Notes from Megameeting 22nd June 2009

Attendees

BruceAmbacher UM
JohnGarrett GSFC
KatiaThomaz INPE
MarkConrad NARA
RobertDowns CIESIN, Columbia University
SimonLambert STFC
DavidGiaretta STFC
MarieWaltz CRL

Summary

The meeting focussed on the Simon's response to the document on the Wiki containing Marks comments - with David's further comments.

It was agreed that future MegaMeetings will be on Wednesdays at the same time, from July 1st.

Actions (see transcript for details):

  1. Responses to the document with DG/MC comments.
  2. section editors to correct their sections i.e. clarify unclear phrases and missing words etc which Mark has pointed out. Highlight points where there are disagreements
  3. DG to do other tidy up e.g. references
  4. all to think about how to handle "preservation" i.e. to concisely describe what we are trying to certify - see discussion below
  5. David to produce a cross-walk of CNES docuemnt to current draft
  6. review of CNES doc and cross-walk from DG

Transcript of chat

David Giaretta >> (All): Looking at the day/time choice either Tuesday or 
Wednesday would be better than today
David Giaretta >> (All): Marie - did you get a chance to look at Mark's 
comments?
David Giaretta >> (All): Simon - it looks as if you agreed with most of Mark's 
comemnts but had a question about one or two
Marie Waltz >> (All): I've only had time to glance at Mark's comments, which 
appear to be extensive and pithy in my section
SimonLambert >> (All): David - that's right
SimonLambert >> (All): Partic the treatment of authenticity
BruceAmbacher >> (All): Simon, I did email a comment on that
Mark Conrad >> (All): Saw an interesting announcement about some new repository 
software from Portugal. It claims OAIS and TRAC compliance.
David Giaretta >> (All): Oh yes RADA 
SimonLambert >> (All): Bruce - do you think it is sufficient to leave it to the 
auditor's judgement, within the guidance of the current text?
SimonLambert >> (All): Or were you implying that more is needed?
David Giaretta >> (All): Maybe some more Evidence ideas
BruceAmbacher >> (All): A reporitory can really only claim authenticity for an 
object from the time it is received.
David Giaretta >> (All): Yes
BruceAmbacher >> (All): Arepository can assess the authenticity of objects based 
on its understanding of the Producer and the past interactions with the Producer
Mark Conrad >> (All): Which metric in 4.6. are we talking about?
BruceAmbacher >> (All): A repository can only assert the claims made by the 
Producer
David Giaretta >> (All): It can supply any evidence it has collected from the 
Producer
SimonLambert >> (All): 4.6.2
David Giaretta >> (All): Maybe you can add some more examples of evidence Simon
SimonLambert >> (All): Yes, that could be the right approach.
David Giaretta >> (All): There was a mark-up questioning whether "In general the 
level of authenticity required is to be determined by the Designated Community."
was correct
Mark Conrad >> (All): My concern was with the, "This is necessary to ensure..." 
statement. 
SimonLambert >> (All): Mark - what was the problem with that?
Mark Conrad >> (All): The statement says "an authentic copy of the original 
objects, while the metric talks about copies of digital objects that are 
traceable to the originals.
David Giaretta >> (All): AH I see - the metric does not demand "authentic 
copies"
KatiaThomaz >> (All): hi all. sorry for being late.
BruceAmbacher >> (All): Certainly not if the Consumer wanted something extracted 
from multiple AIPs for example
David Giaretta >> (All): In fact I think that in my test audit there was the 
thought that it should say something like "copies or extracts "
David Giaretta >> (All): So we could change the metric to make it clear the DIP 
is not just a copy of the SIP
David Giaretta >> (All): ...and the "This is necessary" might say something 
about "consumer has confidence that what is received is what was wanted
BruceAmbacher >> (All): It is if the Consumer wants an exact copy of the AIP 
(not the SIP);  it is not if the Consumer wants something different.
David Giaretta >> (All): Yes, but the general case is that it will be something 
different
Mark Conrad >> (All): I would not change the metric. I would change the This is 
necessary statement to align it with the metric.
BruceAmbacher >> (All): There is no general case.  The repository fills a 
request with what was requested by the Consumer
SimonLambert >> (All): Mark - but is "copies of" too restrictive?
David Giaretta >> (All): Bruce - yes, but surely we have to allow the case that 
the repository supplies more than just copies of AIPs
BruceAmbacher >> (All): Agreed.  The repository fills the Consumer request for 
all, part, several, etc.
Mark Conrad >> (All): Simon, Copies of digital objects is not too restrictive 
for me.
