Adds the ability to configure things needed by stackdriver

We will eventually need to change the encoder to use their format. Also,
stackdriver does not support shared spans.

spring-cloud-sleuth-zipkin/src/main/java/org/springframework/cloud/sleuth/zipkin2/ZipkinAutoConfiguration.java

@@ -20,6 +20,7 @@ import java.util.concurrent.TimeUnit;
 
 import brave.sampler.Sampler;
 import zipkin2.Span;
+import zipkin2.codec.BytesEncoder;
 import zipkin2.reporter.AsyncReporter;
 import zipkin2.reporter.InMemoryReporterMetrics;
 import zipkin2.reporter.Reporter;
@@ -78,13 +79,20 @@ public class ZipkinAutoConfiguration {
 	public Reporter<Span> reporter(
 			ReporterMetrics reporterMetrics,
 			ZipkinProperties zipkin,
-			Sender sender
+			Sender sender,
+			BytesEncoder<Span> spanBytesEncoder
 	) {
 		return AsyncReporter.builder(sender)
 				.queuedMaxSpans(1000) // historical constraint. Note: AsyncReporter supports memory bounds
 				.messageTimeout(zipkin.getMessageTimeout(), TimeUnit.SECONDS)
 				.metrics(reporterMetrics)
-				.build(zipkin.getEncoder());
+				.build(spanBytesEncoder);
+	}
+
+	@Bean
+	@ConditionalOnMissingBean
+	public BytesEncoder<Span> spanBytesEncoder(ZipkinProperties zipkinProperties) {
+		return zipkinProperties.getEncoder();
 	}
 
 	@Bean