Andreas.Raab at gmx.de wrote:
 > glab server). There's clearly a huge inefficiency (bug?!) in
 > SocketStream's upToAll: implementation which will *always* have to
 > time out unless the server

Good one!!
Because of the problems/hangs I had I changed it to receiveData, if you 
try receiveDataNoWait you get a significant speedup (which was the 
intended version all along). Going into the next version as well...

And as things go I can't reproduce the hangs now, 
the-one-higher-being-we-worship (or not) knows what circumstances are 
needed to reproduce this.

Michael