The IP over ATM Mailing List Archive by date

[Mark Laubach <laubach@terra.com21.com>]

Yes, Interesting.  Hey, if what you describe is easier for implementers
then I'm in favor of changing the statement as part of the rewrite
provided we can get some rough consensus out of the WG. This is what
implementation experience is all about.  This is also how we find out
the "holes" in proposed standards.....<g>

Craig also, can you give some comment on the fixed ATM address sizes
in the ATMARP packets that was raised as an issue last week or so?

Mark


On Thu, 23 Feb 1995, Craig Partridge wrote:

> 
> Mark:
>     
>     Interesting.  I took the other interpretation, i.e. build a regular reply
> and set REPLY or NACK as appropriate.  Seeing the text you quote, this was
> poor reading on my part.  However, pragmatically it is a lot easier.  My code
> starts to build the ARP reply before it learns what the op code is (i.e.,
> one discovers that a NACK is coming back only when the decision to lookup
> and insert the target address fails).
> 
>     Another point is that just sending an ARP NACK means I have to parse
> two different types of ARP messages.  Regular ARP messages with src and
> target in one order, the NACK messages with them reversed.  That's a
> minor nuisance.
> 
> Craig
>