Fixes
- Fixing touch dragging for firefox browsers #1296. A recent firefox bug causes
touchmove
events incapture
event listeners to be broken. I have created a firefox bug report. As a workaround we are using noncapture
listeners fortouchmove
which is fine for our use case. Thanks @kamilmoskal for reporting the problem - Fixing issue where dragging in
jsdom
could cause an error to be thrown ifgetComputedStyle
was not mocked correctly #1297. See alexreardon/css-box-model#9 - Fixing
flow
error #1294. Thanks heaps @TrySound!
Engineering health
- Bumping internal dependencies #1297
Encouraging correct usage of key
s
When rendering items in a list React
recommends using a key
to keep track of the item between re-rendering (docs, guide). In 10.x
we covered over some issues when incorrectly using key
s. 11.x
did not mask incorrect key
usage. In this release we have:
- π©ββ Restored the old (still somewhat broken) behaviour when incorrectly using
key
s - π·ββ Logging a development only warning when we detect incorrect
key
usage - π©βπ« Created a guide to help you using
key
withreact-beautiful-dnd
Other
- Internal linting for
useLayoutEffect
(#1281)
Fixes
- Drag handle will no longer steal focus from children that auto focus #1271. Thanks @mDibyo for raising this
Other
- Improved documentation wording #1270 by @timhaywood
- Adding more function component examples in the docs #1266
- Internal: correctly leveraging
eslint-plugin-react-hooks
withuseMemoOne
#1264
The one with hooks π£ (almost π)
This release is mostly internal changes and our public API remains relatively untouched
The original plan was to move to a totally hooks based api #871. However, we hit a snag: right now it is not possible for a hook
to set up context
. A Droppable
component sets up context
for a Draggable
to consume. So for now we are sticking with our render prop API. Internally we have moved almost everything over to function components and hooks. (Our error boundary still needs to be a class
because function components cannot currently be an error boundary)
Our internal refactor has also resulted in react-beautiful-dnd
now being <StrictMode>
compliant
Sharing the learnings π
I have given a talk at React Sydney about my journey converting react-beautiful-dnd
from class
components over to function
components and hooks.
π¬ Deep sea fishing with React hooks
Breaking changes π₯ #1194
1. React
peer dependency ^16.3.1
β ^16.8.5
In case you wanted to know why we picked
^16.8.5
π΅οΈββοΈ
- Internally we have moved most things to function components and hooks which need
^16.8.0
#871 - We were still using old school
React
context
. We could have moved to the new context api in our previous^16.3.1
dependency, but consumingcontext
throughuseContext
in16.8
has made it a lot easier to shift over. - Need to move to
^16.8.4
to usereact-redux@7
. We proposed to get this lowed to^16.8.0
but the project was not keen. We could still claim ourReact
peer dependency is^16.8.0
, but that could lead to issues with bundlers if using a version>= 16.8.0
and< 16.8.4
. It would at least be a warning, and might even be worse. - Rather than moving to a peer dependency of
^16.8.4
(forreact-redux@7
) we went to^16.8.5
as it contained fixes for shallow rendering which could impact some peoplesreact-beautiful-dnd
tests
2. <Draggable />
β shouldRespectForceTouch
2.1 Name change
shouldRespectForceTouch => shouldRespectForcePress
- shouldRespectForceTouch
+ shouldRespectForcePress
Moving to shouldRespectForcePress
is a clearer name given that force press actions can be fired from mouse inputs too.
