variables, simplified the splayprint function, modified Curl_splayremovebyaddr() to return error code. All in an effort to track down the reported splay problem, but I've failed to do that so far...
nothing matched fine
code rearrange to fit the future better.