On Wednesday, August 27, 2003, at 04:56 PM, Alex Rice wrote:


I think the goal is to get the primary IP address of the computer not of a connection. I may be confused there, too.

But is there such thing as really a primary IP address? There is a default gateway for each network interface. There can be priorities assigned to different routes. However, the particular interface to use isn't decided until the OS needs to send some particular network packets and the OS goes- ok little packet: where's that network- where's that default gateway- what's the route's priority hm.. hm... you go out this interface.


Maybe this is why hostAddress requires a socket to be opened- it is only capable of answering the endpoint IPs of that one socket.

That is a very good point.


I have some computers with several ethernet adaptors, some over-modem adaptors, and several VPN adaptors. And then there is 127.0.0.1. I have some computers which connect to multiple subnets on the same ethernet adaptor. Which is primary?

I think Windows thinks one is primary, but I'm not sure.

This is why I like the idea of an adaptors() or ipAddresses() function. Or the ability of hostAddress to take an unopened connection.

Now if only hostAddress will take a UDP opened socket... That is really in some sense an unopened connection.

Dar Scott



_______________________________________________
use-revolution mailing list
[EMAIL PROTECTED]
http://lists.runrev.com/mailman/listinfo/use-revolution

Reply via email to