Skip to content
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

Spec event edits #1211

Merged
merged 6 commits into from Sep 21, 2022
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
1 change: 1 addition & 0 deletions changelogs/client_server/newsfragments/1211.feature
@@ -0,0 +1 @@
Add `m.replace` relations (event edits), as per [MSC2676](https://github.com/matrix-org/matrix-spec-proposals/pull/2676).
12 changes: 2 additions & 10 deletions content/client-server-api/_index.md
Expand Up @@ -1955,16 +1955,6 @@ rooms, or the relationship missing properties required by the schema below. Clie
handling such invalid relationships should show the events independently of each
other, optionally with an error message.

{{% boxes/note %}}
While this specification describes an `m.relates_to` object containing a `rel_type`, there
is not currently any relationship type which uses this structure. Replies, described below,
form their relationship outside of the `rel_type` as a legacy type of relationship. Future
versions of the specification might change replies to better match the relationship structures.

Custom `rel_type`s can, and should, still use the schema described above for relevant
behaviour.
{{% /boxes/note %}}

`m.relates_to` is defined as follows:

{{% definition path="api/client-server/definitions/m.relates_to" %}}
Expand All @@ -1974,6 +1964,7 @@ behaviour.
This specification describes the following relationship types:

* [Rich replies](#rich-replies) (**Note**: does not use `rel_type`).
* [Event replacements](#event-replacements).

#### Aggregations

Expand Down Expand Up @@ -2643,3 +2634,4 @@ systems.
{{< cs-module name="server_notices" >}}
{{< cs-module name="moderation_policies" >}}
{{< cs-module name="spaces" >}}
{{< cs-module name="event_replacements" >}}
330 changes: 330 additions & 0 deletions content/client-server-api/modules/event_replacements.md
@@ -0,0 +1,330 @@
---
type: module
---

### Event replacements

{{% added-in v="1.4" %}}

Event replacements, or "message edit events", are events that use an [event
relationship](#forming-relationships-between-events)
with a `rel_type` of `m.replace`, which indicates that the original event is
intended to be replaced.

An example of a message edit event might look like this:

```json
turt2live marked this conversation as resolved.
Show resolved Hide resolved
{
"type": "m.room.message",
"content": {
"body": "* Hello! My name is bar",
"msgtype": "m.text",
"m.new_content": {
"body": "Hello! My name is bar",
"msgtype": "m.text"
},
"m.relates_to": {
"rel_type": "m.replace",
"event_id": "$some_event_id"
}
},
// ... other fields required by events
}
```

The `content` of the replacement must contain a `m.new_content` property which
defines the replacement `content`. The normal `content` properties (`body`,
`msgtype` etc.) provide a fallback for clients which do not understand
replacement events.

`m.new_content` can include any properties that would normally be found in
an event's content property, such as `formatted_body` (see [`m.room.message`
`msgtypes`](#mroommessage-msgtypes)).

#### Validity of replacement events

There are a number of requirements on replacement events, which must be satisfied for the replacement to be considered valid:

* As with all event relationships, the original event and replacement event
must have the same `room_id` (i.e. you cannot send an event in
one room and then an edited version in a different room).

* The original event and replacement event must have the same `sender`
(i.e. you cannot edit someone else's messages).

* The replacement and original events must have the same `type` (i.e. you
cannot change the original event's type).

* The replacement and original events must not have a `state_key` property
(i.e. you cannot edit state events at all).

* The original event must not, itself, have a `rel_type` of `m.replace`
(i.e. you cannot edit an edit — though you can send multiple edits for a
single original event).

* The replacement event (once decrypted, if appropriate) must have an
`m.new_content` property.

If any of these criteria are not satisfied, implementations should ignore the
replacement event (the content of the original should not be replaced, and the
edit should not be included in the server-side aggregation).

Note that the [`msgtype`](#mroommessage-msgtypes) property of replacement
`m.room.message` events does *not* need to be the same as in the original event. For
example, it is legitimate to replace an `m.text` event with an `m.emote`.

#### Editing encrypted events

If the original event was [encrypted](#end-to-end-encryption), the replacement
should be too. In that case, `m.new_content` is placed in the content of the
encrypted payload. As with all event relationships, the `m.relates_to` property
must be sent in the unencrypted (cleartext) part of the event.

