Commit message (Collapse) | Author | Files | Lines | ||
---|---|---|---|---|---|
2016-03-02 | Set SYNAPSE_CACHE_FACTOR=1 in jenkins | Erik Johnston | 1 | -0/+1 | |
2016-03-02 | Add SYNAPSE_CACHE_FACTOR to README | Erik Johnston | 1 | -1/+18 | |
2016-03-01 | Add a /replication API for extracting the updates that happened on | Mark Haines | 15 | -4/+846 | |
synapse This is necessary for replicating the data in synapse to be visible to a separate service because presence and typing notifications aren't stored in a database so won't be visible to another process. This API can be used to either get the raw data by requesting the tables themselves or to just receive notifications for updates by following the streams meta-stream. Returns updates for each table requested a JSON array of arrays with a row for each row in the table. Each table is prefixed by a header row with the: name of the table, current stream_id position for the table, number of rows, number of columns and the names of the columns. This is followed by the rows that have been added to the server since the requester last asked. The API has a timeout and is hooked up to the notifier so that a slave can long poll for updates. | |||||
2016-03-01 | Load the current id in the IdGenerator constructor | Mark Haines | 12 | -77/+52 | |
Rather than loading them lazily. This allows us to remove all the yield statements and spurious arguments for the get_next methods. It also allows us to replace all instances of get_next_txn with get_next since get_next no longer needs to access the db. | |||||
2016-03-01 | Limit stream change cache size too | Erik Johnston | 1 | -1/+5 | |
2016-03-01 | Move cache size fiddling to descriptors only. Fix tests | Erik Johnston | 2 | -7/+9 | |
2016-03-01 | Add enviroment variable SYNAPSE_CACHE_FACTOR, default it to 0.1 | Erik Johnston | 3 | -2/+10 | |
2016-03-01 | Reraise exception | Erik Johnston | 1 | -0/+1 | |
2016-03-01 | Correct cache miss detection | Erik Johnston | 1 | -1/+1 | |
2016-02-26 | Add support for changing the actions for default rules | Mark Haines | 3 | -11/+102 | |
See matrix-org/matrix-doc#283 Works by adding dummy rules to the push rules table with a negative priority class and then using those rules to clobber the default rule actions when adding the default rules in ``list_with_base_rules`` | |||||
2016-02-26 | Fix to appease the PEP8 dragon | Gergely Polonkai | 1 | -1/+3 | |
2016-02-26 | Add error codes for malformed/bad JSON in /login | Gergely Polonkai | 1 | -2/+2 | |
Signed-off-by: Gergely Polonkai <gergely@polonkaieu> | |||||
2016-02-25 | Send the invier's member event in room invite state so the invitee has their ↵ | David Baker | 1 | -1/+7 | |
display name and avatar. | |||||
2016-02-25 | Make sure we return a JSON object when returning the values of specif… | Mark Haines | 1 | -1/+3 | |
…ic keys from a push rule | |||||
2016-02-24 | Remove unused get_rule_attr method | Mark Haines | 1 | -8/+0 | |
2016-02-24 | Generate guest access token on 3pid invites | Daniel Wagner-Hall | 4 | -0/+91 | |
This means that following the same link across multiple sessions or devices can re-use the same guest account. Note that this is somewhat of an abuse vector; we can't throw up captchas on this flow, so this is a way of registering ephemeral accounts for spam, whose sign-up we don't rate limit. | |||||
2016-02-24 | Ignore invalid POST bodies when joining rooms | Daniel Wagner-Hall | 1 | -2/+12 | |
2016-02-23 | Report size of ExpiringCache | Erik Johnston | 1 | -2/+13 | |
2016-02-23 | Create a new stream_id per presence update | Erik Johnston | 1 | -6/+9 | |
2016-02-23 | Check presence token interval is less than 100, rather than the token itself | Erik Johnston | 1 | -4/+4 | |
2016-02-23 | Set WORKSPACE if not set | Daniel Wagner-Hall | 1 | -0/+2 | |
2016-02-23 | Allow third_party_signed to be specified on /join | Daniel Wagner-Hall | 7 | -54/+196 | |