We will be calling event.preventDefault()
on the following events from a drag handle if shouldRespectForcePress
is set to true
(the new default)
webkitmouseforcechanged
(mouse)touchforcechange
(touch)
For the next little while we log a development warning if you are using the old
shouldRespectForceTouch
prop
2.2 Flipping the default
See #1160. Opting for a better dragging experience out of the box
- shouldRespectForcePress: true,
+ shouldRespectForcePress: false,
New package: useMemoOne
One byproduct of our move to use hooks was the creation of a new package!
useMemoOne
:useMemo
anduseCallback
with a stable cache (semantic guarantee)
useMemo
and useCallback
cache the most recent result. However, this cache can be destroyed by React
when it wants to:
You may rely on useMemo as a performance optimization, not as a semantic guarantee. In the future, React may choose to βforgetβ some previously memoized values and recalculate them on next render, e.g. to free memory for offscreen components. Write your code so that it still works without useMemo β and then add it to optimize performance. - React docs
useMemoOne
and useCallbackOne
are concurrent mode
safe alternatives to useMemo
and useCallback
that do provide semantic guarantee. What this means is that you will always get the same reference for a memoized value so long as there is no input change.
Other
- Moving to the new GitHub issue templates to streamline issue creation
- Fixed broken external link #1235. Thanks @erik-beus!
- Upgrading to
flow
0.96
π€ - Fixed issue caused by missing
SVGElement
in some testing environment #1249. Thanks @hawkeng for raising this
Open discussions π£
We would love your input on our future direction!
Thanks β€οΈ
- @markerikson for his work on
react-redux@7
which made this release possible - @pwolaq for finding a focus management issue with
11.0
- Everyone who tested the
11.0
beta
- Everyone involved in our hooks discussion #871 (@YurkaninRyan, @gnapse, @trixn86, @shawninder)
- React Sydney for hearing me out
Fixes π©ββοΈ
DraggableStateSnapshot
andDroppableStateSnapshot
are now correctly populated when dropping outside of aDroppable
. In10.1
they where being populated with the home location rather than no location. #1178. Thanks @zmi-jlblatt for finding this one- Respecting
shouldRespectForceTouch
preference for force press touch events that are triggered by some mouse inputs #1192. More information: #1193
Documentation
- Fixing broken link in examples page. Thanks @eliotball
- Updating typo in multi-drag pattern #1171. Thanks @Aarbel
- Cleaning up the language in our
ISSUE_TEMPLATE.md
#1202. Thanks @lukyth!
Engineering health π·ββοΈ
- Refactoring in preparation for hooks π£#871 #1188
- Bumped all
devDependencies
#1195 - Moved to
flow
0.95.1
(Minor) breaking changes are coming soon π₯
We will soon be publishing 11.0
. This will have two minor breaking changes:
- Moving
React
peer dependency from^16.3.1
to^16.8.0
shouldRespectForceTouch
name change and default swap
More details: #1194
π§πΆ Release notes designed to be listed to with: "Once there were dragons" @ 4:05
Highlights π€ΎββοΈ
- π©βπ¨ Stunning animation improvements
- π Overhauled documentation
- π± More predictable touch dragging
- π Simplified home list styling
- π New project design
- π Blogs of our learnings
Beautiful π
This release pushes react-beautiful-dnd
to new summits of interaction design.
We have removed instant movement (snapping) when moving between lists. Snapping breaks the visual language of moving physical objects around that react-beautiful-dnd
is trying to emulate.
β No snapping when moving between lists
β This is what it used to be (snapping in the π home list after a drop)
You do not need to do anything to get the new animation pattern. I have marked this change as a feature release as it is new design behaviour.
Blog: Beautiful interactions π
There was a huge amount of work to remove snapping when moving between lists. Who would have thought? Not me π₯΅. I wrote a blog which goes over the journey of this feature in depth. Personally, I think it well worth your time!
πBlog β Beautiful interactions: Crafting elegant and robust drag and drop animations
New logo π©βπ¨
We have moved to a logo that encapsulates the playfulness and delight that we are striving for with react-beautiful-dnd
Blog: Logo creation π
Maryanne Nguyen, a very talented designer at Atlassian, has detailed her creative process in creating the new react-beautiful-dnd
logo in a blog π¨
πBlog β maryannemade: react-beautiful-dnd
logo
More predictable touch dragging π±
react-beautiful-dnd
tries to respect standard browser interactions as much as possible. Currently, the library plays well with force touch interactions:
If the user force presses on the element before they have moved the element (even if a drag has already started) then the drag is cancelled and the standard force press action occurs. For an anchor, this is a website preview.
