An inline message lets users know when important information is available or when an action is required.
yarn add @atlaskit/inline-message#117363
10a0f7f6c2027 -
This package's peerDependencies have been adjusted for react and/or react-dom to reflect the
status of only supporting React 18 going forward. No explicit breaking change to React support has
been made in this release, but this is to signify going forward, breaking changes for React 16 or
React 17 may come via non-major semver releases.
Please refer this community post for more details: https://community.developer.atlassian.com/t/rfc-78-dropping-support-for-react-16-and-rendering-in-a-react-18-concurrent-root-in-jira-and-confluence/87026
b50c5d5d65ae2 -
Bump to the latest version of @compiled/react4660ec858a305 -
Update React from v16 to v184facbe306465b -
Migrated from @emotion/react to @compiled/react in order to improve performance, align with
the rest of the Atlaskit techstack, and support React 18 Streaming SSR. Please note, in order to
use this version of @atlaskit/inline-message, you will need to ensure that your bundler is
configured to handle .css imports correctly. Most bundlers come with built-in support for .css
imports, so you may not need to do anything. If you are using a different bundler, please refer to
the documentation for that bundler to understand how to handle .css imports. For more
information on the migration, please refer to
RFC-73 Migrating our components to Compiled CSS-in-JS.be58e4bb2e387 -
Migrating usages of UNSAFE types and entrypoints that have been renamed in @atlaskit/icon and
@atlaskit/icon-lab.962b5e77810fb -
Adds side-effect config to support Compiled css extraction in third-party apps3d03c4f1002ab -
Integrate layering and use CloseManager instead770bc26d556f7 -
Migrate to new icons behind a feature flag#127511
db30e29344013 -
Widening range of react and react-dom peer dependencies from ^16.8.0 || ^17.0.0 || ~18.2.0
to the wider range of ``^16.8.0 || ^17.0.0 || ^18.0.0` (where applicable).
This change has been done to enable usage of react@18.3 as well as to have a consistent peer
dependency range for react and react-dom for /platform packages.
ff0815316ab38 -
Removes usage of custom theme button in places where its API is not being used and the default
button is able to be used instead. This should give a slight performance (runtime) improvement.8fb8ca26fb173 -
Integrate layering in inline dialogc733254a2dd6e -
Explicitly set jsxRuntime to classic via pragma comments in order to avoid issues where jsxRuntime
is implicitly set to automatic.45033e519694 -
Add layering support to fix escape keyboard issue in modals3ae5d8dba986 -
Internal change only; update to use primitive components and tokenised typography.ec7c2a38247 - Removed
all remaining legacy theming logic from the Calendar, Form, InlineDialog, InlineEdit and
InlineMessage components.07aa588c8a4 - Reverts
the fix to text descender cut-off, due to incompatibilities with Firefox and Safari.89fccc9c758 - Add
aria-expanded to communicate expanded/ collapsed state56b444b56a8 - Fix a
bug where text descenders were cut off at high zoom levels on Windows4ae083a7e66 - Use
@af/accessibility-testing for default jest-axe config and jest-axe import in accessibility
testing.56507598609 - Skip
minor dependency bump18aeca8c199 - Internal
change to update token references. There is no expected behaviour or visual change.b0f6dd0bc35 - Updated
to use typography tokens. There is no expected behaviour or visual change.25902de2d93 - [ux]
Added pressed (active) state styles for the secondary text.9de88fa1e1e - Internal
changes to include spacing tokens in component implementations.bcf29b33526 - Updates
@emotion/core to @emotion/react; v10 to v11. There is no expected behavior change.#16759
3d6d3a581d6 -
Instrumented @atlaskit/inline-message with the new theming package, @atlaskit/tokens.
