RE: Rework on SOAP 1.2 Part 2: Section 2 and 3

Hi Noah,

Firstly, i think that from the discussion it has become clear to me (i
think) that in this data model nodes are *not* labelled, only edges, which I
think is what I missed.

So my example: 

> Consider this graph:
 > 
 >           edgeB       +-------------+
 >       +-------------->+ "terminalB" |
 >       |               +-------------+
 >       | 
 >  +----+----+  edgeA   +-------------+
 >  | structA +--------->+ "terminalA" |
 >  +----+----+          +-------------+
 >       |
 >       |    edgeC      +---------+ edgeD  +-------------+
 >       +-------------->+ structB +------->+ "terminalD" |
 >                       +----+----+        +-------------+
 >                            |
 >                            |   edgeF     +-------------+
 >                            +-------------+ "terminalE" |
 >                                          +-------------+

Becomes:
            edgeB
          +------->("terminalB")
          |
structA   | edgeA
--------->O------->("terminalA")
          |
          | edgeC       edgeD
          +--------->0-------->("terminalD")
                     |
                     | edgeF
                     +-------->("terminalF")

[I can hear Gudge sigh... "Well it does say "...directed, edge-labeled graph
of nodes..." do we really gave to spell out that only edges get labelled?"
;-)]

Actually, I think a small pictorial example would be helpful... but I'd let
that hang as editorial discretion.

So I'd conclude that nodes don't really have labels in this model, only
edges. Is it the case that terminals are all (possibly typed) literals (I
think).

The thing that then feels awkward is that multirefs then target edges rather
than nodes (because nodes don't have labels or you might wriggle and say
that the id labels the node not the edge). I guess they indirectly target
nodes as being on the end of an edge (and any inbound edge should do).

I think what would be really clear would be to say that edges are
represented by element names and that nodes are represented by element
content, non-terminals have only elements as content and that terminals have
typed literals as content. (of course turned into infoset'ise). Maybe that
is indeed what it says... I'll read it again.

What was throwing me was the notion as a whole that the EII could represent
both an edge and a node.

Thanks,

Stuart

> -----Original Message-----
> From: noah_mendelsohn@us.ibm.com [mailto:noah_mendelsohn@us.ibm.com]
> Sent: 15 March 2002 17:14
> To: Williams, Stuart
> Cc: Martin Gudgin; 'Jean-Jacques Moreau'; xml-dist-app@w3.org
> Subject: RE: Rework on SOAP 1.2 Part 2: Section 2 and 3
> 
> 
> Stuart Williams writes: 
> 
> >>  The first sentence states: "Each graph edge is encoded 
> >> as an element information item and each 
> >> element information item represents a graph edge."
> 
> ...
> 
> >>  The contractiction arises from the first entry
> >> in the itemised list that follows: "...then 
> >> the element information item is said to represent 
> >> a node in the graph and the edge terminates at that 
> >> node."
> >> This now states that element information items can 
> >> also represent graph nodes - contraticting the 
> >> initial sentence.
> 
> I see what concerns you.  I suppose the first two paragraphs might be 
> reworked as follows:
> 
> <original>
> Each graph edge is encoded as an element information item and 
> each element 
> information item represents a graph edge. 3.1.3 Encoding 
> compound values 
> describes the relationship between edge labels and the local name and 
> namespace name properties of such element information items. 
> 
> The node at which an edge terminates is determined as follows:
> </original>
> 
> <proposed note="includes a bit of other cleanup">
> Each graph edge is encoded as an element information item. The 
> representation of outbound edges is described in 3.1.3 
> Encoding compound 
> values.  That section describes the relationship between edge 
> labels and 
> the local name and namespace name properties of such element 
> information 
> items. 
> 
> Each element information item represents (a) a graph edge, as 
> described 
> above, and (b) optionally, a node at which that edge 
> terminates.  Nodes 
> and their associated inbound edges are encoded as follows:
> </proposed>
> 
> I'm a bit torn as to which formulation I prefer.  The latter is more 
> explicit, more clearly balances the presentation of inbound 
> and outbound 
> edges, and probably deals with your concern.  The original is a bit 
> shorter and crisper.  Opinions?
> 
> ------------------------------------------------------------------
> Noah Mendelsohn                              Voice: 1-617-693-4036
> IBM Corporation                                Fax: 1-617-693-8676
> One Rogers Street
> Cambridge, MA 02142
> ------------------------------------------------------------------
> 
> 
>

Received on Friday, 15 March 2002 13:08:19 UTC