Summary:
This diff does a couple things:
- Undo redo with a new undo/redo store that maintains it's own queue of undo/redo tasks. This queue is separate from the TaskQueue because not all tasks should be considered for undo history! Right now just the AddRemoveTagsTask is undoable.
- NylasAPI.makeRequest now returns a promise which resolves with the result or rejects with an error. For things that still need them, there's still `success` and `error` callbacks. I also added `started:(req) ->` which allows you to get the underlying request.
- Aborting a NylasAPI request now makes it call it's error callback / promise reject.
- You can now run code after perform local has completed using this syntax:
```
task = new AddRemoveTagsTask(focused, ['archive'], ['inbox'])
task.waitForPerformLocal().then ->
Actions.setFocus(collection: 'thread', item: nextFocus)
Actions.setCursorPosition(collection: 'thread', item: nextKeyboard)
Actions.queueTask(task)
```
- In specs, you can now use `advanceClock` to get through a Promise.then/catch/finally. Turns out it was using something low level and not using setTimeout(0).
- The TaskQueue uses promises better and defers a lot of the complexity around queueState for performLocal/performRemote to a task subclass called APITask. APITask implements "perform" and breaks it into "performLocal" and "performRemote".
- All tasks either resolve or reject. They're always removed from the queue, unless they resolve with Task.Status.Retry, which means they internally did a .catch (err) => Promise.resolve(Task.Status.Retry) and they want to be run again later.
- API tasks retry until they succeed or receive a NylasAPI.PermanentErrorCode (400,404,500), in which case they revert and finish.
- The AddRemoveTags Task can now take more than one thread! This is super cool because you can undo/redo a bulk action and also because we'll probably have a bulk tag modification API endpoint soon.
Getting undo / redo working revealed that the thread versioning system we built isn't working because the server was incrementing things by more than 1 at a time. Now we count the number of unresolved "optimistic" changes we've made to a given model, and only accept the server's version of it once the number of optimistic changes is back at zero.
Known Issues:
- AddRemoveTagsTasks aren't dependent on each other, so if you (undo/redo x lots) and then come back online, all the tasks try to add / remove all the tags at the same time. To fix this we can either allow the tasks to be merged together into a minimal set or make them block on each other.
- When Offline, you still get errors in the console for GET requests. Need to catch these and display an offline status bar.
- The metadata tasks haven't been updated yet to the new API. Wanted to get it reviewed first!
Test Plan: All the tests still pass!
Reviewers: evan
Reviewed By: evan
Differential Revision: https://phab.nylas.com/D1694
Summary:
If an email is to a group, the reply button should have `To` set the original email sender and `Cc` the group Fixes Phab ticket 2142. Previously, the click on the reply button was not captured unless the click occured on the reply or reply all icon. Changed the span to a div with no padding that captures all clicks
Test Plan: tested manually, unit tests
Reviewers: bengotow
Reviewed By: bengotow
Subscribers: evan
Maniphest Tasks: T2142
Differential Revision: https://phab.nylas.com/D1696
Summary:
Adding spec for message list star. Fixing multi-select in 3-pane starring
I didn't write a spec for thread-list because the thread-list-spec appears to be broken (returns immedietly at the top). I spent ~15 minutes attempting to fix but didn't make much progress
Test Plan: Added spec to message list
Reviewers: bengotow
Reviewed By: bengotow
Maniphest Tasks: T1957
Differential Revision: https://phab.nylas.com/D1697
Summary:
Fixes T1905 - Gost line in recipients field
Fixes T1877 - Collapsed dates should not show time
Test Plan: edgehill --test
Reviewers: bengotow
Reviewed By: bengotow
Maniphest Tasks: T1877, T1905
Differential Revision: https://phab.nylas.com/D1671
Summary: Adds attachment icon to collased message if message has attachment. I just copied the attachment icon in the thread-list to the message-list namespace in the static directory. If you want to do it another way, feel free to update
Test Plan: Tested locally. Not sure if this is worth a test, but I can write one if you want.
