3
0
mirror of https://github.com/ergochat/ergo.git synced 2024-11-10 22:19:31 +01:00
Commit Graph

1321 Commits

Author SHA1 Message Date
Daniel Oaks
fbac8c9a0c Updated changelog 2018-01-22 17:44:33 +10:00
Daniel Oaks
cd6bf3e20c Don't load new language support by default, since we don't have any/enough language files to justify it 2018-01-22 17:42:41 +10:00
Daniel Oaks
e99f22488f Make LANGUAGE support work 2018-01-22 17:30:31 +10:00
Daniel Oaks
a7fdade41d Extend to include example translation stuff 2018-01-21 16:49:17 +10:00
Daniel Oaks
ba77a95c81 Very initial beginning of languages implementation 2018-01-21 16:11:16 +10:00
Daniel Oaks
cf53d27615 travis: Install goreleaser 2018-01-07 18:43:33 +10:00
Daniel Oaks
1c4f316ca9 Remove make clean for now, fixes Travis 2018-01-07 18:35:57 +10:00
Daniel Oaks
2b383145e5 Fix makefile / Travis 2018-01-07 13:20:05 +10:00
Daniel Oaks
3b3c1c733a Support bot mode +B 2018-01-07 12:56:51 +10:00
Daniel Oaks
7f06f25c15 Start making builds with goreleaser 2018-01-07 12:40:26 +10:00
Daniel Oaks
e38d42c1ce Update changelog 2018-01-04 00:28:06 +10:00
Daniel Oaks
3ba8af714e Use base 36 when generating message IDs, gives us full 0-9a-z to use while preserving uniqueness nicely 2018-01-04 00:21:35 +10:00
Daniel Oaks
63bd52d471 Make message IDs much shorter (thanks @prawnsalad!) 2018-01-04 00:13:32 +10:00
Daniel Oaks
713fc59691 CAP: Correctly suspend registration for CAP REQ as per the spec 2017-12-28 13:50:42 +10:00
Daniel Oaks
342a84ba3f Makefile should use var defined at top of file 2017-12-28 13:42:09 +10:00
Daniel Oaks
7c624f9a38 Setup v0.10.4-unreleased devel ver 2017-12-28 13:41:36 +10:00
Daniel Oaks
25c59fdb37 vendor: Updated submodules 2017-12-26 12:57:39 +10:00
Daniel Oaks
5e9f0fd411 Release v0.10.3 2017-12-26 12:43:37 +10:00
Daniel Oaks
b9f7850719 Update changelog 2017-12-26 12:38:54 +10:00
Daniel Oaks
197ab520bf Change casemapping name and advertisement to match new spec 2017-12-26 12:30:04 +10:00
Daniel Oaks
36b26f99be
Merge pull request #173 from slingamn/timeouts.3
refactor idle timeouts again
2017-12-11 14:53:05 +10:00
Shivaram Lingamneni
d8d1705035 refactor idle timeouts again 2017-12-06 23:23:25 -05:00
Daniel Oaks
e0519a6387 caps: Rename maxline cap 2017-12-06 13:11:38 +10:00
Daniel Oaks
b2e5738f08
Merge pull request #171 from slingamn/atime
fix a race condition in client timeouts
2017-12-03 12:59:08 +10:00
Shivaram Lingamneni
ad1e00629b fix a race condition in idle timeouts
squigz on freenode reported an issue where bots were responding to PING
on time, but were occasionally being timed out regardless. This was a race
condition: timeout was detected as idleTime >= it.quitTimeout, but if
the client responded promptly to its PING message and sent no further messages,
but the main loop subsequently slept for longer than expected (i.e., significantly
longer than quitTimeout), this condition would be met through no fault of the
client's.

