[AFS3-std] Re: XDR extensible union type

[Andrew Deason]

On Mon, 19 Sep 2011 21:48:26 -0400
Tom Keiser <tkeiser@sinenomine.net> wrote:

> I think this also begs the question: is the above sufficient
> motivation to revise the ext-union RPC-L grammar to include an
> optional max-leg-length specification, e.g., the following:

Yes, I think so. I feel like I may have mildly objected to this in the
past, but I'm not really sure why. As you pointed out this, is directly
analogous to the array length limitations XDR already has, and the
situation is exactly the same except for memory allocation issues. I
don't think such things are really required to make the spec work, but
I'm sure they help compatibility nontrivially.

However, one point of clarification:

> +      ext-union-opt-max-leg-length:
> +         "max-leg-length" "=" value

I would have thought that this would be "max-unknown-leg-length" or
something. Are you suggesting a max length for _all_ legs, or just the
ones we skip over? I don't see a need to specify a max length for
defined legs, since they should already have a max length implied from
the formation of the elements in the leg.

-- 
Andrew Deason
adeason@sinenomine.net