David Giaretta >> (All): BUt they might be processed or reformatted - surely we 
must allow that
David Giaretta >> (All): ...or would you interpret copies  broadly enough to 
include taht?
BruceAmbacher >> (All): The key is whether the DIP fulfills the Consumer 
request.
David Giaretta >> (All): Yes agreed - with he reuirement about traceability to 
originals  etc
KatiaThomaz >> (All): and is traceable to the original
BruceAmbacher >> (All): It may be "original(s)"
David Giaretta >> (All): yes
KatiaThomaz >> (All): yes
RobertDowns >> (All): yes
Marie Waltz >> (All): This is good
BruceAmbacher >> (All): As long as we are sure the original(s) are the AIPs
Mark Conrad >> (All): I took this metric to mean that the repository should have 
the capability to produce a "roadmap" of how it produced the digital objects 
requested by the Consumer from the AIPs that it holds. The roadmap should 
indicate how evidence of authenticity was maintained across any transformations.
SimonLambert >> (All): ... AIPs rather than SIPs?
KatiaThomaz >> (All): yes
David Giaretta >> (All): Mark - I guess I and others were reading copies as more 
restritcive than that
BruceAmbacher >> (All): Simon, yes.  The repository does not have to retain the 
SIP once it makes a valid AIP
Mark Conrad >> (All): Simon, Yes AIPs. I believe there is a previous metric that 
covers the transformation of SIPs to AIPs.
BruceAmbacher >> (All): Some repositories may also save SIPs in anticipation 
that future technology will allow them to some other processing
David Giaretta >> (All): So what about the "copies" - is that too restrictive?
Mark Conrad >> (All): David, it says copies of digital objects - not copies of 
AIPs.
David Giaretta >> (All): ...or maybe just delete the "copies of"
David Giaretta >> (All): so "disseminated digital objects that are traceable..."
Mark Conrad >> (All): If that removes some confusion, I would be in favor of 
deleting copies of.
JohnGarrett >> (All): But there is no requirement to save SIPs
KatiaThomaz >> (All): could be samples?
David Giaretta >> (All): Samples could be traced to originals
BruceAmbacher >> (All): John, I agree there is no requirement to save SIPs but 
there also is no requirement to dispose of them.
David Giaretta >> (All): Is it agreed to delete "copies of" in the metric?
BruceAmbacher >> (All): Why introduce vague language like samples?
JohnGarrett >> (All): I think copies of digital objects was used to give an 
indication that you are not providing access to the 'original' copy in the AIP.
David Giaretta >> (All): But that is consistent
Mark Conrad >> (All): Is there a requirement that DIPs have to be created from 
AIPs?
JohnGarrett >> (All): It's OK with me to drop "copies of"
BruceAmbacher >> (All): This metric focuses on order fulfillment - did you gave 
the Consumer what he/she/it wanted?  Is it traceable to your AIPs if any 
question arises.
JohnGarrett >> (All): No DIPs could have information added by repository from  
information holdings they have that are not in AIPs.  For example, they could 
have additional format information that was not in an AIP, but would be useful 
to the requester.
BruceAmbacher >> (All): DIPs should meet the Consumer request, whatever the 
source, and be traceable to those sources if need be
Mark Conrad >> (All): I only raised the question in relation to your earlier 
comment about saving the SIPs. If a repository developed a DIP based on an SIP 
rather than an AIP, would that be acceptable?
BruceAmbacher >> (All): Mark, I think it would be acceptable as long as the 
Consumer understands the source(s) of the DIP
David Giaretta >> (All): As long as it's an original - from the Producer
David Giaretta >> (All): What about the "This is necessary" statement?
David Giaretta >> (All): SHould that include Bruce's point about "satisfying the 
consumer's request " 
JohnGarrett >> (All): Depends on if requester needs an authentic copy of the 
data object.  Repository has an authentic copy in the AIP.  Hasn't necessarily 
ensured that SIP copy remains authentic.
Mark Conrad >> (All): David, I don't see that this metric is specifically about 
meeting the Consumers needs. Its more about the quality of the policies and 
procedures used by the repository to produce the DIP.
David Giaretta >> (All): Mark - I see what you mean - I was just wondering if we 
should take bruce's point on board here
David Giaretta >> (All): ...i.e. the procedures should ensure the repository 
satisfies the comsumers
Mark Conrad >> (All): David, See 4.6.2.1.
BruceAmbacher >> (All): I agree with Mark that the metric is about HOW the 
repository provides digital objects and traces those objects to AIP(s) and other 
reliable data objects, then informs consumer about those sources
David Giaretta >> (All): OK
David Giaretta >> (All): ...and then the "traceable to.." is so that comsumer 
