moment/TODO.md

150 lines
5.5 KiB
Markdown
Raw Normal View History

- Refactoring
- Use [Animators](https://doc.qt.io/qt-5/qml-qtquick-animator.html)
- Sendbox
- SignIn/RememberAccount screens
- SignIn must be in a flickable
2019-07-13 19:34:58 +10:00
- Unfinished work in button-refactor branch
- Button can get "hoverEnabled: false" to let HoverHandlers work
- Room Sidepane
2019-07-21 23:26:47 +10:00
- Hide when window too small
- Also save/load its size
Big performance refactoring & various improvements Instead of passing all sorts of events for the JS to handle and manually add to different data models, we now handle everything we can in Python. For any change, the python models send a sync event with their contents (no more than 4 times per second) to JS, and the QSyncable library's JsonListModel takes care of converting it to a QML ListModel and sending the appropriate signals. The SortFilterProxyModel library is not used anymore, the only case where we need to filter/sort something now is when the user interacts with the "Filter rooms" or "Filter members" fields. These cases are handled by a simple JS function. We now keep separated room and timeline models for different accounts, the previous approach of sharing all the data we could between accounts created a lot of complications (local echoes, decrypted messages replacing others, etc). The users's own account profile changes are now hidden in the timeline. On startup, if all events for a room were only own profile changes, more events will be loaded. Any kind of image format supported by Qt is now handled by the pyotherside image provider, instead of just PNG/JPG. SVGs which previously caused errors are supported as well. The typing members bar paddings/margins are fixed. The behavior of the avatar/"upload a profile picture" overlay is fixed. Config files read from disk are now cached (TODO: make them reloadable again). Pylint is not used anymore because of all its annoying false warnings and lack of understanding for dataclasses, it is replaced by flake8 with a custom config and various plugins. Debug mode is now considered on if the program was compiled with the right option, instead of taking an argument from CLI. When on, C++ will set a flag in the Window QML component. The loading screen is now unloaded after the UI is ready, where previously it just stayed in the background invisible and wasted CPU. The overall refactoring and improvements make us now able to handle rooms with thousand of members and no lazy-loading, where previously everything would freeze and simply scrolling up to load past events in any room would block the UI for a few seconds.
2019-08-11 22:01:22 +10:00
- When qml syntax highlighting supports ES6 string interpolation, use them
- Fixes
2019-08-16 23:47:44 +10:00
- (Left?)Banner binding loop
Big performance refactoring & various improvements Instead of passing all sorts of events for the JS to handle and manually add to different data models, we now handle everything we can in Python. For any change, the python models send a sync event with their contents (no more than 4 times per second) to JS, and the QSyncable library's JsonListModel takes care of converting it to a QML ListModel and sending the appropriate signals. The SortFilterProxyModel library is not used anymore, the only case where we need to filter/sort something now is when the user interacts with the "Filter rooms" or "Filter members" fields. These cases are handled by a simple JS function. We now keep separated room and timeline models for different accounts, the previous approach of sharing all the data we could between accounts created a lot of complications (local echoes, decrypted messages replacing others, etc). The users's own account profile changes are now hidden in the timeline. On startup, if all events for a room were only own profile changes, more events will be loaded. Any kind of image format supported by Qt is now handled by the pyotherside image provider, instead of just PNG/JPG. SVGs which previously caused errors are supported as well. The typing members bar paddings/margins are fixed. The behavior of the avatar/"upload a profile picture" overlay is fixed. Config files read from disk are now cached (TODO: make them reloadable again). Pylint is not used anymore because of all its annoying false warnings and lack of understanding for dataclasses, it is replaced by flake8 with a custom config and various plugins. Debug mode is now considered on if the program was compiled with the right option, instead of taking an argument from CLI. When on, C++ will set a flag in the Window QML component. The loading screen is now unloaded after the UI is ready, where previously it just stayed in the background invisible and wasted CPU. The overall refactoring and improvements make us now able to handle rooms with thousand of members and no lazy-loading, where previously everything would freeze and simply scrolling up to load past events in any room would block the UI for a few seconds.