The fix here is to explicitly track the last time the ping was sent, then test
!lastSeen.After(lastPinged) instead (making use of time.Time's monotonicity).
It is sufficient that the measurement of lastPinged happens-before the PING is sent.
2017-12-02 20:30:26 -05:00
Shivaram Lingamneni
f5e7ec525d synchronize accesses to client.atime 2017-12-02 20:05:06 -05:00
Daniel Oaks
ad351268be
Merge pull request #168 from slingamn/clientmanager
refactor ClientManager
2017-11-23 08:35:50 +10:00
Shivaram Lingamneni
b4907dadb9 fix a bug where the uncasefolded nickname wasn't being recorded
Also, rename the nickname vars to hopefully make things clearer
2017-11-22 16:56:14 -05:00
Shivaram Lingamneni
d5a5f939dd review fixes 2017-11-22 16:35:57 -05:00
Shivaram Lingamneni
52b0fb71e7 refactor ClientManager 2017-11-22 05:09:38 -05:00
Daniel Oaks
2cbbec567c Don't override package name 2017-11-19 10:42:40 +10:00
Daniel Oaks
bfcada11dc KLINE: Save name of the banning oper 2017-11-19 10:32:32 +10:00
Daniel Oaks
dedb029272 DLINE: Save name of the banning oper 2017-11-19 10:27:40 +10:00
Daniel Oaks
8036df92fc WHO: Require first param, matching other servers 2017-11-19 10:10:10 +10:00
Daniel Oaks
bec050208d
Merge pull request #166 from slingamn/banlist.3
allow non-operators to list bans
2017-11-17 16:09:21 +10:00
Shivaram Lingamneni
b83d454554 fix ban regexp not being recompiled on removal 2017-11-13 18:25:51 -05:00
Shivaram Lingamneni
b1e12ca1eb allow unprivileged users to list bans 2017-11-13 18:25:51 -05:00
Daniel Oaks
fc8007e6e4 Setup v0.10.3-unreleased devel ver 2017-11-13 22:52:48 +10:00
Daniel Oaks
11ea52b0f0 Release v0.10.2 2017-11-13 22:46:50 +10:00
Daniel Oaks
a33e83de9e
Merge pull request #165 from slingamn/get_fail
fix a wide-ranging assortment of crash bugs
2017-11-13 22:43:24 +10:00
Shivaram Lingamneni
ca8a0b7091 fix a wide-ranging assortment of crash bugs 2017-11-13 02:42:20 -05:00
Daniel Oaks
ff50a25b65
Merge pull request #161 from slingamn/chanreg.3
remove registeredChannelsMutex
2017-11-13 12:46:23 +10:00
Daniel Oaks
075a38d08d developing: Add new release instructions. 2017-11-13 12:45:28 +10:00
Daniel Oaks
efb54163a5 Setup v0.10.2-unreleased devel ver 2017-11-13 12:45:10 +10:00
Daniel Oaks
f2d4698baf Release v0.10.1 2017-11-09 15:07:40 +10:00
Daniel Oaks
44538c3b0b
Merge pull request #162 from slingamn/modeperms
fix channel mode change privilege enforcement
2017-11-09 14:56:53 +10:00
Shivaram Lingamneni
6d619bf411 fix channel mode change privilege enforcement 2017-11-08 23:37:38 -05:00
Shivaram Lingamneni
d4cb15354f remove registeredChannelsMutex
This moves channel registration to an eventual consistency model,
where the in-memory datastructures (Channel and ChannelManager)
are the exclusive source of truth, and updates to them get persisted
asynchronously to the DB.
2017-11-08 22:24:33 -05:00
Shivaram Lingamneni
d5832bf765
Merge pull request #160 from slingamn/cache.4
fix a race in regenerateMembersCache
2017-11-08 22:18:25 -05:00
Shivaram Lingamneni
60b861e07e fix a race in regenerateMembersCache
The rationale for why regenerateMembersCache didn't need to hold the Lock()
throughout was subtly wrong. It is true that at least some attempt to
regenerate the cache would see *all* the updates. However, it was possible for
the value of `result` generated by that attempt to lose the race for the final
assignment `channel.membersCache = result`.

The fix is to serialize the attempts to regenerate the cache, without adding
any additional locking on the underlying `Channel` fields via
`Channel.stateMutex`. This ensures that the final read from `Channel.members`
is paired with the final write to `Channel.membersCache`.
2017-11-07 14:38:18 -05:00