Notes from Megameeting 10th September 2007


DavidGiaretta STFC
DonaldSawyer NASA GSFC
KatiaThomaz INPE, Brazil
MarkConrad NARA
JohnGarrett NASA
HelenTibbo UNC
BruceAmbacher UM

All the discussion at this meeting was conducted by chat, so the following transcript of the meeting (with a few typos corrected) is complete.

Katia Thomaz >> (All): hi.
Don Sawyer >> (All): Hi
David Giaretta >> (All): We needed Bruce - he was in the call then dropped out. I imagine he is goin g to rejoin
Don Sawyer >> (All): yes, hear you
David Giaretta >> (All): Should we start?
David Giaretta >> (All): Let's give it 1 more minute
David Giaretta >> (All): While we wait I guess the firts thing on the agenda is Bruce's email comments - just to give Bruce a heads up
David Giaretta >> (All): We should start
David Giaretta >> (All): Bruce - over to you
David Giaretta >> (All): Yes, I agree we need to keep at the higher level
David Giaretta >> (All): Maybe we should follow Katia's advice and focus on those points where we have written comments and in particular where comparison with other checklists suggests something may be missing
BruceAmbacher >> (All): Mark, What is your take on all of this since you were not involved in developing TRAC?
Katia Thomaz >> (All): i think when someone proposes a change it´s very important to jusify (with sources etc.)
Mark Conrad >> (All): I agree that we need to look at the 10,000 ft level.
David Giaretta >> (All): It also has the advantage of being faster
Mark Conrad >> (All): I think we have been too focused in the weeds and have lost the big picture.
David Giaretta >> (All): Katia - again I agree - let's not do things "on the fly"
Don Sawyer >> (All): My concern is both from a higher level, and the lower level.  From a higher level, there is the need to review the document from the perspective of an auditor.  
                        From a lower level, and from the perspective of an archive, there are some items that I think are not needed and some need clarification.  So, I don't have a problem working at either level.  
                        Those who were not involved in TRAC productions may not have a good overall view.  This may indicate an issues.
JohnGarrett >> (All): I thought we had agreed at the high-level that we liked TRAC and would just make one pass through that in detail and add/change what we needed.  
                        I think we are 2/3 through that process now and perhaps should finish and make that a draft standard.
JohnGarrett >> (All): It would be good to address as many comments in one pass to avoid a lot of RIDs back on the draft that we would then have to address.
David Giaretta >> (All): John - good point. 
Helen Tibbo >> (All): I agree with the gist of this AM's comments - if a standard is too detailed, at least in this point in the history of this type of effort it will be to hard to follow.
Katia Thomaz >> (All): high-level first and low-level after
Don Sawyer >> (All): Regarding the issue of missing things that are in fact elsewhere, this may indicate the need for changed organization or better overview.
David Giaretta >> (All): Bruce - I guess we have got bogged down just recently - a human failing - certainly for me!
Helen Tibbo >> (All): Yes, I can see a "new and improved" version coming out in a few years after this has had a test run.
JohnGarrett >> (All): Are we saying that the level of detail in TRAC is wrong?
Don Sawyer >> (All): I don't think anyone is say trac detail is wrongl
JohnGarrett >> (All): I also believe that we can come out with a better document in a few years but there is no guarantee anyone will have resources in couple of years to lead an effort to update it.
Mark Conrad >> (All): Don - You need the 100,00 ft. view to reorganize the document.
David Giaretta >> (All): So where do we  go from here?
David Giaretta >> (All): Should we press on at the 10000ft level and focus on where we have comments
Don Sawyer >> (All): I would agree with john - we should complete the current pass.  Then step back and look at it from multiple perspectives.  Are we really bogged down?  
                              If so, set a schedule to complete sections
Mark Conrad >> (All): Bruce?
Don Sawyer >> (All): press on
Helen Tibbo >> (All): I actually think we have done quite remarkably with the megameting but a face-to-face always gets more done. 