For example, a replacement for an encrypted event might look like this:

```json
turt2live marked this conversation as resolved.
Show resolved Hide resolved
{
"type": "m.room.encrypted",
"content": {
"m.relates_to": {
"rel_type": "m.replace",
"event_id": "$some_event_id"
},
"algorithm": "m.megolm.v1.aes-sha2",
"sender_key": "<sender_curve25519_key>",
"device_id": "<sender_device_id>",
"session_id": "<outbound_group_session_id>",
"ciphertext": "<encrypted_payload_base_64>"
}
// irrelevant fields not shown
}
```

... and, once decrypted, the payload might look like this:

```json
{
"type": "m.room.<event_type>",
"room_id": "!some_room_id",
"content": {
"body": "* Hello! My name is bar",
"msgtype": "m.text",
"m.new_content": {
"body": "Hello! My name is bar",
"msgtype": "m.text"
}
}
}
```

Note that:

* There is no `m.relates_to` property in the encrypted payload. If there was, it would be ignored.
* There is no `m.new_content` property in the cleartext content of the `m.room.encrypted` event. As above, if there was then it would be ignored.

{{% boxes/note %}}
The payload of an encrypted replacement event must be encrypted as normal, including
ratcheting any [Megolm](#mmegolmv1aes-sha2) session as normal. The original Megolm
ratchet entry should **not** be re-used.
{{% /boxes/note %}}
Comment on lines +125 to +129
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm afraid this still doesn't make any sense to me as a reader - we have zero mention of a "ratchet entry" in the spec, and it's not a term I've heard come up in conversation before. I realize this is coming from the MSC, but it also doesn't make sense there having re-read it.

Is this the "ratchet index" (a defined term in the spec) or can we just say it's "encrypted like any other event" and avoid the problem entirely?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Right, it should be "ratchet value", from https://gitlab.matrix.org/matrix-org/olm/blob/master/docs/megolm.md#the-megolm-ratchet-algorithm.

My general feeling is: if you're implementing edits of encrypted events, you'll know what this means, and if you're not, you don't need to worry about it.

can we just say it's "encrypted like any other event"

Well, at that point, it doesn't seem to give any information at all, so we might as well just omit it.

TBH it seems kindof obvious to me that we wouldn't reuse the old key, so I'd be happy to omit this block, but this text was added in response to @erikjohnston's question at matrix-org/matrix-spec-proposals#2676 (comment). Erik: do you have any thoughts on whether this text is necessary?

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

due to lack of comment I'm inclined to believe it's just a safety net thing. It doesn't feel particularly safe to assume that the person implementing edits of encrypted events is also aware of the intricacies of encryption, however I'm happy to go with whatever at this point.



#### Applying `m.new_content`

When applying a replacement, the `content` of the original event is treated as
being overwritten entirely by `m.new_content`, with the exception of `m.relates_to`,
which is left *unchanged*. Any `m.relates_to` property within `m.new_content`
is ignored.

{{% boxes/note %}}
Note that server implementations must not *actually* overwrite
the original event's `content`: instead the server presents it as being overwritten
when it is served over the client-server API. See [Server-side replacement of content](#server-side-replacement-of-content)
below.
{{% /boxes/note %}}

For example, given a pair of events:

```json
{
"event_id": "$original_event",
"type": "m.room.message",
"content": {
"body": "I really like cake",
"msgtype": "m.text",
"formatted_body": "I really like cake",
}
}
```

```json
{
"event_id": "$edit_event",
"type": "m.room.message",
"content": {
"body": "* I really like *chocolate* cake",
"msgtype": "m.text",
"m.new_content": {
"body": "I really like *chocolate* cake",
"msgtype": "m.text",
"com.example.extension_property": "chocolate"
},
"m.relates_to": {
"rel_type": "m.replace",
"event_id": "$original_event_id"
}
}
}
```

... then the end result is an event as shown below:
turt2live marked this conversation as resolved.
Show resolved Hide resolved

```json
{
"event_id": "$original_event",
"type": "m.room.message",
"content": {
"body": "I really like *chocolate* cake",
"msgtype": "m.text",
"com.example.extension_property": "chocolate"
}
}
```

Note that `formatted_body` is now absent, because it was absent in the
replacement event.

#### Server behaviour

##### Server-side aggregation of `m.replace` relationships

Note that there can be multiple events with an `m.replace` relationship to a
given event (for example, if an event is edited multiple times). These should
be [aggregated](#aggregations) by the homeserver.

