bsbontchev

static void TranscribeSpeechAsyncHandler( const Aws::TranscribeService::TranscribeServiceClient* client, const Aws::TranscribeService::Model::StartTranscriptionJobRequest& request, const Aws::TranscribeService::Model::StartTranscriptionJobOutcome& response, const std::shared_ptr& context) { if ( response.IsSuccess() ) { const Aws::TranscribeService::Model::TranscriptionJob& transcription_job = response.GetResult().GetTranscriptionJob(); if ( ( transcription_job.GetTranscriptionJobStatus()...

[awssdk.zip](https://github.com/aws/aws-sdk-cpp/files/5378968/awssdk.zip) This is a working example to reproduce the issue. Unzip Run "make clean; make" Run "build/test "your_aws_key_id" "your_aws_key" "s3://yourbucket/yourfile.wav" "yourbucket"" The output will be: arg_initialize() result: 1 sdk_initialize() transcribe_initiailaize()...

Tested the latest AWS SDK for C++ with Version:1.8.128, curl 7.68.0, OpenSSL 1.1.1f on Ubuntu. The transcribe issue, e.g. only IN-PROGRESS and no COMPLETE event is still present. [aws_sdk_2021-01-22-19.log](https://github.com/aws/aws-sdk-cpp/files/5867277/aws_sdk_2021-01-22-19.log)

The usual live transcribe stream design is a source(file, audio device with a constant read speed), stream buffer(to accommodate the delays between the source and the sink), and a sink({AWS...