August Documentation Telecon 2014

1)      Follow-up for ESIP summer meeting

·         Went very well with Documentations sessions – they were integrated more into the rest of ESIP

2)      Improving metadata compliance for earth science data records (Ed/ Oliver Chang)

Automating Metadata Compliance Checker

·         CF Checker – console text dump – opaque –“this might be wrong”

·         THREDDS UDDC – detailed, color, descriptions

·         GDS2 Validator – checks for existence/type – lots of text

o   Developed for GRIST – it has a very specific purpose

·         IOOS Compliance checker by Dave Foster – this was used as a model – has detailed messages

·         THIS MODEL

o   Checks CF, ACDD, IOOS

o   CF portion uses IOOS as a black box but rewritten ACDD and GDS2 checkers

o   Checks are designed in data driven way

o   Uses python dics – hierarchical

·         Example – GRIST NetCDF4 file – uses all 3 checkers

o   Color coated – green = complete, yellow = partially complete, and red = fail – also has symbols for color blind people

o   Found problem – one part failed because it wasn’t comma delimited but it was a hierarchical list

·         If you print the report it will expand all categories

·         Uses ACDD 1.1 – this is described on the main page and includes links to the standards used (and the version)

·         GDS2 – LP2 file – also check attributes

·         Has a restful API interface to check files programmatically in JSON – can get HTML output

o   Designed so you can build a web service that people can add parsers

·         Ted and John G both looking for an external filet hat explains how the test is conducted – right now it is just a name and description

·         Ted – suggests separate NUG (NetCDF User Group) from CF

·         Oliver goes back to school … this will freeze the project

·         Ed will work on getting more information out – power point and a few documents will be up on the wiki in the next few weeks, in about a month will have the product released (announced via email)

·         Possibly have a winter meeting session about compliance checkers

3)      Prepare the summary issues for ACDD – (John G.)

1)      Data_*attributes

2)      Summary metadata for geo-spatiotemporal ranges

3)      Organization of ACDD – still some confusion

·         #3 – Phillip Jones – version 1.0 is not up on unidata any more – can it be brought into ESIP wiki

§  Anna – created an ACDD action to do this

o   Ajay (NODC) – do not need to use attributes? – will that affect UUD – Anna – need a new scoring for new version

§  Will that be bundled into a new THREDDS release

§  John is not sure about uptake after approval

·         #2 – add warning text – vote – no objections (DONE)