                              Also, very specific assignments with a time table is critical expecially if we are not meeting face-to-face. 
David Giaretta >> (All): I don't think we have much left to do if we stick to the high level and the points already flagged on the Wiki
BruceAmbacher >> (All): I was not on the last call, where are we in the document?
David Giaretta >> (All): We should be starting B4 I think
Katia Thomaz >> (All): i have a question: what is really the final result of the BOF?
David Giaretta >> (All): IT looks to me as if the introductions of the subsections and alsosome detailed comments on B4.4, B5.1, B5.4 
JohnGarrett >> (All): Well the BOF should be converted to a Working Group which then creates a standard.
Katia Thomaz >> (All): and the final resulta is....
BruceAmbacher >> (All): My hope would be an improved  document that blended TRAC with th generic parts of nestor that could be the first draft to be circulated for standard review
JohnGarrett >> (All): A CCSDS (and ISO) Standard
David Giaretta >> (All): Katia, the final result is something that goes into ISO via CCSDS
Katia Thomaz >> (All): a high-level document or a etailed document?
Katia Thomaz >> (All): detailed
BruceAmbacher >> (All): similar level of detail
David Giaretta >> (All): Katia - a document on which we can build a certification process
Katia Thomaz >> (All): the final standard?
David Giaretta >> (All): ...and from ISO27001 it looks as if it can be very high level
Katia Thomaz >> (All): is it possible to see  examples from other BOFs?
Helen Tibbo >> (All): Who/what organization would do the audits and certifying?
BruceAmbacher >> (All): example is OAIS
Katia Thomaz >> (All): but was OAIS produced in a BOF
Don Sawyer >> (All): However the level of detail of the document needs to be whatever we think it should be
David Giaretta >> (All): Again - we should learn from other certification standards - but it has to be some international body which can first of all accredit organisations
Katia Thomaz >> (All): we need to know where we are going to...
Mark Conrad >> (All): See  Working documents    * RAC-Charter-20070119.doc: Charter agreed at BoF 19 Jan 2007 
Katia Thomaz >> (All): as i´ve never been in such a work i am very confused...
BruceAmbacher >> (All): Is that type of schedule usual - presenting a document that will be amended with more details?
Katia Thomaz >> (All): about the results
Katia Thomaz >> (All): i want to see a template or am example
Don Sawyer >> (All): Again, I think we need, at some point, to look at the certification document as though we were doing the certificaiton.  
                        Does the evidence give what is needed, for example?  
                        We need to give it some test runs, as it where.  Here are NSSDC, we are attempting to state how we would attempt to meet the requirements.  
                        I think if a few archives do this, we can get better insight into whether the document at that point is reasonable.
David Giaretta >> (All): Bruce - we would expect RIDS on any document we submit
Helen Tibbo >> (All): TRAC and DRAMBORA are both undergoing testing of one type or another. Do we envision that phase in our process? 
David Giaretta >> (All): Helen - we never did see any results from the TRAC etc trials
Don Sawyer >> (All): I hope so
BruceAmbacher >> (All): RIDS are one thing, ouradding to the document (as planned) is another.  I am not objecting just asking if there is precedent.
Katia Thomaz >> (All): he word "deliverables" is better than "results"
JohnGarrett >> (All): Examples of the results could be any ISO Standard.  I think if we look at other certification standards from ISO like ISO 9000 series, ISO 27001, etc.
David Giaretta >> (All): Yes, we looked at ISO27001 in some detail
Don Sawyer >> (All): I don't think a very thin document will work:  I think the current TRAC level of detail is about right until testing shows otherwise
Katia Thomaz >> (All): so we must produce a final version of the standard
Katia Thomaz >> (All): not a draft...
JohnGarrett >> (All): Looking at a schedule on the wiki, we are generating an internal draft by March.  Hopefully we are ahead of schedule on that.  
