I believe that diald-99 may have a bug and I wanted to check first
with the experts here first before I stick my neck out and write Mike
Jagdis. I believe the bug is most likely to affect people at home
who are using some sort of Masq/diald combo to share their
Internet connection.
Bottom line: it appears that when using the "dynamic" option, diald
drops the first few packets it receives causing a small delay (a
minute or so) when the link is first brought up. In my experience,
the lost packets are often sufficient to cause the first few Web or
other connections I make to fail.
README.masq states that diald is supposed to send "the buffered
packets *back* to the kernel via the proxy interface rather than
sending to the real interface." This is so the packets can be
properly masqueraded.
The evidence: I have watched sl0 using tcpdump and while
packets go out over the interface, no packets are ever sent back
over the proxy interface. The source code provides a plausible
reason why this might be so: it appears that diald downs the proxy
interface before it gets around to forwarding the packets, thus
guaranteeing the operation will fail!
So what do you all think? Is this a bug and if so, should I write
Mike Jagdis about it?
TIA for your thoughts and info.
Ed G.
-
To unsubscribe from this list: send the line "unsubscribe linux-diald" in
the body of a message to [EMAIL PROTECTED]
