-
Notifications
You must be signed in to change notification settings - Fork 768
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Two "Seen by" dots shown #7882
Comments
I guess this is coming from the recent changes about read receipt for threads... |
Same problem noticed here. Samsung Galaxy S22 |
I believe this is what @SpiritCroc was referring to here: #6342 (comment). They say they believe it is because of a read marker for both "null" and "main" threads causing issues. I am also having issues with notifications on Android that started at the same time as the double read markers appeared. Possibly not notifying because it's wrongly considered read? |
From my experience, it only looks like this due to to reactions. When enabling showing hidden events in the developer options, one receipt is usually on the message itself, and another on the reaction. |
For me, I don't have hidden events enabled. And I see this problem even on regular texts without reactions. But I can see how hidden events could be involved. When I've ended a call, the other person's read markers can sometimes double up without any messages in between, which could mean there's a hidden event between them. |
This does not align with my experience. I've been having this issue for weeks and have not personally observed any interactions with reactions |
On notifications: I simply don't receive them most the time since this behavior began (Google Pixel 5, up-to-date Android, installed from F-Droid). They get vibration for one second with the notification showing briefly before vanishing (Google Pixel 7, up-to-date Android, installed from Play Store). Not sure if this issue is related or if it is a separate problem that started around the same time. |
I seem to get this behavior sometimes too, but not all the time(?) - I'm also Google Pixel 7, current Android, Play Store. |
Note I'm not saying duplicate read receipts are related to reactions. My point is that whenever it looks like duplicated read receipts are attached to the same message, that's usually just because the receipts are "flattened" when there are reactions, and once you have hidden events enabled, the receipts still attach to different events. |
Investigating this issue. Confirming the Event But there I do not see the issue on the device. I guess when sending an Event to a room we add a RR to the database, but not using |
Poked around in some blames. I'm guessing the issue can be tracked down to #7474. Matches the timeline and I don't see more recent changes on thread ids' interactions with read receipts since. Lots of null checking going on because It could also be an issue with migrating the database #7640? Maybe a If this is not a ubiquitous issue, I would guess some edge case happened to a subset of users. Some uncommon logic state, or the migration not properly handling a certain type of events, that sort of thing. |
BTW: possible duplicate of #6342. |
FWIW, I've seen the same chat head placement for the same chat across multiple devices on multiple Elements/Element-based clients since the threads update, and it persists across initial syncs. So it seems 100% deterministic from my experiments / if I haven't missed any case. |
I'd be happy to take a look at the data for the room I'm experiencing this issue in. I'm not an Android dev — is the best approach setting up an emulator through Android Studio, logging into my account, and using Flipper per flipper.md? |
Yes. The main issue is that you will not be able to query the database with Flipper. So if there are lots of entries, it can be painful to inspect. |
Okay, here are my results. I had forgotten how much of a pain android development was to get set up the first time. Not sure what all needs to be censored so I was liberal with it. Note that on the database photos, my friend is censored in pink to distinguish the different TestState AState B
State CFriend's perspectiveI had my friend send a screenshot at the end. They said that my second read receipt was also on the
|
I've just seen 2 read markers from the same user occupy the same position, meaning that they were grouped. Until now, I've only seen them vertically adjacent, never grouped. |
For me there is one read receipt which is real which shows up when the user has actually seen the message, and then there is a fake read receipt that shows up (usually) instantly whenever a message is sent. When the user has actually seen the most recent message, then both read receipts will be on that message, when he hasn't there will be the fake read receipt on the most recent message and the real one somewhere above on the last message he actually saw. And this is only visible on android. On element desktop, I can only see the fake read receipt from android users, so it is impossible to tell if they actually saw the message or not. |
I am still seeing this pretty much daily on the latest version of the app. |
Never stopped for me on all versions since reported. |
Like others said, still affects my experience. Would be happy to provide any logs/debug information per request. |
Also +1 from here. While not mos urgent thing, most irritating it is. |
…_receipts Fix multiple read receipts for the same user in timeline #7882
Steps to reproduce
In the messages view, in a conversation with one person, I see two "Seen by" dots - one after my most recent message the other user has seen, and one after the last message they sent. Something like this (x represents the "Seen by" dot):
Outcome
What did you expect?
I expect to see only one "Seen by" dot, after the most recent of my messages the other user has seen.
What happened instead?
I see two.
Your phone model
Google Pixel 7
Operating system version
Android 13
Application version and app store
Element version 1.5.11
Homeserver
non-public, synapse
Will you send logs?
Yes
Are you willing to provide a PR?
No
The text was updated successfully, but these errors were encountered: