Initial commit

proposals/XXXX-disabled-presence-state.md

@@ -0,0 +1,52 @@
+# MSC0000: Disabled Presence State
+
+In current matrix we have no way to tell clients that we simply do not have presence data at all
+for a given homeserver or user. This proposal addresses this fact by adding a `disabled` state.
+
+By adding a `disabled` state to presence it allows us to kill two birds with one stone. Its used for
+MSC_PRESENCE_OVERRIDE_PLACEHOLDER and for this proposal. In this proposal its simply used to indicate
+a lack of information about a given user. This is going to 9 times out of 10 be because of that presence
+is disabled somewhere in the chain and therefore you cant get data. Be that disabled by the other user
+or their server or your server. If you are on matrix.org for example all presence will return this value
+if this proposal is adopted until they reenable presence. Since they have presence disabled.
+
+
+## Proposal
+
+This proposal proposes the `disabled` presence state. This makes it the 4th presence state together
+with `online`, `offline`, `unavailable`. 
+
+Due to `unavailable` being taken for another use `disabled` became the best candidate due to its dual use.
+
+`disabled` presence should be used if data is missing. Be that due to presence being disabled or because of other
+mechanisms this state was selected. For example due to the user choosing to put this as their state via
+MSC_PRESENCE_OVERRIDE_PLACEHOLDER or other mechanism like this. 
+
+
+## Potential issues
+
+The author can not see any significant potential issues arising from this change except that it can cause
+clients that are not architected to withstand protocol changes to break. 
+
+
+## Alternatives
+
+The alternatives for this proposal are well not that many if any? Like a way to indicate we don't have any data
+is useful. The only change I can think of is splitting intentional lack of data from we just don't have data yet.
+But I fail to find that distinction useful.
+
+
+## Security considerations
+
+Presence should not be security relevant as far as the author is aware. The only exception is the privacy
+discussion. This specific proposal should not have any impact on privacy because of the fact that this proposal
+does not it self change anything in practice. Presence being disabled is not a secret its a well known fact.
+
+## Unstable prefix
+
+Unstable implementations will use the state of `support.feline.mscXXXX.v1.disabled` in place of `disabled`.
+
+## Dependencies
+
+This MSC does not have any direct dependencies but is paired with MSC_PRESENCE_OVERRIDE_PLACEHOLDER due to
+this MSC being a semi dependency for it. These proposals can be adopted independently.