2019-08-11 22:01:22 +10:00
- Reloading config files (cache)
- Tiny invisible scrollbar
2019-07-26 09:19:39 +10:00
- Run import in thread and AsyncClient.olm functions, they block async loop
- Handle import keys errors
2019-07-25 14:10:07 +10:00
- Don't linkify images for outgoing html
- Message position after daybreak delegate
2019-07-21 20:05:01 +10:00
- Keyboard flicking against top/bottom edge
- Don't strip user spacing in html
2019-08-16 16:03:13 +10:00
2019-07-22 09:12:32 +10:00
- [hr not working](https://bugreports.qt.io/browse/QTBUG-74342)
- Terrible performance using `QT_QPA_PLATFORM=wayland-egl`, must use `xcb`
- UI
- Popup:
- label size
- Accept/cancel buttons
- Transitions
- Restoring UI state:
- Sendbox content
- Which element was focused
- Room member filter field
Big performance refactoring & various improvements Instead of passing all sorts of events for the JS to handle and manually add to different data models, we now handle everything we can in Python. For any change, the python models send a sync event with their contents (no more than 4 times per second) to JS, and the QSyncable library's JsonListModel takes care of converting it to a QML ListModel and sending the appropriate signals. The SortFilterProxyModel library is not used anymore, the only case where we need to filter/sort something now is when the user interacts with the "Filter rooms" or "Filter members" fields. These cases are handled by a simple JS function. We now keep separated room and timeline models for different accounts, the previous approach of sharing all the data we could between accounts created a lot of complications (local echoes, decrypted messages replacing others, etc). The users's own account profile changes are now hidden in the timeline. On startup, if all events for a room were only own profile changes, more events will be loaded. Any kind of image format supported by Qt is now handled by the pyotherside image provider, instead of just PNG/JPG. SVGs which previously caused errors are supported as well. The typing members bar paddings/margins are fixed. The behavior of the avatar/"upload a profile picture" overlay is fixed. Config files read from disk are now cached (TODO: make them reloadable again). Pylint is not used anymore because of all its annoying false warnings and lack of understanding for dataclasses, it is replaced by flake8 with a custom config and various plugins. Debug mode is now considered on if the program was compiled with the right option, instead of taking an argument from CLI. When on, C++ will set a flag in the Window QML component. The loading screen is now unloaded after the UI is ready, where previously it just stayed in the background invisible and wasted CPU. The overall refactoring and improvements make us now able to handle rooms with thousand of members and no lazy-loading, where previously everything would freeze and simply scrolling up to load past events in any room would block the UI for a few seconds.
2019-08-11 22:01:22 +10:00
- Combine events so they take less space
- After combining is implemented, no need to hide our own profile changes.
- When starting a long task, e.g. importing keys, quitting the page,
and coming back, show the buttons as still loading until operation is done
2019-07-25 06:44:35 +10:00
- Make invite/left banners look better in column mode
- Messages editing
- Code highlighting
- Support GIF avatars
2019-07-21 20:05:01 +10:00
- When reduced, show the full-window sidepane instead of Default page
- Adapt shortcuts flicking speed to font size and DPI
- Show error box if uploading avatar fails
- EditAccount page:
2019-08-16 16:03:13 +10:00
- Remove account from client
- state: Set UI state page to Default.qml when account is removed
- Device settings
- Multiaccount aliases:
- Warn when conflict with another alias
- Forbid spaces?
- Add an explanation tooltip
- Prevent sending messages with an user not in room
- Support \ escaping
- Improve avatar tooltips position, add stuff to room tooltips (last msg?)