has confidence in the result - or can make a judegment about it
David Giaretta >> (All): ...I meant that something like that should be in the 
Supporting Text
JohnGarrett >> (All): I agree
David Giaretta >> (All): By  the way - if we agree on this that deals with 
mark's comemnts about 4.6 but we need responses to the rest of mark's comemnts
Mark Conrad >> (All): How about, This is necessary to establish an auditable 
chain of authenticity from the AIPs to the disseminated digital objects.
David Giaretta >> (All): Looks good
David Giaretta >> (All): ..or is it "from the Producer"?
Mark Conrad >> (All): I stole most of it from Simon's rewrite of the discussion. 
:)
Mark Conrad >> (All): So from the SIPs?
David Giaretta >> (All): But the procedures could tdetail things from the AIP 
and we already know we can trace the AIPs to the SIPs and also have provenance 
from the Producer
Mark Conrad >> (All): So leave it as I proposed?
RobertDowns >> (All): I think so
David Giaretta >> (All): Yes
Marie Waltz >> (All): Fine
David Giaretta >> (All): Does that fix 4.6?
BruceAmbacher >> (All): I can agree but I think this stretches authenticity.  I 
am not uncomfortable enough to offer a substitute at this time.
Mark Conrad >> (All): Bruce, What do you mean stretches authenticity?
David Giaretta >> (All): How about "chain of evidence of authenticity"?
BruceAmbacher >> (All): Many DIPs produced by this process will be, by 
agreement, not an exact copy and hence not authentic.  They are, however, 
traceable to authentic object(s)
BruceAmbacher >> (All): As I said I can accept this rewrite.
David Giaretta >> (All): OK
Mark Conrad >> (All): Bruce, Agreed. You think this is not covered by the 
rewrite?
Mark Conrad >> (All): How about if we added whether or not the DIP is an exact 
copy of the AIP?
BruceAmbacher >> (All): I can accept what Mark wrote, based on Simon's revised 
text.  We have established elsewhere that DIPs doe not equal AIps in all cases
David Giaretta >> (All): If authenticity is abut being what something is 
purported to be then if a repository claims something is a processed piece of 
data based on an image of the Earth dated 3 SSept 2007 then I should be able to 
see all the evience of that
Mark Conrad >> (All): Yes.
BruceAmbacher >> (All): Agreed.  The consumer or the repository will indicate 
what objects should be processed to    achieve the image of 3 Sept 2007.  The 
consumer will also agree by accepting the DIP
David Giaretta >> (All): So we need evidence about the "oringinal Earth image" 
and then evidence about the processing etc. Is that right?
Mark Conrad >> (All): So. "This is necessary to establish an auditable chain of 
authenticity from the AIPs to the disseminated digital objects." or "This is 
necessary to establish an auditable chain of authenticity from the AIPs to the 
disseminated digital objects, whether or not the DIP is an exact copy of the 
AIP." ?
David Giaretta >> (All): I don't think we need that last phrase. But perhaps we 
may want to add something to the discussion
Mark Conrad >> (All): Bruce, What do you think?
BruceAmbacher >> (All): The first rewrite is preferable.  We have qualified what 
a DIP is elsewhere.
Mark Conrad >> (All): So is everybody ok with the first rewrite?
KatiaThomaz >> (All): yes
Marie Waltz >> (All): OK
RobertDowns >> (All): yes
David Giaretta >> (All): yes
SimonLambert >> (All): yes
David Giaretta >> (All): Great
David Giaretta >> (All): We need the rest of Mark's comments to be addressed
David Giaretta >> (All): and also Daniele's document
Marie Waltz >> (All): What do you want us to do with Daniele's document?
David Giaretta >> (All): It is a little difficult since it refers to an older 
version
BruceAmbacher >> (All): I discovered my comments on his comments for 4.1 did not 
save in the original.  I guess that's what "Read only"means.  Duh.  I will 
transmit my comments in an email after this.
Mark Conrad >> (All): Simon has done a great job of rewriting the discussion 
section. The only comment I have is that there is a typo. It should be "an 
auditable trail..." not "and auditable trail".
David Giaretta >> (All): I was planning to try to do a cross-walk and see if 
there were any points that seemed not to have already been addressed.
BruceAmbacher >> (All): Lets talk about how to handle the CNES product.  I am 
uncomfortable dealing with comments that are not enlightened by participation in 
the webchats,  I would prefer to get their comments on the final draft before 
submission to ISO pricess
David Giaretta >> (All): Then we can sweep things up together with the 
discussion of Mark's comemnts about the other sections
Mark Conrad >> (All): I agree with Bruce.
David Giaretta >> (All): Bruce - I agree but it is a matter of timing. As I say, 
if they have made comemnts that we have not already thought of then we can 