2016-02-23 | Handle get_all_entities_changed returning None | Erik Johnston | 1 | -0/+3 | |
2016-02-23 | Add comments | Erik Johnston | 1 | -0/+4 | |
2016-02-23 | Change the way we figure out presence updates for small deltas | Erik Johnston | 2 | -17/+52 | |
2016-02-23 | Don't include defer.returnValue in Measure block | Erik Johnston | 1 | -9/+12 | |
2016-02-23 | Measure PresenceEventSource.get_new_events | Erik Johnston | 1 | -34/+35 | |
2016-02-23 | handlers/sync: style fix, line too long | Patrik Oldsberg | 1 | -1/+5 | |
Signed-off-by: Patrik Oldsberg <patrik.oldsberg@ericsson.com> | |||||
2016-02-23 | Pick up currently_active across federation | Erik Johnston | 1 | -0/+1 | |
2016-02-23 | handlers/sync: fix SyncResult not counting account_data change when ↵ | Patrik Oldsberg | 1 | -1/+1 | |
converting to bool This fixes account_data events not triggering an immediate /sync response Signed-off-by: Patrik Oldsberg <patrik.oldsberg@ericsson.com> | |||||
2016-02-22 | Check that the disable_registration config key is handled correctly | Mark Haines | 1 | -0/+21 | |
2016-02-22 | address review comments | Richard van der Hoff | 1 | -7/+2 | |
drop commented-out special casing for historyvisibility event s/he/they/ for users | |||||
2016-02-22 | Add a test for TreeCache.__contains__ | Mark Haines | 2 | -0/+9 | |
2016-02-22 | Remove superfluous call to get_state_at when we already have an event for ↵ | Erik Johnston | 1 | -4/+12 | |
that stream position | |||||
2016-02-19 | Test Filter.filter_rooms | Mark Haines | 1 | -0/+18 | |
Also check that the __repr__ method for FilterCollection does something sensible. | |||||
2016-02-19 | Interpret unknown visibilities the same as shared | Richard van der Hoff | 1 | -11/+13 | |
2016-02-19 | Fix flake8 warnings for tests | Mark Haines | 39 | -364/+359 | |
2016-02-19 | Correctly filter states | Erik Johnston | 1 | -2/+2 | |
2016-02-19 | Only send presence updates to remote hosts if user is ours | Erik Johnston | 1 | -2/+10 | |
2016-02-19 | Add Measures to presence | Erik Johnston | 1 | -57/+62 | |
2016-02-19 | Allow guest users access to messages in rooms they have joined | Richard van der Hoff | 3 | -14/+42 | |
There should be no difference between guest users and non-guest users in terms of access to messages. Define the semantics of the is_peeking argument to filter_events_for_clients (slightly) better; interpret it appropriately, and set it correctly from /sync. | |||||
2016-02-19 | Add presence metric | Erik Johnston | 1 | -0/+3 | |
2016-02-19 | Remove spurious extra metrics | Erik Johnston | 1 | -1/+0 | |
2016-02-19 | Don't set currently_active for remote presence | Erik Johnston | 2 | -10/+27 | |
2016-02-19 | More metrics | Erik Johnston | 1 | -0/+6 | |
2016-02-19 | Add presence metrics | Erik Johnston | 2 | -0/+14 | |
2016-02-18 | "You are not..." | Erik Johnston | 1 | -1/+1 | |
2016-02-18 | Close cursor | Erik Johnston | 1 | -0/+1 | |
2016-02-18 | Remove unused param from get_max_token | Erik Johnston | 6 | -13/+11 | |
2016-02-18 | WheelTimer: Don't scan list, use index. | Erik Johnston | 1 | -5/+8 | |
2016-02-18 | Remove dead code for setting device specific rules. | Mark Haines | 10 | -141/+45 | |
It wasn't possible to hit the code from the API because of a typo in parsing the request path. Since no-one was using the feature we might as well remove the dead code. | |||||
2016-02-18 | Remove invalid arg. | Erik Johnston | 1 | -1/+0 | |
2016-02-18 | Add unit test | Erik Johnston | 2 | -93/+560 | |
2016-02-18 | Review comments | Daniel Wagner-Hall | 3 | -7/+7 | |
2016-02-18 | Remove old tests. | Erik Johnston | 6 | -2098/+0 | |
2016-02-18 | Add stuff pulled from the DB to the cache | Erik Johnston | 1 | -2/+4 | |
2016-02-18 | PEP8 | Erik Johnston | 1 | -2/+2 | |
2016-02-18 | Prefix TS fields with _ts | Erik Johnston | 4 | -43/+44 | |
2016-02-18 | Remove status_msg when going offline. Don't offline -> online if you send a ↵ | Erik Johnston | 1 | -5/+15 | |
message | |||||
