[DFDL-WG] Telcon minutes 2006-09-20
Mike Beckerle
beckerle at us.ibm.com
Wed Sep 20 12:09:18 CDT 2006
Attendees: Mike Beckerle, Bob McGrath, Geoff Judd
Agenda:
Report on OGF DFDL session last week
Discuss ongoing document edits.
Review ongoing actions:
Action: [Geoff] Put together a first pass description
Action: [Mike] Look at DCGs
Action: [Tom] Modifications to the scoping section.
Discussion of next steps with multidimensional arrays
Report on DFDL session at OGF last week:
- only 6 people attended, of them only 2 were "new people", one from
Boeing, one from National Archives. Presentation was uneventful.
Doc Edits:
- Geoff Judd reports that uncertainty topic has been split into core +
supplement and the results posted on GridForge. The defaults and nulls
split is still being figured out. Unclear how to split it up, what stays
core, etc.
- Tom Sugden had scoping reorg - Not on call, no report.
Ongoing Actions
- Geoff has written up an example of a group description. Still under
review/discussion with Mike and Martin
- Mike has taken a stab at starting a complementary more methodical
description. To circulate to Geoff and Martin first.
- Mike grabbed some material on Definite Clause Grammars. Seems to have
potential as a formalism for us. The issue is that DCGs are actually
programs, i.e., they're runnable when crafted in the usual way, hence, one
would expect to have to put in all the details. That is, DCGs as a
notation aren't naturally used as a summarization which hides complexity,
but rather to make all the complexity explicit. Conclusion, we need to try
using them and see how it works out.
MultiDimensional Arrays
- we reviewed the status of the work on multidimensional arrays.
- we had been assuming we needed more than just 1d array support, but this
is now unclear. We've pushed back on "top-down" use of XML Schemas, i.e.,
DFDL must be used bottom up, and transformation into the "logical form you
wanted" is not our job. With respect to multi-dimensional arrays having
complex implementations we can adopt the exact same position. The DFDL
describes the representation, and the sttructure of the DFDL schema will
end up matching the shape of the representation of the array. Transforming
that into something that looks and acts like a multi-dimensional array is
a transformation that is out-of-scope for us. E.g., if the array is stored
as a run-length encoded vector, then it is DFDL's job to describe this
run-length encoded vector, but not to project/transform it so that it can
be accessed in a manner that hides the run-length encoding and makes it
look like an ordinary dense array.
- Mike agreed to write this point up and put it before the WG on our email
list.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.ogf.org/pipermail/dfdl-wg/attachments/20060920/0564ff53/attachment.html
More information about the dfdl-wg
mailing list