charlie mail dec 3

Date: Wed, 03 Dec 2003 16:12:40 -0600
To: "Jackson, Scott M" <Scott.Jackson@pnl.gov>,
   Jennifer Schopf <jms@mcs.anl.gov>, jian@xcerla.com, bekah@nas.nasa.gov,
   mkoo@nas.nasa.gov
From: Charlie Catlett <catlett@mcs.anl.gov>
Subject: RE: survey document comments
Cc: bill@computer.org, cec@mcs.anl.gov

All-
I'm terribly sorry for such a late response.  I was reminded to look at these documents and come up to speed on the thread.  See my note below and let
 me know if it makes sense.
CeC

At 10:52 AM 11/10/2003, Jackson, Scott M wrote:
> The remaining document is simply a table that I would propose
>   be included as an appendix to the draft-rec-ur-natlang document
>   rather than as a stand alone document.

I am somewhat disinclined to do so for the following reason. The Usage Record WG plan calls for a three step approach: baseline assessment of current 
accounting practices by means of a survey paper, a natural language description, and a specification document in the selected (XML) format.
We have always considered the usage survey to be one of our deliverables (as a separate document).

I am not obstinately against it. If attaching the survey as an appendix to another document would clearly be seen as fulfillment of our first 
deliverable then I might be persuaded.

I see no reason that each working group milestone must translate into a discrete document.  So there is no issue with respect to fulfillment of 
deliverables.

But it is still my preference to see this as a separate document -- that can be appropriately referenced by the other two. Its purpose is different. 
It's author is different. It is a standalone work.

I can understand what you are saying.  The trade-off is figuring out at what level something should be included in the document series.  Ideally the 
only documents in the series are ones that you would expect to be referenced by other papers at some later date.  This is impossible to predict a 
priori, of course, but in this case I would expect the only citations to come from the other 2 documents.

I view an appendix as adding stand-alone information to a document, i.e. an internal reference, that is more or less specific to that document and 
unlikely to be referenced by other documents.  If you have an appendix that is needed in two documents I don't think that is sufficient reason to 
create a separate document, unless you anticipate that many future documents will want to reference that appendix.

We erred on the other side of this discussion when we put GFD.1 together,  because we included both the IP policy and the (c) statement as appendices.
  We are considering breaking GFD.1 into 3 documents because the IP policy and (c) statement are being used all the time (included rather than cited, 
but used as discrete items).

If you believe that lots of people will want to reference the survey results then it probably should be a discrete document in the series.  But even 
the abstract suggests that it is primarily supporting data for the record format.

I could be wrong about this- if you tell me you expect lots of people working in this field to reference the survey results I'll relax.  If the only 
reason is that the author is different or that two documents need this information, then it strikes me as an appendix.

CeC