2016-02-18 | Check presence state is a valid one | Erik Johnston | 2 | -1/+6 | |
2016-02-18 | Move if statement | Erik Johnston | 1 | -8/+8 | |
2016-02-18 | Comment | Erik Johnston | 1 | -0/+2 | |
2016-02-17 | Some cleanup | Daniel Wagner-Hall | 2 | -69/+61 | |
I'm not particularly happy with the "action" switching, but there's no convenient way to defer the work that needs to happen after it, so... :( | |||||
2016-02-17 | Remove spurious comment | Erik Johnston | 1 | -1/+0 | |
2016-02-17 | Initial cut | Erik Johnston | 16 | -1136/+933 | |
2016-02-17 | Respond to federated invite with non-empty context | Daniel Wagner-Hall | 3 | -46/+65 | |
Currently, we magically perform an extra database hit to find the inviter, and use this to guess where we should send the event. Instead, fill in a valid context, so that other callers relying on the context actually have one. | |||||
2016-02-17 | Add wheeltimer impl | Erik Johnston | 2 | -0/+162 | |
2016-02-17 | api/filtering: don't assume that event content will always be a dict | Patrik Oldsberg | 1 | -1/+4 | |
Signed-off-by: Patrik Oldsberg <patrik.oldsberg@ericsson.com> | |||||
2016-02-17 | client/v1/room: include event_id in response to state event PUT, in ↵ | Patrik Oldsberg | 1 | -2/+2 | |
accordance with the spec Signed-off-by: Patrik Oldsberg <patrik.oldsberg@ericsson.com> | |||||
2016-02-16 | Some cleanup, some TODOs, more to do | Daniel Wagner-Hall | 1 | -68/+60 | |
2016-02-16 | Fix typo in request validation for adding push rules. | Mark Haines | 1 | -1/+1 | |
2016-02-16 | Make adding push rules idempotent | Mark Haines | 1 | -82/+86 | |
Also remove the **kwargs from the add_push_rule method. Fixes https://matrix.org/jira/browse/SYN-391 | |||||
2016-02-16 | Some cleanup | Daniel Wagner-Hall | 2 | -23/+30 | |
2016-02-16 | Tidy? up room creation event sending | Daniel Wagner-Hall | 1 | -63/+72 | |
2016-02-16 | Add comment | Daniel Wagner-Hall | 1 | -1/+1 | |
2016-02-15 | Branch off member and non member sends | Daniel Wagner-Hall | 5 | -74/+99 | |
Unclean, needs tidy-up, but works | |||||
2016-02-15 | Simplify room creation code | Daniel Wagner-Hall | 2 | -52/+28 | |
2016-02-15 | Simplify room creation code | Daniel Wagner-Hall | 2 | -55/+31 | |
2016-02-15 | Use update_membership for profile updates | Daniel Wagner-Hall | 2 | -20/+12 | |
2016-02-15 | Use update_membership to kick guests | Daniel Wagner-Hall | 2 | -16/+19 | |
2016-02-15 | Reuse update_membership from /join | Daniel Wagner-Hall | 2 | -19/+14 | |
2016-02-15 | Inline _do_local_membership_update | Daniel Wagner-Hall | 1 | -18/+7 | |
2016-02-15 | Inline _do_join as it now only has one caller | Daniel Wagner-Hall | 1 | -48/+47 | |
Also, consistently apply rate limiting. Again, ugly, but a step in the right direction. | |||||
2016-02-15 | Merge implementation of /join by alias or ID | Daniel Wagner-Hall | 5 | -70/+71 | |
This code is kind of rough (passing the remote servers down a long chain), but is a step towards improvement. | |||||
2016-02-15 | Merge some room joining codepaths | Daniel Wagner-Hall | 2 | -5/+11 | |
Force joining by alias to go through the send_membership_event checks, rather than bypassing them straight into _do_join. This is the first of many stages of cleanup. | |||||
2016-02-12 | Revert "Merge two of the room join codepaths" | Daniel Wagner-Hall | 5 | -73/+69 | |
This reverts commit cf81375b94c4763766440471e632fc4b103450ab. It subtly violates a guest joining auth check | |||||
2016-02-12 | Merge two of the room join codepaths | Daniel Wagner-Hall | 5 | -69/+73 | |
There's at least one more to merge in. Side-effects: * Stop reporting None as displayname and avatar_url in some cases * Joining a room by alias populates guest-ness in join event * Remove unspec'd PUT version of /join/<room_id_or_alias> which has not been called on matrix.org according to logs * Stop recording access_token_id on /join/room_id - currently we don't record it on /join/room_alias; I can try to thread it through at some point. | |||||
