VegBank

The short format should include:
Observation: AccessionDate (do we have this yet?)
Observation: AccessionNumber (I think this had been added, perhaps diff name)
Observation: AuthorObsCode
Observation: ObsStartDate
Project:ProjectName (-> Observation:Project_ID)
Party:givenname+middlename+surname (_> ObservationContributor:role_id=PI)
Country [We talked of moving this to Plot from named place]
State/Provence [We talked of moving this to Plot from named place]
County/Parrish [We talked of moving this to Plot from named place]
Plot:RealLatitude
Plot:RealLongitude
Plot:ConfidentialityStatus
Plot:Area
Count of number of taxa recorded

The long form might consist of the above, plus the long summary view of the
plot you get after you search for a plot and then view it.

This is not definitive. Perhaps Mike or Don have ideas about what next.

I transfer this to ML for completion, with expectation that he will pass to
Gabe for implementation.

• Bug 777 has been marked as a duplicate of this bug. ***

the example output that I have thus far is here:
http://tekka.nceas.ucsb.edu/~lee/example_short_receipt.html
The xsl file that runs against the receipt.xml is here:
http://tekka.nceas.ucsb.edu/~lee/short-receipt.xsl (view source!)

I suggest that we drop county/parish from the list, as it's not in the receipt
xml, nor in denormalized format in plot table.

We do have accession number and date entered, which is what I assume you mean,
Bob, by accession date. The date in the accession number is the date accessed,
which is obviously dynmaic.

Gabe, we need to add the following fields to the receipt.xml so that they can be
styled into this form:
*project name (observation.project_ID -> project.project_ID, project.projectName)
*observation PI (observation.observation_ID ->
observationContributor.observation_ID where role_ID = (role_ID for "PI"))
*realLatitude and realLongitude (perhaps they are already being funnelled
there, but would be good to label them as such and include lat and long, which
would include any fuzzing of location)
*ConfidentialityStatus
**plot.area

Additionally, Gabe, I need some examples of XML receipts that I would get in the
event the plots had failed. In my example, I've only made it "FALSE" but we
need to return the error message to the user in cases where there are errors,
too. So we need an xml schema or dtd for the receipt.xml

the example output that I have thus far is here:
http://tekka.nceas.ucsb.edu/~lee/example_short_receipt.html

Looks good except that we have the WRONG format for accession numbers.
In our group minutes from June 2002 we asserted we wanted:
Rule: Accession # = VB-unique_obs_seq-YYYYJJJHHMM.
I think this could be split into two components, the unique obs seq AND the
date the plot was submitted.

I suggest that we drop county/parish from the list

OK

We do have accession number and date entered, which is what I assume
you mean, Bob, by accession date. The date in the accession number
is the date accessed, which is obviously dynmaic.

This statement is not clear.
Accession date is the date the data were accepted into the sysem, not the date
they were accessed.

Quoting Bob:

the example output that I have thus far is here:
http://tekka.nceas.ucsb.edu/~lee/example_short_receipt.html

Looks good except that we have the WRONG format for accession numbers.
In our group minutes from June 2002 we asserted we wanted:
Rule: Accession # = VB-unique_obs_seq-YYYYJJJHHMM.
I think this could be split into two components, the unique obs seq AND the
date the plot was submitted.

This is an old xml doc that I was using to test the stylesheet. The stylesheet
is all I'm working on right now. I agree that the accession number format is
wrong here and I believe it has been fixed by Gabe already. At any rate, there
is a bug for that.

We do have accession number and date entered, which is what I assume
you mean, Bob, by accession date. The date in the accession number
is the date accessed, which is obviously dynmaic.

This statement is not clear.
Accession date is the date the data were accepted into the sysem, not the date
they were accessed.

Your above statement is correct, Bob,

I think this could be split into two components, the unique obs seq AND the
date the plot was submitted.

This is done. There is a problem in the way that we think about this, however,
in that there is both a plot accession number (plot.accession_number) and an
observation one (observation.obsaccessionnumber). The latter is the only one
that means anything to me. Plot accession numbers are relatively useless. We
now track plot.dateEntered. But we probably need to be tracking
observation.dateEntered, which as far as I can tell, we are not.

However, dateEntered doesn't have anything to do with accession Number. The
accessionNumber is formed when someone accesses the plot. The date the plot (or
observation) is irrelevant to the accession code, which, just to beat a dead
horse, I will give an example:
I enter a plot last month, called "Mike's plot #2"
It gets a observation.obsccessionnumber of "VB.791" because it's the 791st
observation in the database. It gets a dateEntered value of "03-MAR-2003
11:43:21 GMT". I query this plot today, and in my XML downloaded stream, I
should read the accession number to be "VB.791.29-APR-2003 15:41:13 GMT" which
is the concatenation of the observation.obsaccessionnumber and the date I
accessed the plot in VegBank's system.

I would prefer not to use the phrase "accession date" as it seems to be
ambiguous. I prefer "date entered" and "date accessed" as unambiguous.

So, other than all that, I hear from you, Bob that this generally looks good and
I'll be working with Gabe when he's not doing more important things to get this
plugged in.

Bob's email response, r.e. accession numbers:
----------------------------------------------
I think the problem is in your usage of the term "accession". This would
generally be a fixed code generated by the institution that accessioned
the plot, which is unambiguously VegBank. The term accession should not
apply to the user. Accession date is exactly equal to the date vegbank
accepted the plot which is equal to the date the plot was entered. The
accession number is exactly equal to the VB.791 in the above example and
is NOT dynamic. We need a NEW and unambiguous term for the concatenation
of the plot identification number and date of extraction from VegBank. It
is really something like an Observation Export Identifier. Is there a
better term in common usage out there in the IT world? Gabe, do you know?

We should create a UserDataset for the plots that have just been submitted.
Through this dataset, the user can come back and see tha entire batch. The link
to that batch should be in the receipt. The link should go via the
accessionCode for the dataset.

need to also provide user with file that updates his/her VegBranch
accessionCodes with new VegBank accession codes so that they are in sync

This should be handled in the form of a link to users, showing them the stuff
they just added to vegbank, but referencing the dataset created when they added
stuff.

using the dataset created on load. # of plots, parties, plants, comms added,
also min # of taxonObservations per plot, max # per plot would be nice. (this
shows that plants were added to all observations). % of observations with
commClass records would be nice. Link to higher amounts of detail.

• Bug 1857 has been marked as a duplicate of this bug. ***

done, but still need finish dataset view so that link works.

Original Bugzilla ID was 780