Hi Neil and others,
As requested some cost and time estimations for the GBIF bid:
- 3 days: Internationalization: some (minor) adjustments to ensure that the
database
is suitable for international use
- 2 days: Upgrade to support NCD 0.4 (minor work)
- 3 weeks: Add some import and export options. Perhaps a few specific ones,
like EAD
and NoDIT import and/or a more generic one where a database can be mapped
against the NCD schema, like in DiGIR or PyWrapper.
- for Moore bid, not for GBIF bid: Add RDF functionality
- 2 weeks: Create an installation executable for easy installation on
multiple
platforms
- 2 weeks: Create documentation for usage and installation (English only?)
- PM (matched with NLBIF funding): Establish a helpdesk and maintenance
(this could perhaps be matched with
NLBIF money)
- 2 weeks for testing and reporting
This makes 10 weeks in total. Our normal tariff (ETI) is Euro 81,50/hour,
but for for these developments you can take Euro 50,- (cost price).
This makes a total of 20K Euro (10*40*50). Perhaps some external testing by
other NCD members would also be useful?
Wouter
----- Original Message -----
From: "Neil Thomson" <N.Thomson(a)nhm.ac.uk>
To: "Wouter Addink" <wouter(a)eti.uva.nl>
Sent: Friday, November 03, 2006 11:55 AM
Subject: RE: Report back from TDWG 2006
Hi Wouter,
Many thanks for your helpful comments - I have forwarded them to the RAVNS,
since it looks as though only members can send messages to that list.
I will get the TDWG mailing list up and running so that this hassle goes
away.
Yes, I would indeed like some help in making cost and time estimations,
since this will need to be part of the bid.
Thanks again,
Neil
-----Original Message-----
From: Wouter Addink [mailto:wouter@eti.uva.nl]
Sent: 02 November 2006 13:42
To: Neil Thomson; rlg-nhsc-collection-description
Cc: Markus Döring (E-mail); Larry Speers (E-mail); marc(a)eti.uva.nl
Subject: Re: Report back from TDWG 2006
Hi Neil,
thanks for your report. I will limit my comments to the GBIF bid here,
because NLBIF will play a role in that. To make the NLBIF database and
editor available and usefull for GBIF and GBIF nodes, we need to achieve a
few things, which could be done under the GBIF bit.
- Internationalization: some (minor) adjustments to ensure that the database
is suitable for international use
- Upgrade to support NCD 0.4 (minor work)
- Add some import and export options. Perhaps a few specific ones, like EAD
and NoDIT import and/or a more generic one where a database can be mapped
against the NCD schema, like in DiGIR or PyWrapper.
- Add RDF functionality (not sure what is needed for the LSID support, yet,
I suggest to put this part under the Moore bid.
- Create an installation executable for easy installation on multiple
platforms
- Create documentation for usage and installation (English only?)
- Establish a helpdesk and maintenance (this could perhaps be matched with
NLBIF money)
Creation/Fill of the central database and GBIF index are tasks that should
be done by GBIF, I think?. We should take in consideration with this that it
will never be the case that all nodes adapt the NLBIF database (we need the
two-way solution as discussed during the workshop).
I don't know if GBIF would like adding OAI-PMH functionality, but I like the
idea and some harvesting option is needed for GBIF anyway. Implementing the
full specification however would be a conciderable amount of work. Since
there exist a few implementations already, we could perhaps incorporate an
existing implementation. I think OAIbiblio could be a good canditate. We
should research these possibilities first, if it is problematic to include
an existing OAI-PHM implementation, we could also concider to implement only
OAI-PHM export functions for the harvesting by GBIF.
For GBIF Nodes it would probably also be beneficial to add some
(REST-)webservice functions, in order to enable institutes to re-use their
information stored in the nodes database on their websites. This is probably
possible to do with NLBIF money as NLBIF also want to implement such
functions in the NLBIF portal.
Neil,
do you need some estimations of costs or other information? Any idea about
the available budget for the GBIF bid?
regards,
Wouter
----- Original Message -----
From: "Neil Thomson" <N.Thomson(a)nhm.ac.uk>
To: "rlg-nhsc-collection-description"
<rlg-nhsc-collection-description(a)lists2.rlg.org>
Cc: "Markus Döring (E-mail)" <m.doering(a)BGBM.org>; "Wouter Addink (E-mail)"
<wouter(a)eti.uva.nl>; "Larry Speers (E-mail)" <lspeers(a)gbif.org>
Sent: Wednesday, November 01, 2006 12:10 PM
Subject: Report back from TDWG 2006
Hi RAVNS,
Just a quick note to report back from the TDWG conference that took place
recently at the very splendid venue of Missouri Botanical Gardens.