The aggregation format of `m.replace` relationships gives the `event_id`,
`origin_server_ts`, and `sender` of the **most recent** replacement event. The
most recent event is determined by comparing `origin_server_ts`; if two or more
replacement events have identical `origin_server_ts`, the event with the
lexicographically largest `event_id` is treated as more recent.

This aggregation is bundled under the `unsigned` property as `m.relations` for any
event that is the target of an `m.replace` relationship. For example:

```json
turt2live marked this conversation as resolved.
Show resolved Hide resolved
{
"event_id": "$original_event_id",
// irrelevant fields not shown
"unsigned": {
"m.relations": {
"m.replace": {
"event_id": "$latest_edit_event_id",
"origin_server_ts": 1649772304313,
"sender": "@editing_user:localhost"
}
}
}
}
```

If the original event is
[redacted](#redactions), any
`m.replace` relationship should **not** be bundled with it (whether or not any
subsequent replacements are themselves redacted). Note that this behaviour is
specific to the `m.replace` relationship. See also [redactions of edited
events](#redactions-of-edited-events) below.

##### Server-side replacement of content

Whenever an `m.replace` is to be bundled with an event as above, the server
should also modify the content of the original event according to the
`m.new_content` of the most recent replacement event (determined as above).

An exception applies to [`GET /_matrix/client/v3/rooms/{roomId}/event/{eventId}`](#get_matrixclientv3roomsroomideventeventid),
which should return the unmodified event (though the relationship should still
be bundled, as described above).

#### Client behaviour

Clients can often ignore `m.replace` events, because any events returned
by the server to the client will be updated by the server to account for
subsequent edits.

However, clients should apply the replacement themselves when the server is
unable to do so. This happens in the following situations:

* The client has already received and stored the original event before the
message edit event arrives.

* The original event (and hence its replacement) are encrypted.

Client authors are reminded to take note of the requirements for [Validity of
message edit events](#validity-of-message-edit-events), and to ignore any
invalid edit events that are received.

##### Permalinks

When creating [links](/appendices/#uris) to events (also known as permalinks),
clients build links which reference the event that the creator of the permalink
is viewing at that point (which might be a message edit event).

The client viewing the permalink should resolve this reference to the original
event, and then display the most recent version of that event.

#### Redactions of edited events

When an event using a `rel_type` of `m.replace` is [redacted](#redactions), it
removes that edit revision. This has little effect if there were subsequent
edits. However, if it was the most recent edit, the event is in effect
reverted to its content before the redacted edit.

Redacting the *original* message in effect removes the message, including all
subsequent edits, from the visible timeline. In this situation, homeservers
will return an empty `content` for the original event as with any other
redacted event, and as
[above](#server-side-aggregation-of-mreplace-relationships) the replacement
events will not be bundled with the original event. Note that the subsequent edits are
not actually redacted themselves: they simply serve no purpose within the visible timeline.

#### Edits of replies

Some particular constraints apply to events which replace a
[reply](#rich-replies). In particular:

* In contrast to the original reply, there should be no `m.in_reply_to`
property in the the `m.relates_to` object, since it would be redundant (see
[Applying `m.new_content`](#applying-mnew_content) above, which notes that
the original event's `m.relates_to` is preserved), as well as being contrary
to the spirit of the event relationships mechanism which expects only one
"parent" per event.

* `m.new_content` should **not** contain any [reply
fallback](#fallbacks-for-rich-replies),
since it is assumed that any client which can handle edits can also display
replies natively. However, the `content` of the replacement event should provide
fallback content for clients which support neither rich replies nor edits.

An example of an edit to a reply is as follows:

```json
turt2live marked this conversation as resolved.
Show resolved Hide resolved
{
"type": "m.room.message",
richvdh marked this conversation as resolved.
Show resolved Hide resolved
// irrelevant fields not shown
"content": {
"body": "> <@alice:example.org> question\n\n* reply",
"msgtype": "m.text",
"format": "org.matrix.custom.html",
"formatted_body": "<mx-reply><blockquote><a href=\"https://matrix.to/#/!somewhere:example.org/$event:example.org\">In reply to</a> <a href=\"https://matrix.to/#/@alice:example.org\">@alice:example.org</a><br />question</blockquote></mx-reply>* reply",
"m.new_content": {
"body": "reply",
"msgtype": "m.text",
"format": "org.matrix.custom.html",
"formatted_body": "reply"
},
"m.relates_to": {
"rel_type": "m.replace",
"event_id": "$original_reply_event"
}
}
}
```