Oh neat! So would that use MappedByteBuffer? And in that regard, would it make sense to use ByteBuffers instead of byte arrays? I wasn't very sure about how possible it was to convert a direct ByteBuffer back into a byte array if necessary.
I'm not super experienced with the NIO API, but there must be some parts of it we can take advantage of. I really like the buffers APIs. We use Charsets, of course, so there's another part. Then again, Java 1.7 made NIO a lot more useful, so I don't know how much we could really do to take advantage of NIO without NIO.2. On 4 August 2014 19:51, Remko Popma <[email protected]> wrote: > Personally I'd like to work on some new features. Specifically, binary > logging, a memory-mapped file appender and providing configuration support > for some system properties (async loggers, JMX). > > So I'd like to work towards a 2.1 release. > > As you said, if some critical bug pops up we can branch off the 2.0.1 tag, > and do a fix on that branch. > > Sent from my iPhone > > On 2014/08/05, at 9:10, Matt Sicker <[email protected]> wrote: > > Are there any outstanding issues we'd like to address in a 2.0.2 release, > or should we just start working toward 2.1 now instead? Because if we go > the 2.1 route of focus, I've got a few branches to merge back together > (thankfully, git-svn will help a lot in that regard) into trunk. > > As Ralph (IIRC) pointed out, we don't need to make an explicit 2.0 branch > since we can just branch from the 2.0.1 tag itself if necessary. > > -- > Matt Sicker <[email protected]> > > -- Matt Sicker <[email protected]>
