World Wind 1.4.1 Notes

From World Wind Wiki

Jump to: navigation, search

Projects for 1.4.1 World Wind Code

Please treat this as a Brief Road Map and use JIRA for feature details and tagging commits.

[edit] Notes on 1.4.1's Goals

IMHO, we should have no more than 4 or 5 major projects to work on. We need to remain focused and not doing "this and that" just because. Minor fixes should be done after the major issues (unless we get more devs on board). 1.4's development and release really did become a garbled mess near the end and I would like to not repeat that again.

My ideas for the major issues to focus on:

  • Fix camera.cs see the irc transcript for more details, this is a major problem that must be fixed.
  • KML Support
    • KML Time (Integration with World Wind Time API)
    • KML Models (Collada)
    • KML Models (non-Collada) run much slower in WW (un-useable) also ugly wen compared to GE.
    • Robust network link and nested network links
  • 3D Model support / Planet9
    • Remote Model loading and https-auth
    • Occlusion culling of models
  • Downloader and DEM rendering
  • Time support
  • DSTile GUI

This is just my idea for the list, but I think that is we focus on these first and get them rock solid, we will have a tool that is of great use to a lot of people and bring more interest to World Wind for more application use.

[edit] Current Ideas List

What_Nick:

  • Dstile GUI - Possibly migrate DStile to C# with new gdal c# bindings
  • Better ModelFeature
    • Loading Models from a server, possibly with https-auth
    • Model shadow casting
  • More ingrained time support
    • Add-ons with time support (TomServo)
      1. Satellite Tracker
      2. Earth Quakes
  • QTS-T
    • Quadtilesets with time stamp
  • Use Nightlights textures as an emmissive map together with sun shading to give more realistic view of the Earth at Night.(Idea from Celestia)


Stepman (stephan mantler):

  • Download management (substantial changes)
    • already in progress, but terrain support is causing major headaches
  • More flexible rendering (experimental)
    • Extended shader support
    • Separation of texture & terrain data
      • Allow rendering of higher-res textures with lower res terrain data and vice versa (also see download management *sigh* )
      • addresses various rendering artifacts
    • Probably through a separate entity ("QuadTile2"), switchable at startup time? (i.e., loader allocates layer definitions as either QuadTileSet or QuadTile2Set)

Withak

  • File -> Load... gui
    • XML, zip, cs: pass to quickload
    • image (jpg, png, tif, etc.): launch image overlay plugin? add xml save capability?
    • shp: send the shapefile loader
    • kml: send to kmlimporter
    • model (x, dae): add to LM, add xml save capability?
    • misc. data (csv?, xls?, ???): pass to one of the icon-creating plugins?
  • Reorganize menus?
  • Layer manager restructure to allow renderable reordering/addition/deletion
    • maybe make the reordering permanent
  • WMS Browser
    • add better dimension support
    • manually add parameters to the url


Selar

  • Documentation
  • KML Support - use the standard Google XSD and support all the features that have a renderable alternative in World Wind. Most features do. (Selar, thanks for letting us blame this one on you!)
  • Unit tests for code - MBUnit suites


HamletAlive

  • KML needs to work, and work well
  • Shapefiles should be a no-brainer
  • Tile downloading needs to be flawless
  • Access to local cache *offline* needs to be flawless
  • Documentation (Yes but who? Who's the mythological hero that will do this thankless yet divine task?)
  • DSTile GUI
  • Collision (Why do I have to say this?)
  • Physics engine :) (Sorry, just had to say it.)

TomServo

  • Documentation - working on that.. have two WW books in the planning stages. NOT giving an ETA though. :P
  • Local Tile Cache Speedup - Ok, this one has been a thorn in my side for sure, as well as others. But Local cache MUST take first place over downloads. Even on a fast connection, what you alread have downloaded can take a while to resolve.. as you wait for unrelated imagery or DEM data to download. Maybe this will be fixed with Step's download code. But it is something that we need to look at and fix though.

[edit] See also

Personal tools