Unfortunately, heavy presses can cancel a drag. It can be tricky at times to get the right amount of pressure to correctly use touch dragging. This can be frustrating for users.
We have added an option for you to opt-out of respecting force touch interactions. This option will disable force press interactions on the drag handle. However, it allows for a more consistent touch dragging interaction.
We have added shouldRespectForceTouch
as a prop to <Draggable />
. It defaults to true
in order to respect force touch interactions, but you can set it to false
to opt out of respecting force touch
<Draggable
draggableId="draggable"
+ shouldRespectForceTouch={false}
>
</Draggable>
This change resulted in a feature release. The default behaviour matches the behaviour of previous versions.
Should force touch be respected by default?
Right now shouldRespectForceTouch
defaults to true
in order to respect browser interactions. However, for this specific use case, I am thinking that the default could be to not respect force touch in order to have a more consistent touch dragging experience out of the box.
I have created a request for discussion to collect other people's thoughts on what the default should be.
Simplified home list styling π
We have added a new value to the DroppableStateSnapshot
: draggingFromThisWith
type DroppableStateSnapshot = {|
isDraggingOver: boolean,
draggingOverWith: ?DraggableId,
+ draggingFromThisWith: ?DraggableId,
|};
draggingFromThisWith
is populated with the draggableId
of the dragging item in the π home list for the entire drag. This allows you to style the home list differently from the other lists if you like. We recommend you do this to make it clear which list the user is dragging from. You can read more about the thinking behind this in my blog
In this example we set the background-color
of the home list to pink
when we are dragging over the list. We set the background-color
of the home list to blue
when not dragging over the home list.
Overhauled documentation π
We have totally overhauled our documentation to be easier to find and use information.
Our README.md
was getting too big. It was hard for new comers to know where to start, and it was hard for consumers of the library to quickly find information. We have simplified our README.md
and added a comprehensive documentation section which links off to categories of information.
The README.md
now contains a high-level introduction to the project, feature set, how to get started and then links to the rest of the documentation.
For new comers and evaluators πΆ
Information in the README.md
:
- Big
gif
showing off what the library is about - Reduced badges to the critical subset to reduce noise
- High level motivation of the project
- Feature set and limitations
- Getting started guide
- An
About
section which links off information useful to understand, evaluate and get started with the library.
For consumers π©βπ€
- The documentation has been broken down into more logical independent chunks to make it clearer where to find topical information. Previously
API
was under a wall of design thinking information which would not be useful to a non-evaluator - All documentation pages have a link back to the documentation root for quickly jumping around
A big thank you to @Noviny for his incredible efforts in helping to create this new information architecture
Documentation engineering health π¨ββοΈ
- Added a test to ensure that there are no more broken links π. This is by far the cause of the most issues
- Added test to ensure consistent usage of component names
- Added test to ensure all documentation pages end with a backlink to the documentation root
Refreshed examples π¦ΈββοΈ #1096
We have made our examples brighter and more colourful
Previous | New style |
---|---|
We have also upgraded to Storybook v5
and added our new logo to the header
Other improvements
- Only attaching to the
redux
devtools whenprocess.env.NODE_ENV === 'development'
#1085. Thanks @teleaziz!! - More collision engine refinements
Fixes π
- A
transition
for a property other thantransform
that finished during a drop animation on aDraggable
could cause the drop to end suddenly #1096 position:fixed
lists on ie11 #1091 #1088. Thanks @InvisibleBacon for raising this one- Typo in docs #1102. Thanks @jDeppen
Engineering health
Happy new year! π
I have been off for a month and I had a great time. I could go for another month...
Here is what we plan on working on next: Up next milestone
Also, a reminder that we have two planned breaking changes coming up:
- Moving to a
React
peer dependency of 16.6+ #984 - Moving from
flow
toTypeScript
#982. This might not be a breaking change if we have a goodflow
story π€
Unplanned but possible:
- Going all in on hooks and creating a new hook based api π€π
Improvements
doctype
warning π¨ββοΈ #1068
We have added a dev only warning to notify you if the pages doctype
is supported #1068 #1040. We only support the html5 doctype
(<!doctype html>
). The doctype
impacts a number of measurement calculations. Previously if you used another doctype
things might work, but there could be some strange issues depending on your css configuration (eg if you did not have body height: 100%
). Rather than trying to account for every strange doctype
out there - we print a dev warning to the console
! π We also now call this out in the docs.
My hope is that this change makes it easier to fall into the pit of success
Just a public service announcement: if you do not add a doctype
tag, browsers will use some other doctype
such as "Quirks mode". Please add a doctype
to your pages!
<!doctype html>
Other
- More robust ref type checking #1067. Allows usage of
react-frame-component
#796. Thanks @jdillick for the ideas - Fixing typos in docs #1059 #1062 #1042 #1063 #1011 #1070 #1071. Thanks @niwsa, @mego and @hannesvdvreken!
Engineering health
Improvements
π Delicious auto scrolling #349
Previously when a user picked up a draggable item that was near the edge of a scrollable area it could trigger a large and jarring auto scroll. We have created a solution that gets around this problem in a really elegant way.
I have created a video which goes over what we did in a bit of detail πΊ
A big thank you to @ajaymathur and @jjcm for their huge help with this one
β€οΈ Docs
- New: Common setup guide: Common setup issues that are hard for us to catch in code. We also linked to this from our
ISSUE_TEMPLATE.md
- New: Identifiers guide: How to use identifiers (
draggableId
anddroppableId
) successfully inreact-beautiful-dnd
- Fixing typo in
README.md
link #970. Thanks @fzembow!! - Fixing links #955. Thanks @Blasz!!
- Removing outdated reference to
Draggable > type
#952. Thanks @DmitryGonchar!! - Fixing incorrect drop animation naming in our drop animation guide
Other
- Avoiding extra DOM read for
Droppable
s #966. Thanks @woohling! - Adding extra dev only checks for
Draggable
andDroppable
boolean props #945.
π Remember:
null
is not a validboolean
prop type. Either pass intrue
,false
orundefined
to get the ReactdefaultProp
value.
- Added new media section #943
- Updated collaborators list in
README.md
#943
π©ββοΈ Engineering health
- Adding additional server-side rendering test to ensure that client hydration works correctly #960. Thanks @petegleeson for the prior work on this
- Fixed flaky builds on CircleCi caused by
flow
crashing. Thanks @TrySound for your help with this one! #1008 #1007 - Moved to
flow
0.88
- Moved to storybook
4x
#989 - Moved back from
emotion
to my first love:styled-components
4x
for examples. - Upgrading dev dependencies #1005
π£ Announcement: breaking change coming
We have two planned breaking changes coming up in early 2019:
My hope is to batch these two breaking changes into a single breaking change release
Fixes
- Better scroll container detection for the
<body>
element. Fixes compatability issue with Semantic UI #904. We also wrote a new guide: how we detect scroll containers. Thanks @MohammedAl-Mahdawi for raising this and @seancurtis for helping me work through it.
Improvements
- Improved cross axis movement into an empty list #936. Allows for movement of bigger items into a list that might not have enough current space for it. Thanks @lukyth for raising this one!
- Even more spacing accuracy for different box model configurations #919. Also created a spacing stress test example
- Adding more detail to issue creation template #934
- Removing unneeded
Array.prototype.find
message from docs #933. Thanks @wuweiweiwu! - Adding note about
styled-components
4.0
api change to docs #926. Thanks @balonsom!! - Improvements to
README.md
#943