2022-03-19 06:11:35 +08:00
|
|
|
This is the draft release notes for 0.19.0, scheduled to be released around
|
2022-05-20 11:01:43 +08:00
|
|
|
2022-07-01.
|
2022-03-19 06:11:35 +08:00
|
|
|
|
|
|
|
# Breaking changes
|
|
|
|
|
|
|
|
# Deprecated features
|
|
|
|
|
2022-04-09 18:45:15 +08:00
|
|
|
Deprecated features will be removed in 0.20.0.
|
|
|
|
|
|
|
|
The following deprecated features trigger a warning whenever the code is parsed
|
|
|
|
and compiled, even if it is not executed:
|
|
|
|
|
|
|
|
- The `float64` command is now deprecated. Use `num` for constructing a typed
|
|
|
|
number, or `inexact-num` for constructing an inexact number.
|
|
|
|
|
|
|
|
The documentation has advertised it as deprecated since the 0.16.0 release,
|
|
|
|
but deprecation warnings were never added.
|
|
|
|
|
2022-03-19 06:11:35 +08:00
|
|
|
# Notable bugfixes
|
|
|
|
|
2022-06-06 05:39:48 +08:00
|
|
|
- Temporary assignment on an unset environment variables no longer leave it
|
|
|
|
set to an empty string ([#1448](https://b.elv.sh/1448)).
|
|
|
|
|
2022-03-19 06:11:35 +08:00
|
|
|
# Notable new features
|
2022-04-09 18:45:15 +08:00
|
|
|
|
|
|
|
- A new `inexact-num` converts its argument to an inexact number.
|
|
|
|
|
|
|
|
It is functionally identical to the now deprecated `float64` command since
|
|
|
|
the Go float64 type is the only underlying inexact number type for now. Its
|
|
|
|
behavior may change in future if there are more underlying types for inexact
|
|
|
|
numbers.
|
2022-05-20 11:01:43 +08:00
|
|
|
|
|
|
|
- A new type of interactive abbreviation: `edit:command-abbr`
|
|
|
|
([#1472](https://b.elv.sh/1472)).
|