> From: Oskar Sandberg <md98-osa at nada.kth.se>
> Reply-To: freenet-dev at lists.sourceforge.net
> Date: Mon, 24 Apr 2000 00:48:34 +0200
> To: freenet-dev at lists.sourceforge.net
> Subject: Re: [Freenet-dev] Build 120 -- Mac bug.
> 
> On Fri, 14 Apr 2000, Paul Kappler wrote:
> <snip> 
>> You are correct line 71 SHOULD setSoTimeout back to zero.  It seems it does
>> not.  If comment out line 71 the problem disappears.  Its very repeatable
>> but only on a mac.  This is why I thought it might be a JVM bug.
>> 
>> If a stream dies won't it throw an exception from the read or write within
>> the message handlers or conduit?  It would if we set a long SO_Timeout.
>> Maybe 5 min?
> 
> I see now that you misunderstood me when I said to change this. I meant to
> change that the second time the SoTimeOut is set it is set to something like
> 10
> minutes (rather then 0 which the mac seemed to choke on), it still needs to be
> short for that loop.
> 
> I did it this way now, does that reintroduce this bug on the Mac?

Yup, I can't insert large files to servers that are not "near me" again.  It
seems the Mac only acknowledges the first setSoTimeout.  I get the same
error as before.

I need to wonder why we don't just setSoTimeout to Core.connectionTimeout.
It would fix the problem save a little CPU.  Maybe by some command option to
fix this problem.  You said there is no way to tell if a stream dies except
by reading it.  I don't think you would know if the stream dies any later if
you had the full connection timeout for soTimeout.  I think a dead
connection ought to return -1 when the computer knows its dead.

Does this deserve a timeout testing program and along with some yanking
cables out of the switch?

So much for the platform independence of this aspect of Java.  :)

Paul


Note: Node does not compile right now without some exception hacking.


_______________________________________________
Freenet-dev mailing list
Freenet-dev at lists.sourceforge.net
http://lists.sourceforge.net/mailman/listinfo/freenet-dev

Reply via email to