Commit message (Collapse) | Author | Files | Lines | ||
---|---|---|---|---|---|
2018-04-26 | Fix media admin APIs | Erik Johnston | 1 | -2/+2 | |
2018-04-25 | Also fix reindexing of search | Erik Johnston | 1 | -1/+1 | |
2018-04-25 | Fix quarantine media admin API | Erik Johnston | 1 | -1/+1 | |
2018-04-24 | Update CHANGES.rst | Neil Johnson | 1 | -2/+1 | |
fix formatting on line break | |||||
2018-04-24 | Bump version and update changelog v0.28.0-rc1 | Neil Johnson | 2 | -3/+44 | |
2018-04-24 | Revert "Bump version and update changelog" | Neil Johnson | 2 | -41/+3 | |
This reverts commit 08b29d45749d1944965e8d203155a4ea45a15423. | |||||
2018-04-24 | Bump version and update changelog | Neil Johnson | 2 | -3/+41 | |
2018-04-23 | Remove spurious param | Richard van der Hoff | 1 | -1/+0 | |
2018-04-20 | Document contrib directory | Erik Johnston | 2 | -0/+14 | |
2018-04-20 | accept stupid events over backfill | Richard van der Hoff | 1 | -4/+9 | |
2018-04-20 | Reinstate linearizer for federation_server.on_context_state_request | Richard van der Hoff | 1 | -5/+11 | |
2018-04-20 | Refactor store.have_events | Richard van der Hoff | 3 | -27/+55 | |
It turns out that most of the time we were calling have_events, we were only using half of the result. Replace have_events with have_seen_events and get_rejection_reasons, so that we can see what's going on a bit more clearly. | |||||
2018-04-18 | Add some comments | Richard van der Hoff | 2 | -5/+35 | |
2018-04-18 | Check events on backfill too | Richard van der Hoff | 1 | -15/+42 | |
2018-04-18 | Reject events which have too many auth_events or prev_events | Richard van der Hoff | 1 | -4/+28 | |
... this should protect us from being dossed by people making silly events (deliberately or otherwise) | |||||
2018-04-16 | Avoid creating events with huge numbers of prev_events | Richard van der Hoff | 4 | -54/+162 | |
In most cases, we limit the number of prev_events for a given event to 10 events. This fixes a particular code path which created events with huge numbers of prev_events. | |||||
2018-04-16 | Replace Queue with six.moves.queue | Adrian Tschira | 2 | -4/+6 | |
and a six.range change which I missed the last time Signed-off-by: Adrian Tschira <nota@notafile.com> | |||||
2018-04-16 | Make tests py3 compatible | Adrian Tschira | 5 | -5/+6 | |
This is a mixed commit that fixes various small issues * print parentheses * 01 is invalid syntax (it was octal in py2) * [x for i in 1, 2] is invalid syntax * six moves Signed-off-by: Adrian Tschira <nota@notafile.com> | |||||
2018-04-15 | fix spurious changelog dup | Matthew Hodgson | 1 | -1/+0 | |
2018-04-15 | add __bool__ alias to __nonzero__ methods | Adrian Tschira | 3 | -0/+9 | |
Signed-off-by: Adrian Tschira <nota@notafile.com> | |||||
2018-04-15 | Use six.itervalues in some places | Adrian Tschira | 4 | -11/+18 | |
There's more where that came from Signed-off-by: Adrian Tschira <nota@notafile.com> | |||||
2018-04-15 | Add some more variables to the unittest config | Adrian Tschira | 1 | -0/+4 | |
These worked accidentally before (python2 doesn't complain if you compare incompatible types) but under py3 this blows up spectacularly Signed-off-by: Adrian Tschira <nota@notafile.com> | |||||
2018-04-15 | Use str(e) instead of e.message | Adrian Tschira | 2 | -10/+10 | |
Doing this I learned e.message was pretty shortlived, added in 2.6, they realized it was a bad idea and deprecated it in 2.7 Signed-off-by: Adrian Tschira <nota@notafile.com> | |||||
