Folks, I've had some email from people recently about the approach we're taking and combined with some other things I'm afraid to say I've changed my mind about things. Don't get me wrong, I still think an kernel land stack will be the way to go, but I think it's a long way off and for obos we really should be considering the nearer future. Considering the nearer future basically means that we look at creating a userland network stack that's compatible with beos 5. While this is a change of course, as we've not really done much so far it isn't a big change is it? :) The change will also lower the entry requirements for everyone and hopefully start to provide soem code that people can work on in a familiar environment rather than having to jump into a totally new environment at the same time as trying to wrestle with the complexities of a network stack. When we get towards an R2 release (if we ever get there) and we're using the newos kernel then we can look at moving over to a kernel based stack, which should be much easier having already done the userland code and having a bunch of people used to working with network protocols. I realise this is a change, but think it's one we should make. david