Community Meeting 2005-04-29

From World Wind Wiki

Revision as of 17:50, 29 April 2005 by Dpatton (Talk | contribs)
Jump to: navigation, search

Meeting will be on Friday, April 29, at 20:00 UTC (1 p.m. PDT, 4 p.m. EDT) (UTC time converter)

Meeting will be held in #worldwind-meeting. Forum-wide announcement has been posted. No idea how many people will turn up, but I've set most of the "important people" to have Op flags so if needed we can set +m and control things. - Beanie.

Agenda Topics: Please, have your name after the agenda topic so we know who posted what.

PART 1 - Administrative / Policies / Procedures (1 hour)

  • Roadmaps / Ideal Release Dates (TomServo)
    • 1.3.2
    • 1.3.3
    • 1.3.4
    • 1.4
  • Bugs
    • ones that NEED fixing ASAP (TomServo)
  • Documentation
    • User Doc (TomServo)
    • dataset preparation (dpatton)
  • NASA Policies (dpatton)
    • It would be nice to have a "policies" section on the NASA website that explains any "policies" related to WW. For example, it could include a copy of the WW licence, but I'm mainly thinking of things such as the policy regarding "where data is hosted", etc.
  • Dataset Naming (dpatton)
    • "dataset" is used to avoid confusion with "layer". Aside from changes in the new rendering engine, the current WW "layers" already consist of two datasets - imagery and terrain data
    • Every "dataset" should have a unique "identifier". This applies to any dataset(collection of data) used by WW, such as a set of image tiles, terrain data, or other types of data such as placenames.
    • The namespace for the "identifier" could simply be a numeric value such as <DatasetID>01020003</DatasetID>
    • The namespace for the "identifier" could encode things such as the "world ID", "dataset type", etc., so 01020003 could mean Earth(01), Satellite Imagery(02), Dataset ID number 3
    • In addition to the "dataset ID", each dataset should also have a "short name" and a "long name" - something like this:
      • <DatasetID>01020003</DatasetID>
      • <DatasetShortName>BMBase</DatasetShortName>
      • <DatasetLongName>Blue Marble Base (2048x1024)</DatasetLongName>
    • Once all the above is established in the XML(without breaking the existing XML shema), developers can start retrofitting the use of "dataset" into the code, and use it in new code
    • The dataset ID should be the only value used in code to reference a dataset
    • The dataset ID should be used to name Cache directories.
    • If possible the "1.3.2 version for the WWC XML change" release should start using this new scheme for the Cache directory names
    • The dataset short and long names can be used in the WW user interface
    • The dataset short and long names can be translated to other languages without breaking any code because all references to datasets would be done using the ID(although the code/XML has to support language encodings properly)
    • http://worldwind URLs could be allowed to use either the ID or "short name" to reference datasets
  • Ways to get more Open Source community help
    • There has been a lot of constructive criticism on the forums of the new NLT data set, but nobody has offered to help with the 15m set, we need some way of encouraging actual help. (Bull)
  • Image Suppliers
    • Ways to entice them to open up some of their imagery (TomServo)
    • Could NLT create some documentation on setting up a WW compatible server, with example scripts etc. (Bull)
    • "Creation" of an 'Image/Data Master' for Image Supplies to contact? See this forum posting (TomServo)
    • I had suggested to Patrick a format for a 'can we have some sample image data' letter (dpatton)
      • My idea was to have this "official NASA letter" that could be sent to image supplies. If one goes for it, it may encourage other supplies.
      • I suggested asking for their best resoltion imagery, for an area of their chosing, on each continent
  • WorldWindCentral
    • The Wiki needs to be enhanced(if possible) so that when a person is directed to a page such as NLT_Landsat_Processing_Techniques they get a "breabcrumb" trail at the top of the page so they know what 'section' the current page belongs to
  • Other topics

PART 2 - Technical Discussions (45 minutes)

  • WMS/SVS Issues (TomServo)
    • 1.1.x compatability (TomServo)
  • Terrain Data
    • How to best fix terrain data errors (Rich Signell)
    • Status of new World Wind terrain engine (Rich Signell)
    • Problem with rendering engine not displaying negative elevation although the new elevation code allows it (Bull)
  • CVS Issues
    • Broken code in HEAD (TomServo)
      • Broken trails unless 0 elevation is used (Bull)
      • No chevrons (little arrows) (Bull)
      • Menu paint issue (menu items do not dissapear when a new menu is selected) (Bull)
    • Version splits (TomServo)
    • Other (TomServo)
  • Programming Documentation
    • Legacy Code (adamhill)
    • New code (adamhill)
  • dataset bounding box issues (dpatton)
    • what's the "right way" to deal with non-rectangular datasets, particularly when they overlap with other datasets
    • what's the "practical way" to deal with non-rectangular datasets, particularly when they overlap with other datasets(i.e. so such datasets can be used in the 'current' WW
  • dataset 'priority' (dpatton)
    • not sure how this relates to the "Z-order handling" in the new rendering engine, but there needs to be a defined mechanism to deal with overlapping datasets
    • this mechanism needs to deal with the 'priority' for datasets, and can't be tied just to what may be specified for a dataset in it's XML definition. For example, in an area with overlapping datasets, WW needs to select a default, but the user needs to be able to select to use the other dataset instead
  • Other topics
Personal tools