On May 3, 2010, at 13:10 , Kenton Varda wrote:
Interesting. Since this seems like a JVM implementation issue, I wonder if the results are different on Dalvik (Android). Also, the extra code sounds undesirable for lite mode, but my guess is that you had to place this code inside CodedOutputStream which is shared by lite mode. So yeah, there are some down sides. Not sure how I feel about it.

Yes, I actually changed ByteString, since ByteString.copyFromUtf8 is how protocol buffers get UTF-8 encoded strings at this point.

I can understand that viewpoint. As I said, I'm somewhat surprised that this is faster, as I don't think there is a good reason for it. It also uglifies the code a fair amount. I'm somewhat tempted to file a JDK bug on this, but its unclear to me if that is actually a useful process.

What I should probably do is look at the JDK source, and figure out *why* my code actually seems faster ...


Evan Jones

You received this message because you are subscribed to the Google Groups "Protocol 
Buffers" group.
To post to this group, send email to proto...@googlegroups.com.
To unsubscribe from this group, send email to 
For more options, visit this group at 

Reply via email to