2018-04-14 | Clarify that SRV may not point to a CNAME | Silke | 1 | -0/+3 | |
Signed-off-by: Silke Hofstra <silke@slxh.eu> | |||||
2018-04-13 | fix federation_domain_whitelist | Matthew Hodgson | 1 | -6/+6 | |
we were checking the wrong server_name on inbound requests | |||||
2018-04-13 | revert last to PR properly | Matthew Hodgson | 1 | -6/+6 | |
2018-04-13 | correctly auth inbound federation_domain_whitelist reqs | Matthew Hodgson | 1 | -6/+6 | |
2018-04-13 | Bump version and Change log v0.27.4 github/release-v0.27.0 release-v0.27.0 | Neil Johnson | 2 | -1/+9 | |
2018-04-13 | Revert "Use sortedcontainers instead of blist" | Richard van der Hoff | 3 | -10/+10 | |
This reverts commit 9fbe70a7dc3afabfdac176ba1f4be32dd44602aa. It turns out that sortedcontainers.SortedDict is not an exact match for blist.sorteddict; in particular, `popitem()` removes things from the opposite end of the dict. This is trivial to fix, but I want to add some unit tests, and potentially some more thought about it, before we do so. | |||||
2018-04-13 | Bump canonicaljson to 1.1.3 | Richard van der Hoff | 1 | -1/+1 | |
1.1.2 was a bit broken too :/ | |||||
2018-04-13 | ResponseCache: fix handling of completed results | Richard van der Hoff | 1 | -13/+19 | |
Turns out that ObservableDeferred.observe doesn't return a deferred if the result is already completed. Fix handling and improve documentation. | |||||
2018-04-12 | Update canonicaljson dependency | Richard van der Hoff | 1 | -1/+1 | |
1.1.0 and 1.1.1 were broken, so we're updating this to help people make sure they don't end up on a broken version. Also, 1.1.0 is speedier... | |||||
2018-04-12 | Refactor ResponseCache usage | Richard van der Hoff | 6 | -67/+87 | |
Adds a `.wrap` method to ResponseCache which wraps up the boilerplate of a (get, set) pair, and then use it throughout the codebase. This will be largely non-functional, but does include the following functional changes: * federation_server.on_context_state_request: drops use of _server_linearizer which looked redundant and could cause incorrect cache misses by yielding between the get and the set. * RoomListHandler.get_remote_public_room_list(): fixes logcontext leaks * the wrap function includes some logging. I'm hoping this won't be too noisy on production. | |||||
2018-04-12 | Format docstring | Erik Johnston | 1 | -2/+3 | |
2018-04-12 | Doc we raise on unknown event | Erik Johnston | 1 | -2/+3 | |
2018-04-12 | s/list/tuple | Erik Johnston | 1 | -2/+2 | |
2018-04-12 | Set all metrics at the same time | Erik Johnston | 2 | -10/+10 | |
2018-04-11 | Fix tests | Erik Johnston | 1 | -0/+1 | |
2018-04-11 | Track last processed event received_ts | Erik Johnston | 4 | -0/+52 | |
2018-04-11 | Track where event stream processing have gotten up to | Erik Johnston | 4 | -0/+24 | |
2018-04-11 | Add GaugeMetric | Erik Johnston | 2 | -1/+38 | |
2018-04-11 | bump version/changelog | Neil Johnson | 2 | -1/+9 | |
2018-04-11 | Send events to ASes concurrently | Erik Johnston | 1 | -2/+19 | |
2018-04-10 | Add metrics for ResponseCache | Richard van der Hoff | 6 | -7/+21 | |
2018-04-10 | fix typo | Neil Johnson | 1 | -2/+2 | |
2018-04-10 | Use run_in_background instead | Erik Johnston | 1 | -1/+1 | |
2018-04-10 | Note why we're limiting concurrent event sends | Erik Johnston | 1 | -0/+5 | |
2018-04-10 | Preserve log contexts correctly | Erik Johnston | 1 | -1/+4 | |
