lucene/src/main/java/com/twitter/elephantbird/mapreduce/input/LuceneIndexCollectAllRecordReader.java (4 lines):
	- line 26: // TODO: This bit set can get big (up to about Integer.MAX_VALUE bits),
	- line 27: // TODO: ideally it should get thrown away when its no longer in use.
	- line 28: // TODO: However, it's difficult to know when that is, and the jvm will be torn down after
	- line 29: // TODO: each map task anyway.


core/src/main/java/com/twitter/elephantbird/util/ThriftToProto.java (3 lines):
	- line 24: * TODO: handle complex fields. Only primitives are mapped right now.
	- line 81: // TODO: Handle structs
	- line 87: // TODO: Handle Lists that are more than 1 level deep. Be all pro-style with the recursion.


core/src/main/java/com/twitter/elephantbird/mapreduce/input/MultiInputFormat.java (2 lines):
	- line 47: // TODO need handle multiple input formats in a job better.
	- line 111: // TODO: merge all the switch (fileFormat) cases below


core/src/main/java/com/twitter/elephantbird/util/ThriftToDynamicProto.java (2 lines):
	- line 409: // TODO: proper enum handling
	- line 561: // TODO: proper enum handling. For now, convert to strings.


hadoop-compat/src/main/java/com/twitter/elephantbird/util/HadoopCompat.java (2 lines):
	- line 79: //TODO : private static final Method H2_IS_FILE_CLOSED_METHOD;
	- line 332: // (TODO Raghu) figure out how achieve such a build with maven


pig/src/main/java/com/twitter/elephantbird/pig/util/ThriftToPig.java (1 line):
	- line 277: //TODO we should probably implement better naming, the current system is pretty nonsensical now


pig/src/main/java/com/twitter/elephantbird/pig/load/LzoBaseRegexLoader.java (1 line):
	- line 85: // TODO: stop doing this, as it can slow down the job.


pig/src/main/java/com/twitter/elephantbird/pig/util/PigToThrift.java (1 line):
	- line 62: // may be TODO : compare the schemas to catch errors early.


pig/src/main/java/com/twitter/elephantbird/pig/load/LzoBaseLoadFunc.java (1 line):
	- line 33: * TODO: Rename "BaseLoadFunc" as this class has nothing to do with LZO.


pig/src/main/java/com/twitter/elephantbird/pig/load/SequenceFileLoader.java (1 line):
	- line 199: // TODO fix Pig's silent ignorance of FrontendExceptions thrown from here


rcfile/src/main/java/com/twitter/elephantbird/mapreduce/output/RCFileOutputFormat.java (1 line):
	- line 30: * TODO: contribute this to PIG.


core/src/main/java/com/twitter/elephantbird/mapreduce/io/BinaryConverter.java (1 line):
	- line 7: /* TODO : What about exceptions?


pig/src/main/java/com/twitter/elephantbird/pig/store/LzoProtobufB64LinePigStorage.java (1 line):
	- line 32: // TODO: should remove the default constructor, after updating codegen.


pig/src/main/java/com/twitter/elephantbird/pig/store/LzoProtobufBlockPigStorage.java (1 line):
	- line 56: // TODO Auto-generated catch block


rcfile/src/main/java/com/twitter/elephantbird/mapreduce/input/RCFileThriftInputFormat.java (1 line):
	- line 146: //TODO : add error tracking


rcfile/src/main/java/com/twitter/elephantbird/mapreduce/output/RCFileThriftOutputFormat.java (1 line):
	- line 113: // TODO: handle errors


core/src/main/java/com/twitter/elephantbird/mapreduce/input/LzoJsonRecordReader.java (1 line):
	- line 24: * TODO: Fix that, and keep Hadoop counters for invalid vs. valid lines.