                        Then we have another year in this group before we go out for public review.
BruceAmbacher >> (All): The big blend was to address nestor and the CRL, etc. crosswalk
David Giaretta >> (All): Yes
Don Sawyer >> (All): Approval by ISO means it goes out for formal review.
David Giaretta >> (All): And getting RIDS
David Giaretta >> (All): ...which may lead to some changes
BruceAmbacher >> (All): Yes
JohnGarrett >> (All): Process is that we develop an internal draft for group, then work that draft until you get a document with technical consensus from the group.  
                         We then send that draft for public and formal reviews by CCSDS Agencies.
David Giaretta >> (All): Shall we dive into B4.4?
David Giaretta >> (All): ---------------------------------
BruceAmbacher >> (All): Yes B4.4
BruceAmbacher >> (All): I was reading the crosswalk and do not see any issues
David Giaretta >> (All): Katia has suggested a change - adding some text to the metric, obtained from Nestor - this is from Katia's analysis
David Giaretta >> (All): See
David Giaretta >> (All): Bruce - you are looking at Robin's cross walk
BruceAmbacher >> (All): yes
Don Sawyer >> (All): Katia wrote: " It lacks “repository actively monitors authenticity, storage and readability of archival objects” 
                               (see NESTOR CCTDR 7.2, 10.3)".  I think these attributes are mostly covered.  Perhaps we should address each of the three to see that they are there.
BruceAmbacher >> (All): Should all 3 be addressed here in B4.4 or separated?
David Giaretta >> (All): The supporting text mostly focusses on Fixity - Katia's addition would add authenticity
JohnGarrett >> (All): I think it might be easier to separate them.
Mark Conrad >> (All): I believe authenticity is addressed across several other bullets
David Giaretta >> (All): Mark: Maybe a case for cross-referencing if you can identify those metrics
Don Sawyer >> (All): Authenticity is address during the ingest process.  It seem to me that once that is accomplished, it is a matter of maintaince and the content.
Don Sawyer >> (All): maintenance of the content..
BruceAmbacher >> (All): Agreed
David Giaretta >> (All): The logic was: have a plan, implement it, preserve the important thin then monitor
Mark Conrad >> (All): You have to monitor authenticity every time you change the object (e.g., migration).
Don Sawyer >> (All): This section 4 is dealing with maintenance of the content.
Helen Tibbo >> (All): This is all pretty complex and those auditing and those being auditing are facing this for the first time. 
                        I think cross referencing of fundamental concepts/functions/activities is essential to making all of this reasonably clear and complete.
David Giaretta >> (All): Mark, Don: Is the question - what should be monitored?
Helen Tibbo >> (All): I know we are dealing with writing this now but I keep thinking ahead to how people are going to use it.
Don Sawyer >> (All): Let's see where transformations are dealt with
Don Sawyer >> (All): I also agree with Helen  - we need t hinkg about the 'users' of this
David Giaretta >> (All): Helen - yes we have done a bit of cross-referencing to help in that way but we have not been 100% consistent
BruceAmbacher >> (All): The idea/need for cross-references is good to help navigate the document.  Looking at the raw criteria after the text also helps with this - moew barebones
Mark Conrad >> (All): Some aspect of authenticity is addressed in almost every single criteria in th edocument.
David Giaretta >> (All): bruce, it was also to help avoid confusion and the sort of discussion we are having
Don Sawyer >> (All): Section 4.3 specifically mentions transformation (migrations).  
Don Sawyer >> (All): Recall that we were going to do an index to help look across the document for key terms and concepts.
David Giaretta >> (All): DOn - even if nothing is supposed to have changed, should things be monitored?
BruceAmbacher >> (All): Yes - sampling or more on a set periodic schedule
JohnGarrett >> (All): Seems like spot checks would be done to ensure checksums didn't change.
Don Sawyer >> (All): Yes, the integrity of the objects needs to be monitored.
