-
Notifications
You must be signed in to change notification settings - Fork 3.9k
ReleaseNotes
xetorthio edited this page May 30, 2011
·
10 revisions
- remove watch from transaction (Jonathan Leibiusky)
- Add binary responses to binary transaction (Jonathan Leibiusky)
- binary jedis watch command accepts several keys (Jonathan Leibiusky)
- On reconnection, select the correct db index (Jonathan Leibiusky)
- hvals response type consistency, was Collection should have been List (Jonathan Leibiusky)
- MasterSlave consistency and old mode compatibility with shard names (Dario Guzik)
- list command receive now long parameters to be consistent with return type (Jonathan Leibiusky)
- add Tuple compareTo (Jonathan Leibiusky)
- add getDB() which return the db number we are connected to (Jonathan Leibiusky)
- add publish command to Pipeline (Jonathan Leibiusky)
- add publish to transaction (Jonathan Leibiusky)
- fix md5 hashing as MessageDigest is not threadsafe, now using ThreadLocal (Jonathan Leibiusky)
- throw JedisDataException when sending NULL values to redis as it is not a valid value in the protocol (Jonathan Leibiusky)
- throw a JedisConnectionException if unsubscribing from a not subscribed JedisPubSub instance (Jonathan Leibiusky)
- handle quit command response as it was leaving the socket in an unconsistent state (Jonathan Leibiusky)
- jedis monitor should set client socket timeout to infinite (Jonathan Leibiusky)
- adding bit commands to ShardedJedisPipeline (ewhauser)
- add del() to ShardedJedis (Jonathan Leibiusky)
- don't check if jedis is connected as PING will connect if necessary (Jonathan Leibiusky)
- fix reversed boolean logic for setbit (Eric Hauser)
- multi/exec block return formatted responses (Jonathan Leibiusky)
- pipeline return formatted values (Jonathan Leibiusky)
- made shard iterable in the same order as in config during creation (Dmytro)
- fix issue 108, brpoplpush set infinite timeout before waiting for a reply from the server (Jonathan Leibiusky)
- should always default to Murmur to be consistent everywhere (Jonathan Leibiusky)
- ISSUE 78: Removed logic that waits forever till all shards are connected in ShardedJedisPool Ensuered that all commands connect at the beginning if necessary. (Dmytro)
- Compare Tuples based on the element, not the element & score (Jon Bracy)
- Implementation of ZREVRANGEBYSCORE command (lmar)
- Generate shard nodes by shard position. Shards are now host/port independent (Dario Guzik)
- Generate keys for different shards (Dario Guzik)
- add public setrange and getrange (Jonathan Leibiusky)
- pubsub should flush commands ASAP since it won't read from the socket (Jonathan Leibiusky)
- adding support for bit manipulation in to pipeline (ewhauser)
- Adding support for bit commands get/setrange (Eric Hauser)
- Adding setbit/getbit to ShardedJedis (Eric Hauser)
- Fixing Jedis.(get|set)bit to use use booleans as input and output to better match Redis commands (Eric Hauser)
- Use #discard on the transaction object (Pieter Noordhuis)
- Don't wait for QUEUED replies in MULTI (Pieter Noordhuis)
- Use regular HashSet instead of LinkedHashSet for unordered replies (Pieter Noordhuis)
- Flush to socket when starting to read (Pieter Noordhuis)
- Connect.connect() now honors timeout value (Mike Hobbs)
- add JedisDataException and JedisConnectionException
- add binary support for pubsub
- pubsub command check if we are connected
- update missing sort command in transactions api
- add defaults to pool config
- added JedisPoolConfig getters/setters so object pool can be configured with Spring/IoC
- Binary client delegates for Pipline
- add missing sort overloads to Transaction
- setbit and getbit receive long offset in BinaryJedis
- update transaction api with all the new commands of jedis 1.5.1
- updates on debug command