The end of the beginning is in sight

[Eric S. Raymond]

Hal Murray <hmurray at megapathdsl.net>:
> 
> > > That sounds uncomfortably plausible.  I can think of a workaround: add a
> > > padding extension long enough that the packet can't have any of the magic
> > > lengths. 
> > I've read that.  I've even implemented it myself once, in the Python
> > protocol back end.  Is there advice in there that I missed on how to avoid
> > magic-length interactions? 
> 
> It has a couple of minimum length constraints.  I think those are enough to 
> avoid the screwup cases.
> 
> Mostly, I pointed it out to show the complexities of that area, at least with 
> the current setup.  There may be room for something like NTPv4.1 which drops 
> compatibility with the old stuff that caused problems since it is rarely used.

And by "old stuff" I think you mean specifically Autokey, don't you?
To the extent I understand these length interactions from having coded
the Python support, I don't believe either MD5 or SHA-1 MACs are
implicated.

I think you're right.  The first thing to do about this is probably to
strengthen the non-interoperablity warnings around Autokey.
-- 
		<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>