Document a bit that UDP is not what you might think.

Subject: Re: IO::Socket::INET bug sending large UDP packets/fragmentation
Message-ID: <20000724085915.B13172@alanya.lupe-christoph.de>

p4raw-id: //depot/perl@6534

diff --git a/pod/perlipc.pod b/pod/perlipc.pod
index 6467a29..47556a5 100644 (file)
--- a/pod/perlipc.pod
+++ b/pod/perlipc.pod
@@ -1246,6 +1246,11 @@ find yourself overly concerned about reliability and start building checks
 into your message system, then you probably should use just TCP to start
 with.
 
+Note that UDP datagrams are I<not> a bytestream and should not be treated
+as such. This makes using I/O mechanisms with internal buffering
+like stdio (i.e. print() and friends) especially cumbersome. Use syswrite(),
+or better send(), like in the example below.
+
 Here's a UDP program similar to the sample Internet TCP client given
 earlier.  However, instead of checking one host at a time, the UDP version
 will check many of them asynchronously by simulating a multicast and then
@@ -1296,6 +1301,11 @@ with TCP, you'd have to use a different socket handle for each host.
        $count--;
     }
 
+Note that this example does not include any retries and may consequently
+fail to contact a reachable host. The most prominent reason for this
+is congestion of the queues on the sending host if the number of
+list of hosts to contact is sufficietly large.
+
 =head1 SysV IPC
 
 While System V IPC isn't so widely used as sockets, it still has some