Standards WG - Local Data Dictionary - 05/21/03

9:06 Start

Steve Joy leaves as this is only a Working Group (WG) meeting

Initial requirements

1) There needed to be faster data element approval. (i.e. not CN-PDSMC but 
simply DN and peer review committee.)

2) Scope of use of local data elements was to be severly limited. There was to 
be no thought of use for correlative search using DE proposed for a LDD.
Else the data element should go through the std process.


Ron Joyner (RJ):
Presented two requirements

o improve process for getting keywords into PSDD Rafael took responsibility.
	
o RETA - any keyword should be allowed into PSDD Ron Joyner requests 5 minute 
  summary of each member's position. But first...

Laverne Hall:
Need to move forward rather than argue indefinitely. Believes we need to move 
forward urgently, so that standards progress can be shown quarterly.  Feels SCR 
progress shouldn't be bogged down by working out complete solution (i.e., all 
technical details) before moving on. Notes that deadlocked issues are to be 
forwarded to System Engineer, by MC decision at last meeting.
  
Todd King (TK):
Feels SCR process has "dramatically improved" progress (LH agrees).  Notes that 
long-term debates may reflect a) lack of time to concentrate, and/or b) real 
complexity in the issues.  LH agrees a single fixed time limit is not applicable 
equally to all subjects.
  
Lyle Huber (LH):
Notes that if discussions last long enough, people can and do change their 
minds.  Suggests we should strive for "good enough" when perfection is not 
achievable within time constraints.
  
TK:
wants to clarify that a vote against is not a deadlock or a reason to divert to 
System Engineer.  LH agrees.
  
Dick Simpson (RS):
Notes that details are important, and changing situations may require changing 
opinions. Agrees that the key to progress is keeping the standards group 
working.
  
RJ:
Back to the current SCR.  Two requirements existed: faster ingest of keywords; 
and  desire for accommodating keywords for very localized use within a dataset. 
(Keyword ingest was formerly tracked and handled by a single person no longer 
here.)  Feels the current proposal meets the above requirements
  
Proposes each WG member take five minutes to comment.
  
TK:
It was once stated that once WG passes an SCR on, WG members may no longer 
comment.  Steve Hughes notes this seems inappropriate and unenforcable.
    
RJ:
If WG reaches a compromise, it is inappropriate for WG member to re-invoke the 
original issue.  If Tech Group members do, the agreed compromise should be 
presented. SH notes the WG members who voted in favor become defenders to the 
whole Technical Group.
  
TK:
Suggests recording vote of "yes with a concern," so that concerns can be raised 
with the broader group.
  
TB:
Announces Lyle Huber is a member of WG

LH: SCR looks good as is, notes a minor duplication error.  Agrees with SCR as 
it is now, but worries that this may not work with Joe Zender's work (Rosetta, 
Mars Express and ESA generally).

Anne Raugh (AR):
No comments - position should be well known and documented.

Steve Hughes (SH):
Initial requirements were clear:
1. Faster approval (discipline node and peer review sufficient).  Might
   need to be clearer in SCR.
2. Scope was initially severely limited.  Thinks this needs to be 
   strengthened in current SCR.
   
Additional issues:
1. This is a significant change to standards (ODL syntax change) as
   written.  PDS_VERSION_ID change needs to be considered.  That 
   discussion will be continued.
2. Prototype showed reasonable chance of success.

Withdraws the "control authority" tracking provisions in favor of severely 
limited scope.
  
Joe Zender (JZ):
Requirement for local DD was strong, in preparation for >1 year, including 
planning with instruments teams who started implementation. Current solution is 
a problem, but no decision is also a problem - possibly a bigger one.

Notes that keyword names might be re-used in different name space, and doesn't 
think this should be allowed (e.g., "MEX:PDS_VERSION_ID").

Also is concerned that multi-level approval process will result in delays 
similar to the current system. Strongly recommends some decision.

Registration authority is not acceptable to ESA, adds new level of discussion 
debate, time lags to process.
	
Joel Wilf (JW):
From data modelling perspective and looking to long-term, the addition of 
namespaces is technically the right thing to do. The GROUPS implementation does 
this on a local level, the current SCR is more general and robust.  It would be 
better in long run for Rosetta/ESA to adopt namespaces.

TK:
Noticed typo in ELEMENT_DEFINITION of NAME (in comments). Should be restricted 
to 30 characters in length.
  
RJ:
This was agreed on by group.
  
SH:
Clarifies that 61 is max length if namespace is included. 30 if not. This will 
be fixed.

TK:
Agrees with Dick Simpson's comment that "Background" needs re-writing. It 
implies we are lowering quality standards.

Feels namespaces shouldn't be limited to missions; nodes should be able to 
define their own namespaces (node space, for example).

Not clear on why, logically, the pdsdd.full NAME includes NAMESPACE_ID, when 
present.  Seems to be based on implementation plan.

ODL syntax descriptions need some polishing.

Agrees with SH's restrictions on use of local keywords in cross-correlative 
searches.

Is very concerned that the Rosetta/ESA work be supported if they are not able to 
change to the current SCR standard.

Rafael Alanis is not present to comment.

RS:
Notes that there probably won't be a final decision for months, so Joe Z. can't 
have a quick decision in any event.  Thought the ESA solution was going to be 
supported.  Re: registration authority: thought that CN would do sanity-check on 
proposed local keywords, but that the nodes would do a quick review for local 
keywords.  Also concerned that the versioning aspect is going to be used as a 
trigger for PDS4.

VOTE: Approve the current SCR for referral to the whole Technical Group?
  RJ - Yes
  AR - No
  SH - Yes
  LH - Yes
  JW - Yes
  RA - not present
  TK - Yes
  JZ - Yes
  
TB: SCR will be rewritten and forwarded up to tech group. 

TK:
Asks if list of concerns will be generated as well, and can that be circulated 
among the WG well before forwarding it to Tech Group.

RJ:
Reviews his notes for changes.  Will make revisions and compile open 
issues/concerns and distribute to WG.

TK:
Requests updates and paperwork be posted to Web site first, then emailed.
    
Next meeting: two weeks time.

Planning - Ron requests resolution on Interface Control Document (ICD)
  so that work can continue on the PAG.  Other working groups are
  (presumably) proceeding apace and can continue to do so during the
  next two weeks.  Usually weekly telecon slot will be open.
  
9:55pm  End