- Accept drag and dropping a picture in account settings to set avatar
2019-07-18 22:10:36 +10:00
- Show something when connection is lost or 429s happen
- "Rejoin" LeftBanner button if room is public
2019-07-09 02:52:07 +10:00
- Daybreak color
Big performance refactoring & various improvements Instead of passing all sorts of events for the JS to handle and manually add to different data models, we now handle everything we can in Python. For any change, the python models send a sync event with their contents (no more than 4 times per second) to JS, and the QSyncable library's JsonListModel takes care of converting it to a QML ListModel and sending the appropriate signals. The SortFilterProxyModel library is not used anymore, the only case where we need to filter/sort something now is when the user interacts with the "Filter rooms" or "Filter members" fields. These cases are handled by a simple JS function. We now keep separated room and timeline models for different accounts, the previous approach of sharing all the data we could between accounts created a lot of complications (local echoes, decrypted messages replacing others, etc). The users's own account profile changes are now hidden in the timeline. On startup, if all events for a room were only own profile changes, more events will be loaded. Any kind of image format supported by Qt is now handled by the pyotherside image provider, instead of just PNG/JPG. SVGs which previously caused errors are supported as well. The typing members bar paddings/margins are fixed. The behavior of the avatar/"upload a profile picture" overlay is fixed. Config files read from disk are now cached (TODO: make them reloadable again). Pylint is not used anymore because of all its annoying false warnings and lack of understanding for dataclasses, it is replaced by flake8 with a custom config and various plugins. Debug mode is now considered on if the program was compiled with the right option, instead of taking an argument from CLI. When on, C++ will set a flag in the Window QML component. The loading screen is now unloaded after the UI is ready, where previously it just stayed in the background invisible and wasted CPU. The overall refactoring and improvements make us now able to handle rooms with thousand of members and no lazy-loading, where previously everything would freeze and simply scrolling up to load past events in any room would block the UI for a few seconds.
2019-08-11 22:01:22 +10:00
- Conversation breaks: show time of first new msg after break instead of big
blank space
- Replies
- `pyotherside.atexit()`
- Sidepane
- Header back button when reduced
- Better look for arrows and option button when collapsed
- Way to put it back to auto-sizing (snap)
- Show it when hovering on the left when collapsed/reduced
- Ability to drag on any place of the pane to resize
- Reducable room sidepane, swipe to show full-window
2019-07-09 02:52:07 +10:00
- Invite to room
- Server selection
- Register/Forgot? for SignIn dialog
- Add room
- Leave room
- Forget room warning popup
- Prevent using the SendBox if no permission (power levels)
Big performance refactoring & various improvements Instead of passing all sorts of events for the JS to handle and manually add to different data models, we now handle everything we can in Python. For any change, the python models send a sync event with their contents (no more than 4 times per second) to JS, and the QSyncable library's JsonListModel takes care of converting it to a QML ListModel and sending the appropriate signals. The SortFilterProxyModel library is not used anymore, the only case where we need to filter/sort something now is when the user interacts with the "Filter rooms" or "Filter members" fields. These cases are handled by a simple JS function. We now keep separated room and timeline models for different accounts, the previous approach of sharing all the data we could between accounts created a lot of complications (local echoes, decrypted messages replacing others, etc). The users's own account profile changes are now hidden in the timeline. On startup, if all events for a room were only own profile changes, more events will be loaded. Any kind of image format supported by Qt is now handled by the pyotherside image provider, instead of just PNG/JPG. SVGs which previously caused errors are supported as well. The typing members bar paddings/margins are fixed. The behavior of the avatar/"upload a profile picture" overlay is fixed. Config files read from disk are now cached (TODO: make them reloadable again). Pylint is not used anymore because of all its annoying false warnings and lack of understanding for dataclasses, it is replaced by flake8 with a custom config and various plugins. Debug mode is now considered on if the program was compiled with the right option, instead of taking an argument from CLI. When on, C++ will set a flag in the Window QML component. The loading screen is now unloaded after the UI is ready, where previously it just stayed in the background invisible and wasted CPU. The overall refactoring and improvements make us now able to handle rooms with thousand of members and no lazy-loading, where previously everything would freeze and simply scrolling up to load past events in any room would block the UI for a few seconds.
