Discussion JPorter/RNottrott on minimum viable systems for LTER Data catalog

As a starting point in LTER NIS design/implementation, we discussed four systems that could be used to implement a successor to the existing LTER Catalog of Core Data Sets, 1) table-based, 2) Harvest-Web page retrieval and indexing system, 3) LTER Catalog entry and 4) Use NASA Global Change Master Directory.

1. Table-based
   1. link Hartman table to actual data
   2. query capabilities
      1. none
   3. directory tree capabilities
      1. requires selection of suitable data categories and consistent coding
      2. one dataset might fall into multiple categories
         1. can be done using multiple entries in the table per dataset
   4. ease of establishment
      1. can be done by one person over network
   5. ease of update
      1. difficult, requires re-accessing WWW sites individually
      2. could be semi-automated using Webcrawler-like software to detect new URLS
2. Harvest-Web page retrieval and indexing system
   1. query capabilities
      1. free text only
   2. directory tree capability
      1. none
   3. ease of establishment
      1. not clear at this point (to be explored)
      2. may be automated
      3. central site accesses site harvester
   4. ease of update
      1. automated
   5. RNott and KBaker have looked at some existing Harvest systems
3. LTER Catalog entry (direct successor to the Core Data Set Catalog)
   1. create our own system focusing on very minimal metadata (catagories, title of dataset, authors, URL(s))
   2. query capabilities
      1. text search based on title and keywords
   3. directory tree
      1. based on catagories/keywords
   4. ease of establishment
      1. requires sites to input
      2. or could be done by central site where sufficient metadata is online
   5. ease of updates
      1. requires manual updates by sites
      2. could be facilitated by WWW forms
   6. comments
      1. need to make sure keywords etc. match up with other similar systems (e.g. GCMD)
4. Use NASA Global Change Master Directory (GCMD)
   1. work with GCMD on developing modified data input and query forms that give the appearance of a stand-alone LTER system (Porter and Nottrott have worked on this with NASA in the past. As a result, present LTER Core Data Set Catalog entries are in GCMD)
   2. query capabilities
      1. both full-text and fielded searches
   3. directory-tree
      1. use facilities under development with GCMD
   4. ease of establishment
      1. would require sites to input appropriate descriptions, either via automated creation of DIF's or using WWW or PC-based forms
      2. GCMD would provide curation in the selection of themes, topics and terms
   5. ease of update
      1. sites would need to submit updates when datasets were added or changed
   6. Comments: requires cooperative relationship be developed with GCMD