> bryan@terran.org said:
> } I know this question has been asked a few times and I have seen the
> } URL's flying around a bit. However, linux.kernel.org is still
> } reporting 2.0.33 as the latest stable kernel and I'll take it's word
> } for it. Are there plans for a formal 2.0.34 release, and if so, how
> } close are we? I'm actually on the bleeding 2.1 kernel series myself,
> } but I'm doing some work with 2.0 in a few places and I'd like to
> } appraise the 33-34 diff, if/when one becomes available. :)
>
> Much as I hate to say it, it looks like we are going to produce some stuff
> that *ought* to go into the 2.0.x series.
>
> Justification:-
> --------------
>
> Masq has a set of bugs in 2.1.x and I believe (but have not tested,
> although some recent bug reports certainly tend to confirm) that the same
> set also exist in 2.0.x. The problem is that (some of) the checksum
> generate/check routines do not correctly handle ip options resulting in
> mangled/dropped packets. This is a pretty rare problem at the moment, but
> David Miller's funky new network code slops options all over the place.
> This would mean that without a fix we would have an inter-operability
> problem between 2.0/2.2 kernels.
Not IP options, but TCP options.
Note that Win98 uses TimeStamp/Window Scaling too, so when there are really
bugs in this regard you should receive lots of bug reports soon. Also alot
of other OS do [at least IRIX, 4.4BSD]
-Andi
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu