Skip to end of metadata
Go to start of metadata

USGS CDI TSWG Meetings 2-22-2011 Meeting

There's a prettier Etherpad that you can get a link to back on the agenda page.

Some links related to the Catalog demo:

GI-CAT documentation:http://essi-lab.eu/cgi-bin/twiki/view/GIcat/GIcatDocumentation

ncISO:http://www.ngdc.noaa.gov/eds/tds/

Geoportal Serverhttp://gptogc.esri.com/geoportal/catalog/main/home.page

GeoNetworkhttp://geonetwork-opensource.org/

make sure someone from GIN attends GeoData 2011
GeoData 2011 Participant List: http://tw.rpi.edu/web/Workshop/Community/GeoData2011/Participants

MAD interacting with NGP WDIF - maybe NGP could help lead this group to guide thinking about how to structure their efforts. Matt Tricomi, John Hollister (CIDA)

SAD: get back to Falgout about who might be interested in this, build a stub of a wiki, get some names.

CDI BPDM group is spinning up three sub groups.
Scott outlined purpose of "Policy Development and Review Subteam": How to encourage data sharing and integration at the RGE/EDGE policy level. see https://my.usgs.gov/CDIWiki/wiki/BPDM#section-BPDM-SubTeams for more info.
-ask BPDM re: experience with Powell Center proposal process, do we need to focus on "research" more than "technology"
-Sky has been working with PC to get more thinking on geoinformatics (not just traditional "earth" sciences). Sky wants to keep geoinformatics proposals competing in the same pool as other PC projects.
Look at previous proposals-gotta do our homework. Look at current state of affairs that synthesizes what's really here and working.
-Proposal has to be clear on what the contribution to science will be. They're looking for major journal articles (like Nature, Science). Could address more informatics oriented journals (IJGeographic Information Science?)
Proposal shouldn't be "engineering"-while building something might well be part of the work, this is not the purpose
-think about what informatics means, and what it does for the Agency.

re: develop almost general proposal as an exercise to mature TSWG thoughts on what is important, maybe modify/tailor according to who we are shopping/pushing it to.

re: is there a different wavelength that we need to be on if proposing to CDI itself for funding?
-might separate "research" portion from technical implementation in our thinking.
-are there ideas that span different technologies that we want to understand or pick a solution?
-Sky: not really about "shopping" proposal around. Just get TSWG thinking together while PC figures their end out.
-Next PC proposal set is due at the end of June or July (can someone look this up?)

-Rich: was the idea a bad fit? Sky: not necessarily if it was more than an engineering type of project. Focus on what is the current state of cataloging. Why doesn't this ideal of a giant "catalog of catalogs"? Marty wants to know why we can't just have a giant cattle call of USGS databases. Some of this might be packaging.

-Rich: there are definitely a set of tools folks are using, a number of works-in-progress. Would be nice to get support to work out some of the tech problems. E.g. feeding the data to geodata.usgs.gov, etc? This kind of development would be a win for all involved.
-Sky: could you put time/energy into the development of something like this as a deployable product?
-Rich: This might be missing the focus. We have metadata issues (does our stuff fit into ISO standards, etc)? There's also training and education part.
-Greg: separation of technology from our development of domain-specific metadata hard to do because all the technology is evolving so rapidly. To be fluent enough to develop the metadata means being very committed and up to date on the technology. It requires efforts in both technology development and evaluation of how well your data can utilize this technology.
-Dave B: Current state of data integration is too formative to think about coalescing around a single "uber" solution. We need to work as a community to test, experiment, and educate. Very difficult to unify efforts with the decentralized highly varied landscape we live in. Yes, we are just looking for a bit of extra resource to support this distributed development of cataloging technology.

-Sky sees three models sounds fair:
-all of us as a community of practice doing our individual jobs; we value venue to share, communicate, etc (this is the status quo TSWG)
-EGIS model - people officially supported to be part of a tech support/consulting team that can be tapped to help folks.
-fund a substantial chunk of one or a small number of folks to provide support; when TSWG

  • which of these tech support models do we pursue? Sky has talked with Jennifer Sieverling. do we get CSS to help grow this group of experts.

-start small because this is not a mainstream business need. The environment is heterogeneous and evolving. Sky thinks that this is worth corporate support if TSWG can describe what should be supported (like his second and third support models, above). This could be put to both CDI and CSS. CSS in particular could be the organizational unit that sponsors this.

-research vs. technology: should we get tech support together first and then focus on research?

  • Greg is more into research, Rich: geoinformatics is the research for figuring out how to make these tools work together effectively and getting our science into the communities.
  • Sky asks of Greg: CMSP is similar to Geoscience Info Network (GIN). Should we collaborate with other groups to promote an open-source project? This adds complexity to the process but could reap benefits.
    • Greg would be interested. He is not doing too much coding, mostly about population of the pieces with metadata and a lot of configuration. Dave B. agrees that configuration is a HUGE portion of this world. Rich concurs:

Focus next meeting on Cataloging demos (Gi-Cat, GeoPortal Server, GeoNetwork, THREDDS/NcISO, CSW issues):

how about 10-15 minutes each from:
Blodgett
Miller
Signell
Bristol
others?

Get Rick Brown onto TSWG email list. He's doing GI-Cat.

NGP wants a slot the meeting after the 24th. Approximately, <=20 minutes for a demo. Greg Smocyk.

Tim will post a wiki page on Stan Smith's Data Modeling Standards proposal.