consider them - otherwise they will just be made at the ISO review stage.
David Giaretta >> (All): One purpose of what we are doing in an open process is 
to try to anticipate and deal with coments before the ISO review stage
BruceAmbacher >> (All): If that process is followed, some group (with some/all 
of us) would adjudicate all comments received?
David Giaretta >> (All): Yes, we would have to.
JohnGarrett >> (All): I think we can probably take a look at the comments CNES 
have and see what applies to this version.  
David Giaretta >> (All): I guess I am anxious about the timing because IF we 
need another meeting to make those decisions I would like it to be while there 
is the possibility of having some funding for a meeting.
BruceAmbacher >> (All): David, I agree, but this document is changing so fast 
that comments on a month-old version may be superseded.  Will you address that 
in your cross-walk?
David Giaretta >> (All): Bruce - I'll try to but it will only be on a best 
efforts basis.
David Giaretta >> (All): As I understand it CNES will do another review based on 
the final docuemnt - and I assume that will feed into the ISO review
David Giaretta >> (All): John - I remember that we were told that it might take 
8 months for the ISO review. If we can get it in before the end of July that 
would mean March 2010
David Giaretta >> (All): The danger is that the CCSDS editor has to have some 
time to prepare the docuemnt.
BruceAmbacher >> (All): What does that 8 moths include? Dissemination and 
comments?  Dissemination, comments, adjudication?  Or the whole process to ISO 
status by March 2010?
David Giaretta >> (All): I assumed that meant to receive the coments - John any 
views?
David Giaretta >> (All): Clearly if some countries have show-stopper comemnts 
then we are in trouble! 
JohnGarrett >> (All): I think we plan on receiving comments
BruceAmbacher >> (All): That would mean that the standard would not be issued 
until 2011 - the comments would have be reviewed, the draft standard revised and 
another round of review and comment before standard status
David Giaretta >> (All): OK - then we as a group resolve them - but, unless 
there are showstoppers we don't need another round of reviews
BruceAmbacher >> (All): ok
David Giaretta >> (All): ...I think that's how it went with OAIS
Marie Waltz >> (All): Whatever it takes
JohnGarrett >> (All): For the ISO process there are set periods of time for 
countries to respond with comments to the documents.
Mark Conrad >> (All): Action items for next week?
BruceAmbacher >> (All): John, does that timeframe vary from TC to TC or is it an 
ISO set time frame for all standards?
JohnGarrett >> (All): I don't think we need another round of reviews if we can 
respond to the comments that come in
David Giaretta >> (All): How about (1) responses to mark'scomemnts (2) review of 
CNES doc and cross-walk from DG (3) any other actions from last week
JohnGarrett >> (All): The ISO process is a set time period across all TCs.
BruceAmbacher >> (All): I expect to be available Monday but have another meeting 
set for Tuesday if we change days
David Giaretta >> (All): Also when do we meet next? Should we make it Tuesday or 
Wednesday?
David Giaretta >> (All): Bruce - should we make it WEednesday?
Marie Waltz >> (All): Yes please, I'm not here Monday either
JohnGarrett >> (All): We can get to a CCSDS standard first and then send it in 
to ISO and the ISO process is shorter then, but the overall time is about the 
same.
JohnGarrett >> (All): Wednesdays are much better for me.  I haven't responded 
yet on the wiki.
BruceAmbacher >> (All): This standard has broader interest than just CCSDS and 
should be exposed to wider review/comment
David Giaretta >> (All): So looks like next week Wednesday same time
BruceAmbacher >> (All): Wednesday will be ok
Mark Conrad >> (All): So the next meeting will be on July 1?
David Giaretta >> (All): Are the actions I suggested OK?
Marie Waltz >> (All): Yes
David Giaretta >> (All): But we need comemnts on Mark's comemnts for sure
Marie Waltz >> (All): I hear you
BruceAmbacher >> (All): Same start time on 7/1?
David Giaretta >> (All): Yes - the responses indicated that would be OK - unless 
anyone has a problem with taht
Mark Conrad >> (All): That would be good for me.
KatiaThomaz >> (All): ok. bye and have a nice week.
Marie Waltz >> (All): Bye
BruceAmbacher >> (All): ok.  My comments on 4.1 will out today
David Giaretta >> (All): I'll take an action to email section editors directly
David Giaretta >> (All): Bye all
-- DavidGiaretta - 22 Jun 2009
Edit | Attach | Watch | Print version | History: r2 < r1 | Backlinks | Raw View | Raw edit | More topic actions
Topic revision: r2 - 2009-06-22 - SimonLambert
 
This site is powered by the TWiki collaboration platform Powered by PerlCopyright © 2008-2018 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback