From: Godmar Back (gback@cs.utah.edu)
Date: Mon Feb 08 1999 - 11:37:38 EST
>
> On Feb 8, 1999, Godmar Back <gback@cs.utah.edu> wrote:
>
> >> On Feb 8, 1999, Godmar Back <gback@cs.utah.edu> wrote:
>
> >> > Alexandre, why did you add a timeout value to connect?
> >> > There's no use for it, is there?
>
> >> Of course there is. Do you think I'd have spent time on it if I
> >> didn't really need it? :-)
>
> > Can you tell me how JacORB even accesses it?
>
> No :-)
>
> I've just talked to Hilson (the guy who's trying to run JacORB on
> Kaffe) and he told me that JacORB was probably waiting for a timeout
> on a read(), not a connect(), so I probably misunderstood him before.
> So, we can safely revert this change.
>
> > The connect is done in the socket constructor, and there's no timeout
> > parameter. In fact, in the patch you checked in, timeout is always
> > zero for connect. Am I missing something?
>
> No, I think you're absolutely right, and my patch was just an attempt
> to complete a work that was already completed :-)
>
> Should I revert it or are you going to do it yourself (I'd prefer the
> latter :-)
>
I haven't even checked out your patch, besides wasn't there something
you really needed to fix with EISCONN?
- Godmar
This archive was generated by hypermail 2b29 : Sat Sep 23 2000 - 19:58:01 EDT