David Giaretta >> (All): Don - yes we planned an index  also, but we also started cross-referencing metrics
David Giaretta >> (All): OK - so maybe the action should be to see if we can come up with credible cross-references 
David Giaretta >> (All): ...and if not then we add in Katia's words
Katia Thomaz >> (All): david, again i say that people must read comments before the meeting and give their positions
Mark Conrad >> (All): I believe authenticity requirements are addressed elsewhere and do not need to be specifically referenced in 4.4.
Don Sawyer >> (All): I agree with Mark, but important concepts may need to be discussed at a high level to make clear how they are addressed
JohnGarrett >> (All): I suspect that authenticity is covered elsewhere.
JohnGarrett >> (All): Storage not changing the AIPs is covered by use of checksums.
Katia Thomaz >> (All): but in other aspects
David Giaretta >> (All): I think it was the monitoring of authenticity and readability that matters here
Don Sawyer >> (All):  .. so, it may be goo to say how authenticity is addressed.  Readability is something quite different.
JohnGarrett >> (All): by readability do we mean we can access it, in which case the fact that we can get it to run a checksum covers that 
David Giaretta >> (All): Good point
BruceAmbacher >> (All): If the repository has followed the suggested steps to this point they have stablished authenticity.  The monitoring process will ensure it continues.
JohnGarrett >> (All): if by readability we mean understandability, then that is covered elsewhere in ingest and preservation planning to ensure it remains understandable over time.
Don Sawyer >> (All): I Think that 'readability' in general is much broader and goes to 'understandability'.
Helen Tibbo >> (All): Understandability has to do with what context and metadata and related files are saved and referenced
Mark Conrad >> (All): I say again, I think Katia's issues are covered elsewhere in the document and should not be repeated here.
David Giaretta >> (All): John- I think you are right - monitoring understandability is treated elsewhere
Katia Thomaz >> (All): we can get definitions from other digital preservation projects
Katia Thomaz >> (All): and put into the glossary
Don Sawyer >> (All): Yes
BruceAmbacher >> (All): definition of authenticity?
David Giaretta >> (All): OK - I'm being convinced that these points for B4.4 are covered elsewhere 
JohnGarrett >> (All): I think that's fine, does someone want to take an action to get definitions of readability?
Katia Thomaz >> (All): integrity, authenticity, storage and readability 
Helen Tibbo >> (All): I'll do that for next week
Don Sawyer >> (All): I'm not sure that 'storage' is needed, however.  Too generic,
David Giaretta >> (All): So have we agreed that we do not need to add words to the metric?
JohnGarrett >> (All): Agreed
Don Sawyer >> (All): agreed
David Giaretta >> (All): But do we need to add explanatory text?
Katia Thomaz >> (All): no.
Mark Conrad >> (All): agreed
Mark Conrad >> (All): no to explanatory text.
Katia Thomaz >> (All): definitions will clear our position
Don Sawyer >> (All): I think we need more input from Katia on her concerns
Don Sawyer >> (All): o'k
BruceAmbacher >> (All): So, what's the game plan for next week?
David Giaretta >> (All): Katia - I'm not sure what you mean - I agree with DOn
Katia Thomaz >> (All): NESTOR concerns
BruceAmbacher >> (All): nestor concerns beyond what is on the crosswalk?
David Giaretta >> (All): Looks like we look at what Katia writes about B4.4 then we go on to B5 intro and B5.1 then B5.4
Katia Thomaz >> (All): see NESTOR CCTDR 7.2, 10.3
BruceAmbacher >> (All): Both are on the crosswalk for B4.4 
David Giaretta >> (All): We should be able to produce the cross-references to show that these cross-walk concerns are already covered
David Giaretta >> (All): maybe the homework is for those who think B4.4 should not ne changed
BruceAmbacher >> (All): They are covered in that they were raised as covering the same topic as that part of TRAC.  We have not discussed the need, if any, to change language.