The presentation, introducing NCD, seemed to go ok - no-one threw anything,
which I always take to be a good sign. NCD also got mentioned in several
other presentations throughout the week and is on track for becoming a
TDWG/GBIF standard.
The Workshop also went well, starting with an excellent presentation from
Wouter Addink from the Netherlands Biodiversity Information Facility (Dutch
national GBIF node) who explained how they had created a database and editor
from v0.3 of NCD. It looked very good and included the ability to use the
postcode of an institution to conjure up a Google map to show where it is
located. A small number of improvements to NCD are suggested by this demo,
which will be incorporated in the next few days.
This sparked off some good discussion about testing, data sources and how
NCD could be put to good use. There is also the offer of funding from two
sources, if we are quick. One is the second round of Gordon & Betty Moore
Foundation money for which a 2-page bid must be with Lee Belbin by 15th
November. The other is direct from GBIF via Larry Speers and that 2-page bid
must be with Larry by the end of November at latest.
For the long-term objective we came to the conclusion that we need to
accomodate both locally-hosted databases (based on the NLBIF system) which
could be harvested periodically by OAI - and the facility for those that do
not want to, or cannot, host their own data to provide it into a central
database. GBIF do not host data as a matter of principle, but they can host
centralised indexes.
There was particular interest in using NCD to handle organisation data and
part of the testing will include the Index Herbariorum data, which was
agreed at the June Workshop and is still on offer. There are many other
sources.
I would appreciate some help in producing the two bids, in particular to
confirm whether the suggestions that follow are feasible; whether we could
tackle them ourselves or would need to contract out and an idea of costs.
Here are some thoughts:
- Moore bid:
1. There is particular interest in documentation and use of identifiers. It
seems that a bid would be favourably received if it included investigation
into how to work with organisation identifiers, such as using their codens
for the unique part of an LSID and returning RDF as the result of a query.
Also, to determine how NCD will work with the emerging TDWG ontology - for
example in the use of picklists and controlled terminology required for some
elements.
2. NCD is also intended to link to biographical databases through person
identifiers and it would be valuable to work with IPNI (International Plant
Names Index) on developing the return of person information. IPNI author
data could be viewed as "Level 1" with BiogML as the fuller "Level 2" data,
both using the same LSID for the same person. This would give a valuable
triplet of collections, organisations and persons which could be drawn upon
as modules by other TDWG standards. I will contact Sally Hinchcliffe at Kew
about this possibility.
3. Create a really good set of documentation for NCD - Introduction / User
Guide / Data creation and migration guide / ...
- GBIF bid:
1. Determine what modifications are required to the NLBIF database and
editor to make use of NCD v0.4x and to add OAI-PMH export functionality, RDF
delivery and data import facilities (e.g. from EAD). This is with the aim of
being able to provide a toolkit to other GBIF national nodes and to
systematic / thematic organisations for them to record information about
their collections data.
2. Identify sources of existing published information about natural history
collections and relevant organisations (e.g. botanical gardens) that might
be suitable for conversion and inclusion in the overall system.
3. Employ someone to create records from 2 above and contact organisations
for records that are candidates and offer them the opportunity to amend or
otherwise update their information. This would require the modified NLBIF
database and a host for what could turn into the central database system for
those that would prefer not to host their own data.
4. Determine how to create the global index at GBIF from data harvested from
the various NLBIF-based systems
Over the next few weeks, I will be getting to grips with the new TDWG wiki,
mailing list, TYPO3 etc. and will transfer activity to that. I'll keep you
informed of progress. One bit of good news is that the new TDWG Process was
accepted by the conference. The direct effect on us is that we are no longer
tied to the TDWG conference dates for submitting NCD to the approval
process - it can be done at any time. There will be a full external review
of any standard put up for approval and information about how to go about it
and what is required is (or will be) on the new TDWG Website.
** Please do respond with your thoughts about all of this, especially on the
bids to Moore and GBIF.
I think that's all for now - many thanks to Connie and Doug as fellow RAVNS
that came along to support NCD at the conference,
Bye for now,
Neil