Reviewers: bengotow
Reviewed By: bengotow
Maniphest Tasks: T2085
Differential Revision: https://phab.nylas.com/D1687
Summary: Adding star button to message-toolbar view
Test Plan: Added spec for starring message from the message-toolbar view
Reviewers: bengotow
Reviewed By: bengotow
Differential Revision: https://phab.nylas.com/D1683
Summary: Fixes T2024 and renames unnamed files to Unnamed Attachment before appending the attachement component to message list
Test Plan: Added a test to ensure file without filename is renamed to Unnamed Attachment
Reviewers: bengotow
Reviewed By: bengotow
Maniphest Tasks: T2024
Differential Revision: https://phab.nylas.com/D1685
Summary:
Initial message list collapsing
messages can be expanded explicitly
styling message items
composer UI and collapsing
expanding and collapsing headers
style new reply area
adding in message controls
Add message actions dropdown
Test Plan: edgehill --test
Reviewers: bengotow
Reviewed By: bengotow
Differential Revision: https://phab.nylas.com/D1664
Summary: We now show the inspirational quotes only when in list mode and viewing a tag. When you're viewing search results, or when you're in three-pane mode, you now see a more generic empty state.
Test Plan: No tests yet, may want to see if this refactor sticks when we start adding more empty states
Reviewers: evan
Reviewed By: evan
Differential Revision: https://phab.nylas.com/D1642
Summary: Fixes T1191
Test Plan: No new tests to see here.
Reviewers: evan
Reviewed By: evan
Maniphest Tasks: T1191
Differential Revision: https://phab.nylas.com/D1603
Summary:
ScrollRegion with support for tooltips shown as you scroll, custom tooltips for thread list and message list.
fix(specs): all other scrollbars hidden to prevent incompatibility
fix scrollbar sizing when used in conjunction with resizableregion
Test Plan: Need to write tests and docs for ScrollRegion - no tests yet
Reviewers: evan
Reviewed By: evan
Differential Revision: https://phab.nylas.com/D1597
Summary: Add docs for new RetinaImg modes
Test Plan: Not much to test, except that it looks good!
Reviewers: evan
Reviewed By: evan
Differential Revision: https://phab.nylas.com/D1595
Summary:
This diff attempts to improve the responsiveness of the app when you hit "Reply". This is achieved by being smarter about creating the draft and loading it into the draft store, and also by allowing the compose* actions to take objects instead of just IDs (resulting in a fetch of the object).
Allow Actions.composeReply,etc. to optionally be called with thread and message objects instead of IDs. This prevents a database lookup and the data is "right there."
Create DraftStoreProxy for new drafts optimistically—this allows us to hand it the draft model we just created and it doesn't have to go query for it
When we create a new Draft, immediately bind it to a LocalId. This means that when the MessageStore receives the trigger() event from the Database, it doesn't have to wait while a localId is created
When MessageStore sees a new Message come in which is on the current thread, a draft, and not in the localIds map, assume it's a new draft and shortcut fetchFromCaceh to manually add it to the items array and display. This means the user sees the...
...draft instantly.
Remove delays from focusing draft, scrolling to draft after content is ready. I actually removed these thinking it would break something, and it didn't break anything.... Maybe new Chromium handles better?
Fix specs
Test Plan: Run specs - more in progress right now
Reviewers: evan
Reviewed By: evan
Differential Revision: https://phab.nylas.com/D1598
Summary:
Fix T1822 - saving templates not working, not showing template
Fix T1800 - give composers a minimum size
Fix the bottom bar of the composer so the gray bar goes all the way across in popout mode.
Fix T1825 - switch to a more attractive "June 4, 2015 at 3:10 PM" styling for expanded dates
Remove, rather than hide, react components for text fields in composer. Fixes T1147
Fix specs
Switch to 999+ instead of infinity. Fixes T1768
Fix broken TemplateStore specs
Test Plan: Run tests
Reviewers: evan
Reviewed By: evan
Maniphest Tasks: T1147, T1768, T1822, T1800, T1825
Differential Revision: https://phab.nylas.com/D1601
Summary:
Features:
- ThreadListParticipants ignores drafts when computing participants, renders "Draft" label, pending design
- Put the WorkspaceStore in every window—means they all get toolbars and custom gumdrop icons on Mac OS X
Bug Fixes:
- Never display notifications for email the user just sent
- Fix obscure issue with DatabaseView trying to update metadata on items it froze. This resolves issue with names remaining bold after marking as read, drafts not appearing in message list immediately.
