This is due to some kind of bug in ListView that never resettles
properly for bottom-to-top views. This can arise when the pinned message
is loaded (because that squishes the view) or the window is resized
(because that also resizes the view.)
We can work around it by assuming the following:
1. The RoomPage knows the window is resizing because it gets its height
changed before TimelineView.
2. The first height change can be a marker to position the view at the
beginning.
This fixes the issue for me, I did the following in order to test this:
* Switch between many rooms, especially ones with a pinned message. Now
all of them start at the bottom as they should.
* Resize the window, ensure that if you scrolled it stays around that
position - otherwise it sticks at the bottom.
This fixes the annoying "I just scrolled down to the bottom, how come
NeoChat doesn't think I did?"
From what I can tell this is also ListView bug (or something caused by
our style/Kirigami) that creates cases like contentY being -643.2 (for a
ListView of height 643) thus that's not "at Y's end". For our case
though, we don't care and can safely round it.
There's two parts to making this work mainly:
1. Use getEvent instead of findInTimeline so the related event is
actually found.
2. Close the dialog once a reply relation is found, so you can easily
reply in the chat bar.
This is another thing that enables us to view multiple rooms in a single
timeline. Specifically, this improves the experience in room search
going across room versions and getting a correct readOnly status (for
hiding certain hover actions.)
Contrary to popular belief (unfortunately) these reports are *only* sent
to your own server, which is then opaquely handled in some unknownable
way.
See #707
Not all events that are processed in this model belong to the room, e.g.
searching through multiple room versions. Now the model finds the
correct room based on the reported room in the event.
This fixes searching through upgraded rooms, and unblocks searching
through multiple rooms in the future.
We were previously assuming that we always want to parent these menus to
RoomPage, but that only exists on the main window. If you tried to open
the menu for say - the search window - then it would confusingly still
open on the main menu.
Thankfully the way to fix this is simple, by passing a parent QtObject
around.
These are more common than we thought, good examples are pinned or
searched messages - which are not going to be in the room's history
unless you happen to have them loaded. But currently our message menu
infrastructure expects them to be, since its looked up by the room +
event ID.
To fix this is simple, we now move the job of finding the event to the
caller which may use a model instead. I didn't fix all existing
call-sites yet, mainly the message menu opening one since that was the
most obvious bug. But this opens up the door for other assumptions about
room history to be fixed too.
I had to do a bit of C++ re-jiggering in order to expose useful
functions to QML.
You were previously relegated to looking at any avatars or a buggy
tooltip, but suffer no longer! If you tap on a message's read marker, a
dialog will pop up listing the users who have read it.
Uou can also view their profiles from here, etc.
`room()->decryptMessage()` returns null if the message fails to decode. Since elements of `m_pinnedEvents`
get directly dereferenced in getEventForIndex, storing null values leads to a segfault.
In this case we should retain the EncryptedEvent to let the UI report the error.
Some further belt and braces to ensure that we don't crash in the timeline. This does the following:
- Checks all callback inputs
- Makes sure that any objects are cleaned up from incubators when deleting the delegate as it turns out that clear() doesn't do this. Hopefully this stops callbacks into an already deleted parent.
Hopefully helps with https://crash-reports.kde.org/organizations/kde/issues/261627/events/0b6bef500c5641e1924aa0bc2b0c11bd/
Hopefully this stops any crashes around QuickActions.qml and EmojiDialog.qml. Best I can guess this is some race condition where QuickActions are deleted in the time it takes to instnatiate the EmojiDialog popup. I've also rearranged the updateQuickActions function to stop a possible race condition there.
BUG: 509484
I broke this in 409b6da160 making it not
possible to join a succeeding room. I decided to keep the old behavior
of not prompting you to join, because:
1. You are intentionally choosing to tap the button.
2. Room successions aren't random, a privileged user in the room is
creating it and there's already some implicit existing trust
(Also the current join confirmation dialog isn't useful enough yet to
make an actual informed decision anyway.)
This wasn't a thing yet, so the search experience compared to something
like Element Web was much worse. Now you can scroll back as far as any
other client can!
I had to do a bit of refactoring to stop resetting the model all the
time, and append instead. Otherwise, it was quite straightforward to
implement.
This fixes an always-reproducible bug by switching from Bubbles (with
"move local messages to the right" enabled) to Compact, and your
avatar is missing until you switch rooms or restart NeoChat.
The various flags to inform the rest of the delegate of the avatar were
returning true, but the actual avatar item was never created when the
option was switched on.
In other messaging applications (e.g. Discord) this is possible through
text modifiers like "from:@user". We don't support that, and I'm not
super keen on implementing yet-another-parsing-thing, so an action in
the user detail dialog should work for now.
Very useful to sift through large rooms but when you only care about a
specific person's messages (maybe your own?)
The menus have always been split into a menu for file-like content and text-like content
This split makes some things a bit more complicated then necessary.
So I noticed that searching for messages feels "buggy". I dove into the
code and gave it a pretty nice refresh, I know you'll like it once you
try it!
Some of the issues I noticed and are now fixed:
* The search doesn't clear immediately when you clear the text box, it's
delayed like everything else. This also has the knock-on effect of
starting random network requests.
* If you now press the search button (either with a mouse, pen, etc. or
by the keyboard) the search delay timer was still running.
* The "nothing is here" placeholder message appears when the search
timer is running, making you think no messages were found.
These changes also help other places where SearchPage is used, not just
message search.
Apply all the required styling to show links, table, spoilers, etc in cpp. This also updates the method of revealing spoilers so now you can click to reveal then click again to hide.
Separate out a base `MessageContentModel` that can be extended to get the component types from different places. This is used currently for `EventMessageContentModel` but will be used later as part of the rich chat bar.
All display text is now in the text component so it never needs special casing. This also cleans up some of the model parameters so more things come from attributes including location and file data (which was already a qvariantmap anyway).
Also cleaned up the itinerary and file enhancement views,
Add a dev setting to allow relations to be sent to any message using the right click delegate menu. This shouldn't be a main feature, but since it's technically allowed in matrix this will help any future debugging
