feat: expose m.fully_read event ID on RoomInfo by danderson-cont · Pull Request #6569 · matrix-org/matrix-rust-sdk

danderson-cont · 2026-05-13T06:13:50Z

Exposes the m.fully_read event ID on RoomInfo for clients' "jump-to-unread" button

I've documented the public API Changes in the appropriate CHANGELOG.md files.
This PR was made with the help of AI.

Signed-off-by: Daniel Anderson daniel.anderson@toptal.com

codecov · 2026-05-13T06:25:32Z

Codecov Report

❌ Patch coverage is 95.72650% with 5 lines in your changes missing coverage. Please review.
✅ Project coverage is 89.94%. Comparing base (4a26af8) to head (703da91).
✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
crates/matrix-sdk-base/src/sliding_sync.rs	93.97%	5 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #6569      +/-   ##
==========================================
- Coverage   89.95%   89.94%   -0.01%     
==========================================
  Files         382      382              
  Lines      107797   107914     +117     
  Branches   107797   107914     +117     
==========================================
+ Hits        96964    97066     +102     
- Misses       7165     7180      +15     
  Partials     3668     3668

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

codspeed-hq · 2026-05-13T06:39:27Z

Merging this PR will not alter performance

✅ 50 untouched benchmarks

_{Comparing danderson-cont:danderson-cont/expose-fully-read-event-id (703da91) with main (4a26af8)}

stefanceriu · 2026-05-13T07:31:02Z

+                    AnyRoomAccountDataEvent::FullyRead(_) => {
+                        context
+                            .room_info_notable_updates
+                            .entry(room_id.to_owned())
+                            .or_default()
+                            .insert(RoomInfoNotableUpdateReasons::READ_RECEIPT);


The fully read marker isn't a read receipt, is it?

Right. I've added a new flag FULLY_READ, which required switching the field from a u8 to u16

Hywan

Hello,

Thank you for your contribution.

First and foremost, did you read our AI policy? I'm almost certain that this PR has been vibe-coded.

Second, the pull request template asks a couple of questions that you've apparently erased, so I'm copying them here:

<!-- description of the changes in this PR -->

- [ ] I've documented the public API Changes in the appropriate `CHANGELOG.md` files.
- [ ] This PR was made with the help of AI.

<!-- Sign-off, if not part of the commits -->
<!-- See CONTRIBUTING.md if you don't know what this is -->
Signed-off-by:

About the code now: please can you add tests for the new feature?

Hywan · 2026-05-13T07:39:30Z

+                            .room_info_notable_updates
+                            .entry(room_id.to_owned())
+                            .or_default()
+                            .insert(RoomInfoNotableUpdateReasons::READ_RECEIPT);


Please add a test to ensure the notable update reasons is emitted as expected.

Added a test

stefanceriu · 2026-05-13T07:49:02Z

Thanks for looking into this 🙏
I left some comments inline but a couple more general notes:

the description is super long but the commits themselves have close to no info. Let the code and git history talk not the PR, it easier on everybody
please link a final client/meta issue if possible. For this kind of feature there most likely is one
there's formatting errors that need fixing, which is fine but there's a lot of checks on the CI and you probably want to start with a draft PR to avoid notifying everybody fixing them
the FFI layer is supposed to only be used for converting the data types and it shouldn't have business logic in it

danderson-cont · 2026-05-13T15:18:36Z

@stefanceriu @Hywan
Could you chime in on the "Open question for reviewers"

I can shorten it for you here:
The "mark as read" button in the UI takes too long to propagate it's effect for the iOS UI.

This can be fixed in iOS or on the SDK side (optimistic write). Which would you prefer?

stefanceriu · 2026-05-13T15:33:46Z

I can shorten it for you here: The "mark as read" button in the UI takes too long to propagate it's effect for the iOS UI.

This can be fixed in iOS or on the SDK side (optimistic write). Which would you prefer?

Why is that a concern for this PR? Do we even have UI anywhere that requires this to be fast? As a paralel manually marking rooms as unread/read seems fine even without "local echoes".

danderson-cont · 2026-05-13T20:20:49Z

I can shorten it for you here: The "mark as read" button in the UI takes too long to propagate it's effect for the iOS UI.
This can be fixed in iOS or on the SDK side (optimistic write). Which would you prefer?

Why is that a concern for this PR? Do we even have UI anywhere that requires this to be fast? As a paralel manually marking rooms as unread/read seems fine even without "local echoes".

Agreed, out of scope. Dropping it from the PR description.

stefanceriu · 2026-05-14T06:30:30Z

Looks good to me now but I'll let Ivan have the final say as he was the original reviewer.

poljar · 2026-05-14T14:41:22Z

+### Features
+
+- Add `RoomInfo::fully_read_event_id` to expose the user's `m.fully_read` event ID.
+  ([#6569](https://github.com/matrix-org/matrix-rust-sdk/pull/6569))
+


We switched to a towncrier-based setup for changelogs. Sorry for the inconvenience but please move this to the changelog.d folder as a Towncrier fragment.

@poljar Okay I've made these changes

poljar · 2026-05-14T14:41:41Z

+- [**breaking**] `RoomInfoNotableUpdateReasons` is now a `u16` to include a `FULLY_READ` flag to notify on changes of the `m.fully_read` marker.
+  ([#6569](https://github.com/matrix-org/matrix-rust-sdk/pull/6569))
+
+- Add `RoomInfo::fully_read_event_id` and `Room::fully_read_event_id` to expose the user's `m.fully_read` event ID.
+  ([#6569](https://github.com/matrix-org/matrix-rust-sdk/pull/6569))
+


Same for those changelog entries.

@poljar done

danderson-cont · 2026-05-15T22:58:22Z

Seems like the fixup thing is intended while in progress, and will go away when the commits are squashed. The rust tests failed on setup?

Hywan

Thanks! It looks much better.

Can you rebase all your fixup commits please?

Hywan · 2026-05-18T06:47:24Z

@@ -0,0 +1 @@
+- Add `RoomInfo::fully_read_event_id` and `Room::fully_read_event_id` to expose the user's `m.fully_read` event ID.


Suggested change

- Add `RoomInfo::fully_read_event_id` and `Room::fully_read_event_id` to expose the user's `m.fully_read` event ID.

Add `RoomInfo::fully_read_event_id` and `Room::fully_read_event_id` to expose the user's `m.fully_read` event ID.

The format defined in the CONTRIBUTING.md file was incorrect. I've fixed it, sowwy for the inconvenience.

Hywan · 2026-05-18T06:47:42Z

@@ -0,0 +1 @@
+- [**breaking**] `RoomInfoNotableUpdateReasons` is now a `u16` to include a `FULLY_READ` flag to notify on changes of the `m.fully_read` marker.


Suggested change

- [**breaking**] `RoomInfoNotableUpdateReasons` is now a `u16` to include a `FULLY_READ` flag to notify on changes of the `m.fully_read` marker.

[**breaking**] `RoomInfoNotableUpdateReasons` is now a `u16` to include a `FULLY_READ` flag to notify on changes of the `m.fully_read` marker.

Hywan · 2026-05-18T06:56:45Z

        /// Ultimately, we want to clearly identify all the notable update reasons, and
        /// remove this one.
-        const NONE = 0b1000_0000;
+        const NONE = 0b0000_0000_1000_0000;


Could be a good idea to use 0b1000_0000_0000_0000 for NONE, but not mandatory. Do it if you find the motivation, but not necessary here.

Hywan · 2026-05-18T06:58:03Z

+        );
+        assert!(room_info_notable_update_stream.is_empty());
+
+        // …Unless the event id changes!


Nitpicking:

Suggested change

// …Unless the event id changes!

// … Unless the event ID changes!

Add `RoomInfoNotableUpdateReasons::FULLY_READ`, used for when the `m.fully_read` marker for a room changes. Breaking change: the backing bitflag widens from `u8` to `u16` for the new flag to not alter the existing bit positions. Signed-off-by: Daniel Anderson <daniel.anderson@toptal.com>

Persist `m.fully_read` on `BaseRoomInfo` and expose it on `RoomInfo::fully_read_event_id` and `Room::fully_read_event_id`. The `m.fully_read` response processor now stores the event ID on the room info and emits `RoomInfoNotableUpdateReasons::FULLY_READ` Signed-off-by: Daniel Anderson <daniel.anderson@toptal.com>

`m.fully_read` is now available on the FFI `RoomInfo` record Signed-off-by: Daniel Anderson <daniel.anderson@toptal.com>

danderson-cont · 2026-05-19T17:38:11Z

Rebased the fixups into the three feature commits, dropped the leading - from the changelog fragments (including the FFI one), and fixed the test comment. Skipped the NONE bit-position move since you marked it optional.

stefanceriu

Alrighty, I think this one's good to go now, thank you for taking care of it! 👍

danderson-cont requested a review from a team as a code owner May 13, 2026 06:13

danderson-cont requested review from Hywan and removed request for a team May 13, 2026 06:13

danderson-cont mentioned this pull request May 13, 2026

Timeline: add a Jump to Unread button and a new-message badge on the scroll-to-bottom button. element-hq/element-x-ios#5506

Merged

6 tasks

stefanceriu reviewed May 13, 2026

View reviewed changes

Hywan requested changes May 13, 2026

View reviewed changes

danderson-cont marked this pull request as draft May 13, 2026 17:18

danderson-cont force-pushed the danderson-cont/expose-fully-read-event-id branch 2 times, most recently from 60d7404 to 4fcd2cd Compare May 13, 2026 20:14

danderson-cont requested review from Hywan and stefanceriu May 13, 2026 20:35

stefanceriu marked this pull request as ready for review May 14, 2026 06:29

poljar reviewed May 14, 2026

View reviewed changes

danderson-cont force-pushed the danderson-cont/expose-fully-read-event-id branch from 4fcd2cd to d82786c Compare May 14, 2026 16:31

danderson-cont requested a review from poljar May 14, 2026 16:48

Hywan requested changes May 18, 2026

View reviewed changes

danderson-cont added 3 commits May 19, 2026 11:27

feat(ffi): expose RoomInfo::fully_read_event_id

703da91

`m.fully_read` is now available on the FFI `RoomInfo` record Signed-off-by: Daniel Anderson <daniel.anderson@toptal.com>

danderson-cont force-pushed the danderson-cont/expose-fully-read-event-id branch from d82786c to 703da91 Compare May 19, 2026 17:32

stefanceriu approved these changes May 20, 2026

View reviewed changes

stefanceriu merged commit 10c0a86 into matrix-org:main May 20, 2026
60 of 61 checks passed

		@@ -0,0 +1 @@
		- Add `RoomInfo::fully_read_event_id` and `Room::fully_read_event_id` to expose the user's `m.fully_read` event ID.

	- Add `RoomInfo::fully_read_event_id` and `Room::fully_read_event_id` to expose the user's `m.fully_read` event ID.
	Add `RoomInfo::fully_read_event_id` and `Room::fully_read_event_id` to expose the user's `m.fully_read` event ID.

		@@ -0,0 +1 @@
		- [breaking] `RoomInfoNotableUpdateReasons` is now a `u16` to include a `FULLY_READ` flag to notify on changes of the `m.fully_read` marker.

	- [breaking] `RoomInfoNotableUpdateReasons` is now a `u16` to include a `FULLY_READ` flag to notify on changes of the `m.fully_read` marker.
	[breaking] `RoomInfoNotableUpdateReasons` is now a `u16` to include a `FULLY_READ` flag to notify on changes of the `m.fully_read` marker.

	// …Unless the event id changes!
	// … Unless the event ID changes!

Conversation

danderson-cont commented May 13, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

codecov Bot commented May 13, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

codspeed-hq Bot commented May 13, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Merging this PR will not alter performance

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Hywan left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

stefanceriu commented May 13, 2026

Uh oh!

danderson-cont commented May 13, 2026

Uh oh!

stefanceriu commented May 13, 2026

Uh oh!

danderson-cont commented May 13, 2026

Uh oh!

stefanceriu commented May 14, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

danderson-cont commented May 15, 2026

Uh oh!

Hywan left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

danderson-cont commented May 19, 2026

Uh oh!

stefanceriu left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

danderson-cont commented May 13, 2026 •

edited

Loading

codecov Bot commented May 13, 2026 •

edited

Loading

codspeed-hq Bot commented May 13, 2026 •

edited

Loading