.
With a schema like ``` type Bar { property x -> str; } type Baz { property x -> str; } ``` If we do `WITH W := (Bar UNION...
We struggle with migrations of very large schemas (see #3946) and the biggest component of the time spent is pickling the large schemas to send back from the compiler process....
See https://www.edgedb.com/docs/clients/python/api/blocking_client#edgedb.create_client But the client raises an error if you try to do it
The mypy plugin interface is experimental, unstable, and prone to change. In particular, there are no guarantees about backwards compatibility. Backwards incompatible changes may be made without a deprecation period....