RE: [nslu2-linux] Re: NSLU2 - alternative firmware -- R29
- From: Kurt Bernhard Pruenner
>The network chipset used on the slug is capable of doing the crossoverWell I, for one, feel really stupid at this point...
>by itself if neccessary, so it'll work just the same with a normal patch
>cable - and I don't think that there are several hardware revisions of
>the slug with different network chipsets, or are there?
You are correct - a cross-over cable works just as well as a straight
through. I just never checked before. (The LinkSys documentation implies
the NSLU2 doesn't handle this, but clear it does). Indeed my timing tests
suggest that the times to get the ethernet up and running are identical
regardless of which cable is used (timing accuracy is only +/-1s, but the
non-cross-over seemed slower if anything - though not enough to be
From: J.P. Alten
>Well, I'm really not sure if the slug can do the crossover by itself,It can.
>but if it can, both the PC and the slug will notice that there isThat assumes that the PC (or whatever) card can do crossover. The more
>something wrong with a regular cable, try to do the crossover, notice
>that there is still something wrong and negotiate who will do the
likely scenario is that the NSLU2 is plugged into a router which can do the
swap (like the WRT54G). I can see no evidence that the startup time is
longer in this situation.
>Both ways, it will take a few seconds, and if you are trying to use theI can detect no difference to the time at which the ethernet is active in
>TelnetItoRedBoot method to access the slug, you only have 2 seconds to
>gain access. (http://www.nslu2-linux.org/wiki/HowTo/TelnetIntoRedBoot)
any of these scenarios.
Getting into the RedBoot telnet has been difficult for everyone, there is a
three second window about 8 seconds into the bootstrap.
So far as I can see all this stuff about cross-over cables and delays
because of the use of the wrong sex cable is wrong, and irrelevant to the
John Bowler <jbowler@...>