- When you pop out a draft, save it first and *wait* for the commit() promise to succeed.
- If you scroll very fast, you node.contentWindow can be null in eventedIframe
Other:
Make it OK to re-register the same component
Make it possible to unregister a hot window
Break the Sheet Toolbar out into it's own file to make things manageable
Replace `package.windowPropsReceived` with a store-style model where anyone can listen for changes to `windowProps`
When I put the WorkspaceStore in every window, I ran into a problem because the package was no longer rendering an instance of the Composer, it was declaring a root sheet with a composer in it. This meant that it was actually a React component that needed to listen to window props, not the package itself.
`atom` is already an event emitter, so I added a `onWindowPropsReceived` hook so that components can listen to window props as if they were listening to a store. I think this might be more flexible than only broadcasting the props change event to packages.
Test Plan: Run tests
Reviewers: evan
Reviewed By: evan
Differential Revision: https://phab.nylas.com/D1592
Summary:
This fixes T1744 and adds specs for the logic in _attachmentComponents.
- Attachments should display even when the body is an empty string or null
- Attachments should not include files without names (mime parts of the message)
Test Plan: Run 2 new tests
Reviewers: evan
Reviewed By: evan
Maniphest Tasks: T1744
Differential Revision: https://phab.nylas.com/D1566
Summary:
Fixes T1731
Chrome by default on a user-agent stylesheet sets white-space to `pre`.
For `pre` elements. This is useful in most cases since whitespace is taken
literally.
Unfortunately when people send out emails with code in them, the `pre`
blocks can get mangled because there may be unnecessary whitespace in
them.
I personally think the default behavior is much better, however, I think
we should have `pre` blocks render with `white-space: normal` because
anyone who's sending an email with code it in will probably use <span> and
<br> and inline styles to ensure it looks correct on most readers. Since
many places (like Outlook) strip whitespace out of HTML, I think we should
mimic that behavior too.
Test Plan: edgehill --test
Reviewers: bengotow
Reviewed By: bengotow
Maniphest Tasks: T1731
Differential Revision: https://phab.nylas.com/D1561
Summary:
atom-window `sendMessage` was not the same as `browserWindow.webContents.send`. WTF.
Save current namespace to config.cson so that it is never null when window opens
Don't re-create thread view on namespace change unless the namespace has changed
Tests for NamespaceStore state
Push worker immediately in workerForNamcespace to avoid creating two connections per namespace
Allow \n to be put into sreaming buffer, but only one
Clear streaming buffer when we're reconnecting to avoid processing same deltas twice (because of 400msec throttle)
Make `onProcessBuffer` more elegant—No functional changes
Test Plan: Run tests!
Reviewers: evan
Reviewed By: evan
Differential Revision: https://phab.nylas.com/D1551
Summary:
Fixes T1327
Fixes T1294
Fixes T1307
Fixes T1207
- More robus Undo Manager
- Scrolls to the most recent undo location
- Fixes a variety of bugs when the composer's cursor ends up in an unsupported
location
- Fixes auto-scroll detection at end of content
Test Plan: edgehill --test
Reviewers: bengotow
Reviewed By: bengotow
Maniphest Tasks: T1207, T1294, T1307, T1327
Differential Revision: https://phab.nylas.com/D1529
Summary:
Fixes: T1334
remove final InboxApp references
move out all underscore-plus methods
Mass find and replace of underscore-plus
sed -i '' -- 's/underscore-plus/underscore/g' **/*.coffee
sed -i '' -- 's/underscore-plus/underscore/g' **/*.cjsx
Test Plan: edgehill --test
Reviewers: bengotow
Reviewed By: bengotow
Differential Revision: https://phab.nylas.com/D1534
Summary: Re-order the header and toolbar so that the search dropdown appears correctly
Test Plan: Run tests
Reviewers: evan
Reviewed By: evan
Differential Revision: https://phab.nylas.com/D1524
Summary:
Focused Content Store should notify observers when focused items change, not just when they're reassigned
Popovers should have a `direction` and optional event onOpened
Menu divider documentation was wrong, menus should support checked items by default
Pressing escape in a popover's input should dismiss the popover
Other changes
Remove specs that make no sense anymore
Small tweak to report build breaking. Shouldn't happen often ;)
Test Plan: Run tests, which will now phone home if they break
Reviewers: evan
Reviewed By: evan
Differential Revision: https://review.inboxapp.com/D1493
Summary:
Load unread counts from database again, not tags
fix(multiselect-list): Clear selection on esc
fix(onboarding): Make target=_blank links work in onboarding pages
fix(workspace): Items in header and footer regions are in a single column
fix(layout): Critical issue for things not 100% height
fix(activity-bar): Show in dev mode so you know you're in dev mode
fix(quoted-text): Support for #divRplyFwdMsg quoted text marker
Test Plan: Run specs
Reviewers: evan
Reviewed By: evan
Differential Revision: https://review.inboxapp.com/D1484
Summary:
Loads the current user to pre-populate the Salesforce field in the
creators
add tabindex to tokenizing text fields
Test Plan: edgehill --test
Reviewers: bengotow
Reviewed By: bengotow
Differential Revision: https://review.inboxapp.com/D1468
Summary:
This diff makes a couple changes:
- New drafts are no longer created in the composer component. Draft creation is back in the draft store where it belongs!
- This means that the Draft List doesn't *always* contain two extra drafts, which are the unopened hot windows.
- Windows never show until they're loaded, which means windows open slowly vs. opening empty. I think the former is preferable but it's easy to change.
- Login window is now sized *before* it's shown instead of afterwards
- The mailto: behavior is now handled by the DraftStore, which gets rid of the whole draftInitialJSON thing that never made any sense.
fix(login) Make login window show at the correct size
logout from edgehill completely
Fix draft delete issue
Don't show windows, hot or cold, until they've loaded
Move logic for responding to mailto links into the draft store
Always fire `windowPropsChanged` after packages load or the window is shown
Show more error codes in red in activity bar
Make sure DraftStoreProxy doesn't interrupt the rendering of the composer with a new draft
Fix account-sidebar scroll issue, maybe?
Test Plan: Run tests!
Reviewers: evan
Reviewed By: evan
Differential Revision: https://review.inboxapp.com/D1479
Summary:
This diff gives the ComponentRegistry a cleaner, smaller API. Instead of querying by name, location or role,
it's now just location and role, and you can register components for one or more location and one or more
roles without assigning the entries in the registry separate names.
When you register with the ComponentRegistry, the syntax is also cleaner and uses the component's displayName
instead of requiring you to provide a name. You also provide the actual component when unregistering, ensuring
that you can't unregister someone else's component.
InjectedComponent and InjectedComponentSet now wrap their children in UnsafeComponent, which prevents
render/component lifecycle problems from propogating.
Existing components have been updated:
1. maxWidth / minWidth are now containerStyles.maxWidth/minWidth
2. displayName is now required to use the CR.
3. containerRequired = false can be provided to exempt a component from being wrapped in an UnsafeComponent.
This is useful because it's slightly faster and keeps DOM flat.
This diff also makes the "Show Component Regions" more awesome. It displays column regions, since they now
use the InjectedComponentSet, and also shows for InjectedComponent as well as InjectedComponentSet.
Change ComponentRegistry syntax, lots more work on safely wrapping items. See description.
Fix for inline flexbox scenarios (message actions)
Allow ~/.inbox/packages to be symlinked to a github repo
Test Plan: Run tests!
Reviewers: evan
Reviewed By: evan
Differential Revision: https://review.inboxapp.com/D1457
Summary:
#### WIP! ####
This is making it all work with the association endpoint, putting
together the Salesforce Sidebar interfaces, and getting the nested
creators/updaters working.
I still need to do a bunch of UI work and actually debug the whole
workflow still
---
rename SalesforceContactStore to SalesforceSearchStore
rename SalesforceContact to SalesforceSearchResult
salesforce sidebar changes
salesforce association picker
object form store fixes
figuring out newFormItem instigators
Make SalesforceObjectFormStore declarative off SalesforceObjectStore
Make action basd handlers for SalesforceObjectStore
sidebar store create and associate
salesforce sidebar and picker fixes
association works and displays on sidebar
salesforce object form fixes
object form fixes
fix salesforce updating
Test Plan: TODO
Reviewers: bengotow
Reviewed By: bengotow
Differential Revision: https://review.inboxapp.com/D1440