BruceAmbacher >> (All): nestor pints raised in the crosswalk
Mark Conrad >> (All): I am not sure that I agree with Katia's mapping of all three of the nestor requirements to B4.4
David Giaretta >> (All): Mark - does that come down to saying that they actually refer to some other TRAC metric?
BruceAmbacher >> (All): The crosswalk suggests they be addressed at this point in TRAC
Mark Conrad >> (All): David - yes
Mark Conrad >> (All): 7.2 should be mapped to B4.3 and elsewhere.
BruceAmbacher >> (All): Some of the nestor points are also mapped to other parts of TRAC
Don Sawyer >> (All): I agree with Mark: They are address in various parts of TRAC, so some extent.  So, I think the question becomes, are they adequately covered in TRAC?
Katia Thomaz >> (All): no matter if in another metric, but elsewhere
Don Sawyer >> (All): Unfortunately I must leave.
Helen Tibbo >> (All): Sorry for having missed so much this summer, but have we added in other nestor or DRAMBORA concerns along the way or primarily focused on 
                                improving the TRAC document first by itself?
Katia Thomaz >> (All): where?
Don Sawyer >> (All): bye
BruceAmbacher >> (All): I think it has been the latter.
JohnGarrett >> (All): I think we have been addressing the crosswalk concerns if participants have raised the issue.
David Giaretta >> (All): Helen - we had made some changes taking nestor changes on board I think
David Giaretta >> (All): ...but I'd have to check through the edits - Katia has been the main one putting in cross-walk comments
Helen Tibbo >> (All): So, might we not finish going through TRAC, come up with a clean draft of that to see where we are and then intensively look to see if 
                           other nestor and DRAMBORA concerns need to be addressed as a way of "checking" the completeness of our new document?
BruceAmbacher >> (All): I need to go.  Will someone post where we will start next week.
JohnGarrett >> (All): Yes, I think we come up with a clean version and then take a quick look that we covered all concerns.
David Giaretta >> (All): Helen - that second round cross-walking probably should be done - maybe by someone coming to it fresh
David Giaretta >> (All): SO - actions:
Mark Conrad >> (All): I would just like to see us complete one pass through the document. I we change the methodology every week, we will never complete a pass.
David Giaretta >> (All): Mark - agree - I'd like to get though it
Helen Tibbo >> (All): I agree, let's continue on with TRAC as is. It will be easier to deal with the other documents with a new clean copy.
JohnGarrett >> (All): I agree
David Giaretta >> (All): So - do we leave B4.4? or is there something to do?
David Giaretta >> (All): If not then its the intro to B5 and B5.1 and B5.4
Mark Conrad >> (All): See you next week!
JohnGarrett >> (All): Only thing I see now that is needed is to add a log of AIPs to evidence to show you have AIPs you're supposed to have and not extras.
Katia Thomaz >> (All): sorry i need to go. please pay attention on this (extract from the wiki): 
                           "The first step required by CCSDS procedures is the creation of a Birds of a Feather (BOF) group in order to judge the level of effort which is 
                             likely to be available for the writing of this standard. If, as seems likely, there is adequate support then a CCSDS Working Group will be created."
Helen Tibbo >> (All): This sounds good. I must run. I'll be connecting from a hotel next week but I think it should work OK. 
David Giaretta >> (All): Ok - bye all - I'll put the notes on the Wiki
Katia Thomaz >> (All): bye and have a nice week.
JohnGarrett >> (All): bye
Helen Tibbo >> (All): Bye

-- DavidGiaretta - 10 Sep 2007

Edit | Attach | Watch | Print version | History: r2 < r1 | Backlinks | Raw View | Raw edit | More topic actions
Topic revision: r2 - 2008-02-13 - KatiaThomaz
This site is powered by the TWiki collaboration platform Powered by PerlCopyright © 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback