David-

I've already solved the problem like I mentioned- by sending #signal to
the waiting Semaphore instance. And since I cannot duplicate the problem,
I'm not able to test your solution.

What does:
[Socket initializeNetwork] fork.
Socket initializeNetwork.

do that just "Socket initializeNetwork." does not?  I tried simply "Socket
initializeNetwork." but it never had any effect.

Regards,
Aaron

--
"the only difference it makes is some dust on the clay"  :: atmosphere


On Fri, 19 Sep 2003, David Farber wrote:

>
> Aaron - Does the following snippet reproduce your problem?
>
> [Socket initializeNetwork] fork.
> Socket initializeNetwork.
>
> david
>
>
> At 11:29 AM 9/19/2003 -0500, you wrote:
> >Hello,
> >
> >A year and a half ago, I posted this message regarding a networking
> >problem I encountered in Squeak:
> >
> >http://lists.squeakfoundation.org/pipermail/squeak-dev/2002-April/038975.html
> >
> >The gist being that some way, somehow, Squeak's networking stopped working
> >for me then.  I happened to me a couple times over a few months, and I was
> >never sure why.  Today, in a 3.4-based image, the same problem reared it's
> >ugly head.  Like in the 3.2 image mentioned in the above post, the problem
> >is remedied by signalling (by sending #signal) to the semaphore that is
> >waiting.  One can get to that by sending an interrupt (cmd-.) while the
> >net action is being attempted- the topmost object in the debugger is a
> >semaphore.
> >
> >This problem isn't platform specific.  I originally dealt with it on an OS
> >X box, but the problem persisted when i copied the same image to other
> >platforms.  Yesterday, I got the same problem, entirely initiated on Win
> >XP.
> >
> >Anyone have a clue what could cause this?  I have't seen much of a pattern
> >in the cause.  Anyone run into this, especially on newer versions? 3.5 -
> >3.7?
> >
> >Regards,
> >Aaron
> >
> >--
> >"one has a moral responsibility to disobey unjust laws"  :: m. l. king jr.
> >
> >
> >
> --
> David Farber
> dfarber at numenor.com
>