2018-04-10 | Log event ID on exception | Erik Johnston | 1 | -1/+4 | |
2018-04-10 | Ensure slashes are escaped | Erik Johnston | 1 | -1/+1 | |
2018-04-10 | URL quote path segments over federation | Erik Johnston | 1 | -48/+80 | |
2018-04-10 | Don't disable GC when running on PyPy | Vincent Breitmoser | 1 | -1/+7 | |
PyPy's incminimark GC can't be triggered manually. From what I observed there are no obvious issues with just letting it run normally. And unlike CPython, it actually returns unused RAM to the system. Signed-off-by: Vincent Breitmoser <look@my.amazin.horse> | |||||
2018-04-10 | In DomainSpecificString, override __repr__ in addition to __str__ | Vincent Breitmoser | 1 | -1/+1 | |
For some reason, string interpolation on a DomainSpecificString object like "%r" % (domainSpecificStringObj) fails under PyPy, because the default __repr__ implementation wants to iterate over the object. I'm not sure why that happens, but overriding __repr__ instead of __str__ fixes this problem, and is arguably the more appropriate thing to do anyways. | |||||
2018-04-10 | Fix pep8 error on psycopg2cffi hack | Richard van der Hoff | 1 | -6/+5 | |
2018-04-10 | Use psycopg2cffi module instead of psycopg2 if running on pypy | Vincent Breitmoser | 2 | -2/+8 | |
The psycopg2 package isn't available for PyPy. This commit adds a check if the runtime is PyPy, and if it is uses psycopg2cffi module in favor of psycopg2. This is almost a drop-in replacement, except for one place where an additional cast to string is required. | |||||
2018-04-10 | Use sortedcontainers instead of blist | Vincent Breitmoser | 3 | -10/+10 | |
This commit drop-in replaces blist with SortedContainers. They are written in pure python so work with pypy, but perform as good as native implementations, at least in a couple benchmarks: http://www.grantjenks.com/docs/sortedcontainers/performance.html | |||||
2018-04-09 | remove errant print v0.27.3-rc2 | Neil Johnson | 1 | -1/+1 | |
2018-04-09 | Fix msec to sec, again | Neil Johnson | 1 | -3/+3 | |
2018-04-09 | Fix msec to sec | Neil Johnson | 1 | -1/+1 | |
2018-04-09 | v0.27.3-rc2 | Neil Johnson | 1 | -0/+5 | |
2018-04-09 | Fix psycopg2 interpolation | Neil Johnson | 1 | -5/+5 | |
2018-04-09 | version bump v0.27.3-rc2 | Neil Johnson | 1 | -1/+1 | |
2018-04-09 | Handle all events in a room correctly | Erik Johnston | 1 | -1/+2 | |
2018-04-09 | Update CHANGES.rst | Neil Johnson | 1 | -1/+1 | |
2018-04-09 | bump version to release candidate | Neil Johnson | 1 | -1/+1 | |
2018-04-09 | Update CHANGES.rst | Neil Johnson | 1 | -1/+2 | |
2018-04-09 | Update CHANGES.rst | Neil Johnson | 1 | -3/+6 | |
2018-04-09 | bump version v0.27.3-rc1 | Neil Johnson | 1 | -1/+1 | |
2018-04-09 | 0.27.3 | Neil Johnson | 1 | -0/+46 | |
2018-04-09 | Revert "Merge pull request #3066 from matrix-org/rav/remove_redundant_metrics" | Richard van der Hoff | 5 | -28/+52 | |
We aren't ready to release this yet, so I'm reverting it for now. This reverts commit d1679a4ed7947b0814e0f2af9b888a16c588f1a1, reversing changes made to e089100c6231541c446e37e157dec8feed02d283. | |||||
2018-04-09 | Return a 404 rather than a 500 on rejoining empty rooms | Richard van der Hoff | 1 | -0/+8 | |
Filter ourselves out of the server list before checking for an empty remote host list, to fix 500 error Fixes #2141 | |||||
2018-04-09 | Limit concurrent event sends for a room | Erik Johnston | 1 | -49/+50 | |
2018-04-09 | Use create_and_send_nonmember_event everywhere | Erik Johnston | 1 | -6/+1 | |