2016-02-12 | Catch the exceptions thrown by twisted when you write to a closed connection | Mark Haines | 5 | -11/+33 | |
2016-02-12 | Mitigate against incorrect old state in /sync. | Erik Johnston | 1 | -5/+12 | |
There have been reports from the field that servers occasionally have incorrect notions of what the old state of a room is. This proves problematic as /sync relies on a correct old state. This patch makes /sync specifically include in the 'state' key any current state events that haven't been correctly included. | |||||
2016-02-11 | Enforce user_id exclusivity for AS registrations | Daniel Wagner-Hall | 1 | -5/+12 | |
This whole set of checks is kind of an ugly mess, which I may clean up at some point, but for now let's be correct. | |||||
2016-02-11 | Bump version and changelog | Erik Johnston | 2 | -1/+6 | |
2016-02-11 | Fix SYN-627, events are in incorrect room in /sync | Erik Johnston | 1 | -1/+1 | |
2016-02-11 | Bump version and changelog | Erik Johnston | 2 | -1/+7 | |
2016-02-11 | Return events in correct order for /events | Erik Johnston | 2 | -8/+12 | |
2016-02-11 | Add some paranoia logging | Erik Johnston | 1 | -0/+10 | |
2016-02-10 | 0.13.1 v0.13.1 | Matthew Hodgson | 1 | -1/+1 | |
2016-02-10 | bump for 0.13.1 | Matthew Hodgson | 1 | -0/+7 | |
2016-02-10 | try to bump syweb to 0.6.8 | Matthew Hodgson | 1 | -1/+1 | |
2016-02-10 | Update CHANGES | Erik Johnston | 1 | -2/+4 | |
2016-02-10 | Bump version and changelog | Erik Johnston | 2 | -1/+23 | |
2016-02-10 | Batch fetch _get_state_groups_from_groups | Erik Johnston | 1 | -32/+34 | |
2016-02-10 | Use _simple_select_many for _get_state_group_for_events | Erik Johnston | 2 | -17/+11 | |
2016-02-10 | Only calculate initial sync for 10 rooms at a time | Erik Johnston | 1 | -24/+29 | |
This helps to ensure we don't completely starve other requests. | |||||
2016-02-10 | Remove comment | Erik Johnston | 1 | -1/+1 | |
2016-02-10 | s/parent_context/previous_context/ | Erik Johnston | 1 | -9/+9 | |
2016-02-10 | Set parent context on instansiation | Erik Johnston | 1 | -2/+7 | |
2016-02-10 | Rename functions | Erik Johnston | 2 | -4/+4 | |
2016-02-09 | Typo | Erik Johnston | 1 | -1/+1 | |
2016-02-09 | Atomically persit push actions when we persist the event | Erik Johnston | 6 | -65/+49 | |
2016-02-09 | Pass in current state to push action handler | Erik Johnston | 3 | -27/+20 | |
2016-02-09 | Fix test | Erik Johnston | 1 | -9/+1 | |
2016-02-09 | Fix test | Erik Johnston | 1 | -3/+4 | |
2016-02-09 | Change a log from debug to info | Erik Johnston | 1 | -1/+1 | |
2016-02-09 | Remove dead code | Erik Johnston | 1 | -12/+0 | |
2016-02-09 | Invalidate get_last_receipt_event_id_for_user cache | Erik Johnston | 1 | -0/+5 | |
2016-02-09 | Don't load all ephemeral state for a room on every sync | Erik Johnston | 2 | -14/+20 | |
2016-02-09 | Fix unit tests | Erik Johnston | 2 | -5/+9 | |
2016-02-09 | Remove pyc files before running tests | Daniel Wagner-Hall | 1 | -1/+1 | |
2016-02-09 | Revert "Reject additional path segments" | Daniel Wagner-Hall | 3 | -6/+6 | |
This reverts commit 1d19a5ec0fff73af9cee8c21118020b31be47379. iOS Console is apparently relying on these paths. | |||||
2016-02-09 | Measure some /sync related things | Erik Johnston | 2 | -150/+162 | |
2016-02-09 | Revert to putting it around the entire block | Erik Johnston | 1 | -124/+124 | |
2016-02-09 | Don't create new logging context | Erik Johnston | 1 | -12/+15 | |
2016-02-09 | Don't bother copying records on parent context | Erik Johnston | 1 | -8/+0 | |
2016-02-09 | Fix bug in util.metrics.Measure | Erik Johnston | 1 | -1/+3 | |
2016-02-09 | Create new context when measuring | Erik Johnston | 1 | -2/+5 | |
2016-02-09 | Don't measure across event stream call, as it lasts for a long time. | Erik Johnston | 1 | -124/+124 | |
2016-02-09 | Change logcontext warns to debug | Erik Johnston | 1 | -3/+3 | |
2016-02-08 |