- Oct 24, 2017
-
-
Erik Johnston authored
-
Erik Johnston authored
Correctly wire in update group profile over federation
-
Erik Johnston authored
- Oct 23, 2017
-
-
Richard van der Hoff authored
PEP8 fixes
-
Richard van der Hoff authored
E741 says "do not use variables named ‘l’, ‘O’, or ‘I’".
-
Richard van der Hoff authored
what could possibly go wrong
-
Erik Johnston authored
-
Richard van der Hoff authored
Fix a logcontext leak in the media repo
-
Richard van der Hoff authored
Updates to ID checking
-
Richard van der Hoff authored
-
Erik Johnston authored
-
Erik Johnston authored
-
Richard van der Hoff authored
Remove pointless create() method
-
- Oct 20, 2017
-
-
Richard van der Hoff authored
May as well do it whenever we parse a Group ID. We check the sigil and basic structure here so it makes sense to check the grammar in the same place.
-
Richard van der Hoff authored
... because the spec says we should.
-
Richard van der Hoff authored
Factor out a common function for checking user ids and group ids, which forbids capitals.
-
Richard van der Hoff authored
It just calls the constructor, so we may as well kill it rather than having random codepaths.
-
- Oct 19, 2017
-
-
Erik Johnston authored
-
Erik Johnston authored
-
Erik Johnston authored
Add config to enable group creation
-
Erik Johnston authored
-
Erik Johnston authored
-
Erik Johnston authored
-
Erik Johnston authored
Enforce sensible group IDs
-
Erik Johnston authored
-
Erik Johnston authored
-
Erik Johnston authored
-
Erik Johnston authored
Fix typo in thumbnail generation
-
Erik Johnston authored
-
Erik Johnston authored
- Oct 18, 2017
-
-
Richard van der Hoff authored
Fix 500 error when we get an error handling a PDU
-
- Oct 17, 2017
-
-
Richard van der Hoff authored
Fix 500 error when fields missing from power_levels event
-
Richard van der Hoff authored
FederationServer doesn't have a send_failure (and nor does its subclass, ReplicationLayer), so this was failing. I'm not really sure what the idea behind send_failure is, given (a) we don't do anything at the other end with it except log it, and (b) we also send back the failure via the transaction response. I suspect there's a whole lot of dead code around it, but for now I'm just removing the broken bit.
-
Richard van der Hoff authored
-
Richard van der Hoff authored
If the users or events keys were missing from a power_levels event, then we would throw 500s when trying to auth them.
-
Erik Johnston authored
fix thumbnailing (#2548)
-
Richard van der Hoff authored
`os.path.exists` doesn't allow us to distinguish between permissions errors and the path actually not existing, which repeatedly confuses people. It also means that we try to overwrite existing key files, which is super-confusing. (cf issues #2455, #2379). Use os.stat instead. Also, don't recomemnd the the use of --generate-config, which screws everything up if you're using debian (cf #2455).
-