2018-04-09 | Send federation events concurrently | Erik Johnston | 1 | -4/+18 | |
2018-04-09 | Handle exceptions in get_hosts_for_room when sending events over federation | Erik Johnston | 1 | -11/+16 | |
2018-04-07 | Replace some type checks with six type checks | Adrian Tschira | 5 | -8/+18 | |
Signed-off-by: Adrian Tschira <nota@notafile.com> | |||||
2018-04-06 | use python3-compatible prints | Adrian Tschira | 1 | -5/+5 | |
2018-04-06 | Replace old-style raise with six.reraise | Adrian Tschira | 1 | -1/+3 | |
The old style raise is invalid syntax in python3. As noted in the docs, this adds one more frame in the traceback, but I think this is acceptable: <ipython-input-7-bcc5cba3de3f> in <module>() 16 except: 17 pass ---> 18 six.reraise(*x) /usr/lib/python3.6/site-packages/six.py in reraise(tp, value, tb) 691 if value.__traceback__ is not tb: 692 raise value.with_traceback(tb) --> 693 raise value 694 finally: 695 value = None <ipython-input-7-bcc5cba3de3f> in <module>() 9 10 try: ---> 11 x() 12 except: 13 x = sys.exc_info() Also note that this uses six, which is not formally a dependency yet, but is included indirectly since most packages depend on it. Signed-off-by: Adrian Tschira <nota@notafile.com> | |||||
2018-04-06 | de-lint, quote consistency | Luke Barnard | 1 | -2/+2 | |
2018-04-06 | Explicitly grab individual columns from group object | Luke Barnard | 1 | -6/+8 | |
2018-04-06 | When exposing group state, return is_openly_joinable | Luke Barnard | 1 | -0/+5 | |
as opposed to join_policy, which is really only pertinent to the synapse implementation of the group server. By doing this we keep the group server concept extensible by allowing arbitrarily complex rules for deciding whether a group is openly joinable. | |||||
2018-04-06 | By default, join policy is "invite" | Luke Barnard | 1 | -1/+1 | |
2018-04-06 | add_user -> _add_user | Luke Barnard | 1 | -3/+3 | |
2018-04-06 | pep8 | Luke Barnard | 2 | -2/+4 | |
2018-04-06 | Get group_info from existing call to check_group_is_ours | Luke Barnard | 1 | -5/+1 | |
2018-04-06 | Don't use redundant inlineCallbacks | Luke Barnard | 1 | -4/+1 | |
2018-04-06 | join_rule -> join_policy | Luke Barnard | 1 | -1/+1 | |
2018-04-06 | is_joinable -> join_rule | Luke Barnard | 2 | -5/+2 | |
2018-04-06 | Fix federation client `set_group_joinable` typo | Luke Barnard | 1 | -1/+1 | |
2018-04-06 | Factor out add_user from accept_invite and join_group | Luke Barnard | 1 | -41/+29 | |
2018-04-06 | pep8 | David Baker | 1 | -1/+2 | |
2018-04-06 | Implement group join API | David Baker | 5 | -4/+124 | |
2018-04-06 | Port script: Set up state_group_id_seq | Richard van der Hoff | 1 | -0/+13 | |
Fixes https://github.com/matrix-org/synapse/issues/3050. | |||||
2018-04-06 | Port script: clean up a bit | Richard van der Hoff | 1 | -25/+29 | |
Improve logging and comments. Group all the stuff to do with inspecting tables together rather than creating the port tables in the middle. | |||||
2018-04-06 | Port script: avoid nasty errors when setting up | Richard van der Hoff | 1 | -8/+6 | |
We really shouldn't spit out "Failed to create port table", it looks scary. | |||||
2018-04-06 | Add response size metrics | Erik Johnston | 1 | -0/+7 | |
2018-04-06 | use PUT instead of POST for federating groups/m.join_policy | Krombel | 3 | -3/+7 | |
2018-04-06 | more verbosity in synctl | Richard van der Hoff | 1 | -0/+1 | |
2018-04-05 | make prometheus config compliant to v0.28 | Krombel | 3 | -19/+19 | |
