Could you please create a JIRA issue for this and maybe attach your example there?
regards, Karl On Wed, Jul 2, 2008 at 3:22 PM, Craig Phillips <[EMAIL PROTECTED]> wrote: > Hi, > > > > I uncovered an issue in which the Shell.TUI will cause "new > java.io.ServerSocket()" to hang; > > > > Through process of elimination / gutting, pretty much to the point of a > small bundle with an activate() method (I'm using SCR) and a "new > ServerSocket(1234, 1)", with a couple log events pre and post, I > determined that the construction of the ServerSocket will not return > (nor timeout, IIRC), if the Shell.TUI bundle is started first; > > > > In other forms of the telnet shell, which do not use > SCR/DeclarativeServices, I remember I had to put install and start those > bundles prior to the Shell.TUI... However, because I changed the > ordering a la SCR (my bundle is installed and started, but SCR does not > start my "service" and activation thereof until my dependencies are > satisfied, in this case to shell service and log service), my bundle > activation falls behind the shell.TUI activation... > > > > So, the Shell.TUI must be causing some sort of conflict with > ServerSocket ... I did not dig deeper than that... Not sure if someone > would ask me to file a bug or not... I'm just putting this out there... > maybe others have come across this, but I don't know... > > > > I solved my immediate problem by simple removing Shell.TUI from the > auto.start list; > > > > For what it's worth, Craig Phillips, Praxis Engineering > > -- Karl Pauls [EMAIL PROTECTED]
