Doug Zwick wrote:
>Thats amazingly large, and should make most simple commands pretty safe. I
>would think that the Java version is as important as the OS version, but
>1.3.1 and 1.4.2 all show the same limit. I'd try it under 10.2 and 10.3
>too, but I recently switched to an Intel Mac, so doing that is much less
>convenient.
I was amazed at the size, too. I expected it might be as high as 4 KB,
because that's a typical block-size, but 16 KB was a realy eyebrow-raiser.
As for 10.2, 10.3, I guess there's still a post-Leopardian use for my 500
MHz dual G4. Might not get to it for a couple days, though.
>This deadlock is not limited to Java applications, I've seen the same
>deadlock in C/C++ apps using Unix pipes and fork/exec.
Me too.
I wouldn't count on it being 16 KB for C apps. It's an easy C program to
write, because it only needs to implement the Pipelen.write() and atoi() of
argv functionality. I haven't done it, because what I needed to know at
the time was the buffer-size for inter-Java processes.
-- GG
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Java-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/java-dev/email@hidden
This email sent to email@hidden