New tokens will be visible only in applications configured to use the new Tokens API (currently in alpha). These changes are intended to be interoperable with the legacy theme implementation. Legacy dark mode users should expect no visual or breaking changes.
af4bca32ad4 - Internal
changes to supress eslint rules.1489097139d - Added
iconLabel prop to provide useful information to users with screen readers65388bb777d - Removes
the styled-components dependency from the package. Still has a transitive reliance on
styled-components via inline-dialog. There should be no visual or UX change.9c98e8227f6 - Internal
refactor for style declarations.5f58283e1f - Export
types using Typescript's new "export type" syntax to satisfy Typescript's --isolatedModules
compiler option. This requires version 3.8 of Typescript, read more about how we handle Typescript
versions here: https://atlaskit.atlassian.com/get-started Also add typescript to
devDependencies to denote version that the package was built with.#3885
6c525a8229 - Upgraded
to TypeScript 3.9.6 and tslib to 2.0.0
Since tslib is a dependency for all our packages we recommend that products also follow this tslib upgrade to prevent duplicates of tslib being bundled.
954cc87b62 - The readme
and package information has been updated to point to the new design system website.94ba4f2381 - Formatted
testId prop description. The prop description was formatted incorrectly.87f4720f27 - Officially
dropping IE11 support, from this version onwards there are no warranties of the package working in
IE11. For more information see:
https://community.developer.atlassian.com/t/atlaskit-to-drop-support-for-internet-explorer-11-from-1st-july-2020/3953454a9514fcf - Build and
supporting files will no longer be published to npm[patch]68e206c857:
Change imports to comply with Atlassian conventions- Updated dependencies 6b8e60827e:
Updated dependencies 449ef134b3:
Updated dependencies 3a09573b4e:
Updated dependencies 57c0487a02:
Updated dependencies fd41d77c29:
[patch]6548261c9a:
Remove namespace imports from React, ReactDom, and PropTypes- Updated dependencies 6548261c9a:
[patch]4a223473c5:
Removes babel/runtime from dependencies. Users should see a smaller bundlesize as a result- Updated dependencies 28f8f0e089:
Updated dependencies 82747f2922:
Updated dependencies 4a223473c5:
[minor]3c86f3180f:
Adding an optional prop testId that will set the attribute value data-testid. It will help
products to write better integration and end to end tests.
[patch]35d2229b2a:
Adding missing license to packages and update to Copyright 2019 Atlassian Pty Ltd.
[patch]097b696613:
Components now depend on TS 3.6 internally, in order to fix an issue with TS resolving non-relative imports as relative imports
[patch]708028db86:
Change all the imports to theme in Core to use multi entry points
[patch]3b13bd0816:
Both title and secondaryText prop type changed from string to React.ReactNode. This provides more flexibility for consumers to provide i18n components such as FormattedMessage.
[patch]b0ef06c685:
[major]66af32c013:
[major]7c17b35107:
[major]76299208e6:
As a breaking change, all @atlaskit packages will be dropping cjs distributions and will only
distribute esm. This means all distributed code will be transpiled, but will still contain
import and export declarations.
The major reason for doing this is to allow us to support multiple entry points in packages, e.g:
import colors from `@atlaskit/theme/colors`;
Previously this was sort of possible for consumers by doing something like:
import colors from `@atlaskit/theme/dist/esm/colors`;
This has a couple of issues. 1, it treats the file system as API making internal refactors harder, we have to worry about how consumers might be using things that aren't actually supposed to be used. 2. We are unable to do this internally in @atlaskit packages. This leads to lots of packages bundling all of theme, just to use a single color, especially in situations where tree shaking fails.
To support being able to use multiple entrypoints internally, we unfortunately cannot have multiple distributions as they would need to have very different imports from of their own internal dependencies.
ES Modules are widely supported by all modern bundlers and can be worked around in node environments.
We may choose to revisit this solution in the future if we find any unintended condequences, but we see this as a pretty sane path forward which should lead to some major bundle size decreases, saner API's and simpler package architecture.
Please reach out to #fabric-build (if in Atlassian) or create an issue in Design System Support (for external) if you have any questions or queries about this.