2019-08-11 22:01:22 +10:00
- Prevent using an alias if that user is not in the room or no permission
- Spinner when loading account, past room events, images or clicking buttons
- Show account page as loading until profile initially retrieved
- Theming
- Support SVG hue via image provider for icons
2019-07-25 08:41:40 +10:00
- Don't create additional lines in theme conversion (braces)
- Recursively merge default and user theme
- Distribute fonts
- preferredIconPack: accept multiple values
- Find icon packs in user data dir
- Correctly implement uiScale/fontScale + ctrl+-= keys
- See `QCoreApplication::setAttribute(Qt::AA_EnableHighDpiScaling)`
- See [Text.fontSizeMode](https://doc.qt.io/qt-5/qml-qtquick-text.html#fontSizeMode-prop)
- Way to round avatar corners to allow box radius
- If avatar is set, name color from average color?
- Accent color from background
- Settings page
- Message/text selection
2019-07-16 06:14:08 +10:00
- Custom file picker for Linux...
- Major features
- E2E
- Device verification
- Edit/delete own devices
- Request room keys from own other devices
- Auto-trust accounts within the same client
- Import/export keys
- Uploads & proper http thumbnails
- Reduce messages ListView cacheBuffer height once http thumbnails
downloading is implemented
- Read receipts
- Status message and presence
- Links preview
- Client improvements
Big performance refactoring & various improvements Instead of passing all sorts of events for the JS to handle and manually add to different data models, we now handle everything we can in Python. For any change, the python models send a sync event with their contents (no more than 4 times per second) to JS, and the QSyncable library's JsonListModel takes care of converting it to a QML ListModel and sending the appropriate signals. The SortFilterProxyModel library is not used anymore, the only case where we need to filter/sort something now is when the user interacts with the "Filter rooms" or "Filter members" fields. These cases are handled by a simple JS function. We now keep separated room and timeline models for different accounts, the previous approach of sharing all the data we could between accounts created a lot of complications (local echoes, decrypted messages replacing others, etc). The users's own account profile changes are now hidden in the timeline. On startup, if all events for a room were only own profile changes, more events will be loaded. Any kind of image format supported by Qt is now handled by the pyotherside image provider, instead of just PNG/JPG. SVGs which previously caused errors are supported as well. The typing members bar paddings/margins are fixed. The behavior of the avatar/"upload a profile picture" overlay is fixed. Config files read from disk are now cached (TODO: make them reloadable again). Pylint is not used anymore because of all its annoying false warnings and lack of understanding for dataclasses, it is replaced by flake8 with a custom config and various plugins. Debug mode is now considered on if the program was compiled with the right option, instead of taking an argument from CLI. When on, C++ will set a flag in the Window QML component. The loading screen is now unloaded after the UI is ready, where previously it just stayed in the background invisible and wasted CPU. The overall refactoring and improvements make us now able to handle rooms with thousand of members and no lazy-loading, where previously everything would freeze and simply scrolling up to load past events in any room would block the UI for a few seconds.
2019-08-11 22:01:22 +10:00
- Image provider: on failed conversion, way to show a "broken image" thumb?
2019-07-25 07:05:27 +10:00
- Config file format
- Initial sync filter and lazy load, see weechat-matrix `_handle_login()`
- See also `handle_response()`'s `keys_query` request
- Direct chats category
- Markdown: don't turn #things (no space) and `thing\n---` into title,
disable `__` syntax for bold/italic
- Push instead of replacing in stack view (remove getMemberFilter when done)
- `<pre>` scrollbar on overflow
- When inviting someone to direct chat, room is "Empty room" until accepted,
it should be the peer's display name instead.
- Animate RoomEventDelegate DayBreak apparition
- Room subtitle: show things like "*Image*" instead of blank, etc
- nio
2019-07-26 09:19:39 +10:00
- `AsyncClient.share_group_session`: send device batches concurrently
- downloads API
- MatrixRoom invited members list
- Invite events are missing their timestamps (needed for sorting)
- Left room events after client reboot
- `org.matrix.room.preview_urls` event
- `m.room.aliases` event
- Support "Empty room (was ...)" after peer left
- Previewing room without joining
- Distribution
2019-07-16 18:37:39 +10:00
- Include python dependencies in binary with rcc?
- README.md