3dafee6aa6
NetworkSession: explicitly unregister effect manager hooks on dispose,
...
close #3455
2020-05-14 10:59:10 +01:00
36c5d9117d
LoginPacketHandler: properly handle failure to base64_decode stuff from JWT
...
previously this might just return false and blow up in your face.
I considered fixing this on stable too, but it's less useful there because so much stuff on stable just explodes at the first wrong thing anyway.
2020-05-14 09:46:44 +01:00
955b23cc8e
Merge branch 'stable'
...
# Conflicts:
# resources/vanilla
2020-05-14 00:23:36 +01:00
e9ecb9a9fd
fixed build failure
2020-05-13 22:54:24 +01:00
84291e7980
Throw a more specific exception for JWT handling errors
2020-05-13 21:23:04 +01:00
486e0e710b
JwtUtils: fix minor formatting issue
2020-05-13 20:51:24 +01:00
44a7829536
VerifyLoginException: fixed rogue usage in JwtUtils
2020-05-13 20:50:41 +01:00
d7eb4f9651
Extract more general-purpose logic to JwtUtils
...
this code could now be reused for creating custom login packets.
2020-05-13 20:46:01 +01:00
161ac468f3
ProcessLoginTask: properly cater for key parsing errors
2020-05-13 19:23:12 +01:00
934f85841c
asserts 👏 are 👏 not 👏 error 👏 checking
2020-05-13 13:37:49 +01:00
8c2878fe5b
Added JwtUtils::parse(), make ProcessLoginTask more robust
2020-05-13 13:36:42 +01:00
144a66c110
cumulative imports cleanup (again)
2020-05-12 22:38:59 +01:00
6b037d6a4c
RuntimeBlockMapping: these IDs are not static any more
2020-05-11 12:40:55 +01:00
b7cf4f01f9
remove utils\UUID, switch to pocketmine/uuid package
2020-05-11 10:46:48 +01:00
a6c35cab9a
JwtUtils: fix phpstan failure
2020-05-06 22:00:11 +01:00
0b6d6306cf
LoginPacket: fixed error handling edge case with malformed chain data JSON
2020-05-06 21:46:35 +01:00
ed757c7207
consolidate some JWT handling into one class
2020-05-06 21:32:22 +01:00
5d154e43a9
LoginPacket: removed an old hack that's no longer used
2020-05-06 21:18:05 +01:00
a92580b993
StartGamePacket: remove unused imports
2020-05-06 21:14:57 +01:00
acd37b95b8
Merge branch 'next-minor'
2020-05-06 20:41:48 +01:00
88c6dcf46d
ChunkCache: clean up handling of cache cleanup on world unload
2020-05-06 13:41:36 +01:00
e392a6a807
PrepareEncryptionTask: do not store onCompletion as a field
...
pthreads will screw around with it and make it do things it's not supposed to, which is the exact reason why we have thread-local storage to begin with.
2020-05-06 13:32:18 +01:00
ff53ddd3ad
PrepareEncryptionTask: fix wrongly-specified nullability of serverPrivateKey
2020-05-06 13:29:58 +01:00
b4606a4cd0
remove PM resource interaction from StartGamePacket
...
also lose the cache, because it's not very useful ...
2020-05-06 13:17:16 +01:00
dbd51610c3
TypeConverter: use SingletonTrait
2020-05-05 19:22:20 +01:00
81f982a8d9
remove hardcoded legacy entity type ID mapping, load from resources instead
2020-05-04 13:47:39 +01:00
fcd6a69000
cleaning up NBT handling on packet decode/encode
...
now we always decode, because it's not safe to assume that we can just grab the rest of the bytes in the packet.
2020-05-04 13:23:29 +01:00
a73c54bdd0
making tile spawn compound cache use CacheableNbt instead of strings
2020-05-04 12:35:13 +01:00
c2857a91bd
[Network]ChunkSerializer: allow injecting RuntimeBlockMapping
...
this will allow the same serializer to be reused with different mapping tables (will be needed for multi version).
2020-05-04 12:14:27 +01:00
d3dcb8a4e3
moving entity attack sounds to server-side
2020-05-04 11:50:42 +01:00
8682ea35f7
Introduce some (not great) API for entity animations
...
while this API is a bit yucky, it's a step forward for protocol isolation and offers the possibility of controlling animations by adding events.
2020-05-01 13:57:26 +01:00
2964a4be35
making BlockPunchSound server-controlled
2020-05-01 12:23:00 +01:00
1969766b70
Nix some client-sided sounds, control them from the server
...
this is a necessary step to knock out the implicit assumption that every player is using the same protocol.
2020-05-01 11:52:32 +01:00
b74f177958
ChunkCache: destroy cache when world is unloaded
2020-04-29 18:50:33 +01:00
bb11cbd89c
World: relocate packet stuff for time and difficulty behind NetworkSession API
2020-04-29 18:31:54 +01:00
c8c0a1533c
NetworkSession: seal up actor removal behind API
2020-04-29 17:00:47 +01:00
adadd5423d
move force-close crafting grid hack to InGamePacketHandler
...
the crafting transaction implementation has no business caring about this
2020-04-29 16:54:23 +01:00
4ce6525065
NetworkSession: incomplete abstraction of SetActorDataPacket handling
...
this needs to have the metadata properties separated too, but that's a job that's going to get VERY messy.
2020-04-29 16:50:52 +01:00
d1b28ce17a
NetworkSession: allow sending an arbitrary set of attributes for an entity
2020-04-29 16:45:09 +01:00
f6f714c158
NetworkSession: do not mark shared attributes as synchronized, they don't necessarily belong to us
...
if we decided to start sending entity attribute changes to viewers too, this would have caused some unexpected behaviour.
2020-04-29 16:38:10 +01:00
f35b7bf80b
NetworkBinaryStream: remove stale @throws from getAttributeList()
2020-04-29 16:32:56 +01:00
ad70a9e3dc
NetworkSession: allow provision of a custom list of players to syncPlayerList()
2020-04-29 16:27:34 +01:00
d8968e9e40
box up TakeItemActorPacket sending behind NetworkSession API
...
we need to start thinking about moving this into interfaces.
2020-04-29 16:24:04 +01:00
549940d8a7
remove NullPacketHandler
...
this is a waste of LOC
2020-04-29 13:14:23 +01:00
b6214744d5
NetworkSession: inject PacketPool instead of hardcoding it
...
this will make it slightly easier for multi version implementations, but handlers are still quite a big problem.
2020-04-29 12:48:28 +01:00
09e994a026
NetworkBinaryStream: swap BadPacketException for PacketDecodeException
2020-04-29 11:39:44 +01:00
f6f1d31112
do not mess with cached data directly in packets
2020-04-29 11:34:25 +01:00
f9a587d40e
imports cleanup
2020-04-28 17:27:38 +01:00
81044d6aea
updated to latest RakLib
2020-04-28 16:30:29 +01:00
3be9548b1e
net: compressors are now fully dynamic (or at least the potential to be)
...
the compressor used by RakLibInterface when opening a session is still
hardcoded, but that's because we have no way to select the correct
compressor at that point in the login sequence, since we aren't
propagating the protocol information up from RakLib right now.
2020-04-28 16:21:18 +01:00