2018-04-05 | Fix overzealous cache invalidation | Richard van der Hoff | 2 | -26/+84 | |
Fixes an issue where a cache invalidation would invalidate *all* pending entries, rather than just the entry that we intended to invalidate. | |||||
2018-04-05 | NON NULL -> NOT NULL | Luke Barnard | 1 | -1/+1 | |
2018-04-05 | Use "/settings/" (plural) | Luke Barnard | 3 | -3/+3 | |
2018-04-05 | Use DEFAULT join_policy of "invite" in db | Luke Barnard | 1 | -1/+1 | |
2018-04-05 | Document set_group_join_policy | Luke Barnard | 1 | -0/+6 | |
2018-04-05 | Review comments | Neil Johnson | 2 | -5/+6 | |
Use iteritems over item to loop over dict formatting | |||||
2018-04-04 | Remove redundant metrics which were deprecated in 0.27.0. | Richard van der Hoff | 5 | -52/+28 | |
2018-04-04 | phone home cache size configurations | Jan Christian Grünhage | 1 | -0/+3 | |
2018-04-04 | Document the additional routes for the event_creator worker | Travis Ralston | 1 | -1/+3 | |
Fixes https://github.com/matrix-org/synapse/issues/3018 Signed-off-by: Travis Ralston <travpc@gmail.com> | |||||
2018-04-04 | Move the mention of the main synapse worker higher up | Travis Ralston | 1 | -13/+6 | |
Signed-off-by: Travis Ralston <travpc@gmail.com> | |||||
2018-04-04 | Add b prefixes to some strings that are bytes in py3 | Adrian Tschira | 6 | -14/+14 | |
This has no effect on python2 Signed-off-by: Adrian Tschira <nota@notafile.com> | |||||
2018-04-04 | Revert "improve mxid check performance" | Richard van der Hoff | 1 | -4/+3 | |
2018-04-04 | Remove address resolution of hosts in SRV records | Silke | 2 | -122/+10 | |
Signed-off-by: Silke Hofstra <silke@slxh.eu> | |||||
2018-04-04 | Document the behaviour of ResponseCache | Richard van der Hoff | 1 | -0/+32 | |
it looks like everything that uses ResponseCache expects to have to `make_deferred_yieldable` its results. It's debatable whether that is the best approach, but let's document it for now to avoid further confusion. | |||||
2018-04-03 | Use join_policy API instead of joinable | Luke Barnard | 7 | -23/+58 | |
The API is now under /groups/$group_id/setting/m.join_policy and expects a JSON blob of the shape ```json { "m.join_policy": { "type": "invite" } } ``` where "invite" could alternatively be "open". | |||||
2018-04-03 | Fix json encoding bug in replication | Richard van der Hoff | 1 | -1/+1 | |
json encoders have an encode method, not a dumps method. | |||||
2018-03-31 | improve mxid check performance ~4x | Adrian Tschira | 1 | -3/+4 | |
Signed-off-by: Adrian Tschira <nota@notafile.com> | |||||
2018-03-29 | Remove uses of events.content | Richard van der Hoff | 3 | -7/+12 | |
2018-03-29 | Use static JSONEncoders | Richard van der Hoff | 4 | -20/+34 | |
using json.dumps with custom options requires us to create a new JSONEncoder on each call. It's more efficient to create one upfront and reuse it. | |||||
2018-03-29 | Use simplejson throughout | Richard van der Hoff | 3 | -5/+9 | |
Let's use simplejson rather than json, for consistency. | |||||
2018-03-29 | pep8 | Neil Johnson | 1 | -2/+4 | |
2018-03-29 | Remove need for sqlite specific query | Neil Johnson | 1 | -30/+57 | |
2018-03-28 | fix pep8 errors | Neil Johnson | 1 | -3/+0 | |
2018-03-28 | Update README.rst | Neil Johnson | 1 | -2/+2 | |
update docker hub url | |||||
2018-03-28 | This should probably be a PUT | David Baker | 1 | -1/+1 | |
2018-03-28 | remove twisted deferral cruft | Neil Johnson | 1 | -6/+3 | |