release

v1.1.0
1. add metric process for flange;
2. update spill logical for spill-[data] & spill-[idx];

v1.0.6
1. add spill logical with full spanId and batch size record header;

v1.0.5
1. add full spanId into record header;
2. add batch size into record header with format `#size`;
3. add catch error stack log, more catch in serialize;

v1.0.4
1. remove service info fields check in `Init()`, just have `setGlobalConfig()`;
2. check `serviceIP` and `servicePort` in `setGlobalConfig()`;

v1.0.3
1. remove `serviceIP` and `servicePort` check in `Init()`;
2. move `setGlobalConfig()` before init-times-check;

v1.0.2
1. add invalid service info fields check in `Init()`;

v1.0.1
1. move spill meta info from `SerializeSpans` to `spillSpan`;

v1.0.0
1. rebuild flange based on `batch_pb_serialize` and `batch_thrift_sending`;

v0.2.12
1. add 'WithRpcCallType()' to set 'call.type' = 'rpc' tag;

v0.2.11
1. add `WithRpcComponent()` to support rpc tag;

v0.2.10
1. add 'span_version' into each span for post-process;

v0.2.9
1. update traceId ip address to  `bcluster` and `idc` with format [04d%idc];

v0.2.8
1. update span with `traceSpans` to support baseSpan batch flush;

v0.2.7
1. reset `sequeue` to `channel`;
2. reset `ringbuffer` to raw span init, and `time.Now()` in call;

v0.2.6
1. update `seQueue` to use single `Node` saving value(*Span);
2. add `inQueue` into Span to indicate flushed;
3. remove spanId check;

v0.2.5
1. update for-condition from len > 0 to get nil, check may fix hang;
2. check for valid span in consumer/spill goroutine with len(id) > 0;
3. remove unlimit sequeue insert;

v0.2.4
1. add span retrieve debug interface;
2. fix bug for genInputSeQueue cache fixed spans of capacity;

v0.2.3
1. check flume connection before reverse spill append msg;
2. check consumer batch size before next assign;
3. add more metric content;
4. add `flange` as a flag to all log;

v0.2.2
1. add check for `genOutputRingBuffer`;
2. add check for `FromMeta`;
3. check flume connection before append msg;