From grayzeck@chiron.astro.umd.edu Sun Aug 12 13:17 MST 2001
Date: Sun, 12 Aug 2001 16:17:46 -0400 (EDT)
From: grayzeck@astro.umd.edu
To: neese@psi.edu
Subject: Re: more advance liens
Content-Type: text
Content-Length: 4380

I see you are sending these review to Howard/Kopal.  For the XGRS data,
I thought that Tim was still gathering the changes, and then these are 
merged into what APL has produced?   I am attaching the file I created
for our XGRS review.  At this point, I only sent it to Tim.  Should it
now go to the APL folks as well as Tyler and Anne?


Review of XGRS datasets for the review

Note: first transmission to XGRS team on July 31, 2001; update on Aug 7, 2001

1. you have two distinct data sets (given by unique IDs), so a minimum
of two "dataset.cat" files are needed.   As far as I can tell, you attempted 
one for XRS level 2, but left off mention of XRS3; none was made for GRS2.
I am attaching a working example of XRSL23DS.CAT (new name) but the content
needs to be checked by you.  Note I have taken pieces from the aareadme and 
xgrsdataset.cat files and left a number of items blank.   The form is correct.   
(In PDS we would identify the XRS2 and XRS3 by the same data_set_id as the 
files are in the same form except for sampling and organization - correct?)

You have some left over problems to correct the example I sent.  

A. Under Processing are all level 2 products identical to level 3?   If so, then
the sentence "The Level 2/3 data are then combined into integral products within a
relational database and exported as an integral Level-2 product." is correct.
This is what is meant by keeping the data_set_id the same?

B. In the Data and Ancillary Files, the listing is not consistent with what
is in the directories.   For example, DOCUMENTATION should be DOCUMENT,
XGRSDATASET.CAT should be XRSL23DS.CAT and XGRSDEFINITIONS = XGRS_DEFINITIONS.

C. You claim there is a REFERENCES.CAT file but it is no longer in the
DOCUMENT directory

D. You have a left over file XRS_REQS_7_2.LBL% file that should be removed.

E. You have an OBSLOGINFO.TXT file not listed as Contents (OBSERVATION_LOG).

F. You have the same problems in the XRS3 listing; in addition, there is no DATAINFO.TXT file for the DATA directory

G. under FLARE, you should indicate that there is an "XSUMMARY.TXT" file

H. under the XRS3 listing, GRS should be changed to XRS.

2. You need to construct a separate GRSL2DS.CAT file in a similar form.

The above problems B-F are true in GRS dataset description.

3. All the CATALOG names should be less than 8 characters

********************************** update *********************************

4. The data_set_id in the voldesc.cat file should indicate XRS or GRS data.

5. The AAREADME.TXT file for the GRS 2 data (nxeros_2001) has a volume
listing which is for the XRS data.  Although the data is structured the
same, the directory names start with a "G" for the date (not X as given).
The data file extensions are DAT and not XDR as given in the AAREADME.
(This latter problem appears in the AAREADME in all nxeros volumes.)

6. All index file have the wrong data_set_id as forced by APL production;
the data_set_id needs to match that as appearing in the data.

7. All physical labels have the wrong Data_set_id; should match as above.

8. For XRS3 data (nxeros_2012) the Browse files are organized by date
alphabetically, while the FLARE directories are given as a digital date.
There should be a consistency; digital will sort correctly especially
if given in the year/month/day arrangement (e.g., Feb 12, 2001 -01-02-11).

9. For all data directories the organization is by date alphabetically,
yet the directories for a given day are given digitally.  The above
inconsistency for the Browse files is also true.  We suggest digital. 

10. The many small files (each spectra of a given time stamp) could be
reorganized into a daily directory to match the Browse files (other data).
Such a change would still require an index into the table structure.

-------------------------------------------------------------------------

We have performed a spot review for the data in the XGRS 6 review volumes .

The labels and checks can be found at:
/deimos2/taylorp/scratch/XGRS

there is a directory for each cd (the dir. name is the volume number of
the cd)

Common warnings from lvtool and sbnlv are 3-5 statements about "Value is
not a standard value" or "Using PDS GENERIC object definition".  The
common error among the labels is the missing COLUMN sub-object for TABLE.
xcolumn.fmt lists the column sub-objects and their description.

-------------------------