Difference between revisions of "Boston Sprint 2013 Agenda"

From OSGeo
Jump to navigation Jump to search
(Added GDAL driver tasks)
(Added another GDAL driver task)
Line 44: Line 44:
 
## Take care of locale (I needed to export LC_ALL=C to make it work with QGIS)
 
## Take care of locale (I needed to export LC_ALL=C to make it work with QGIS)
 
## Allow user to specify resolution strategy when dealing with different pixel size sources (maybe an option in connection string)
 
## Allow user to specify resolution strategy when dealing with different pixel size sources (maybe an option in connection string)
 +
## Allow user to specify the desired srid when dealing with different srids
 
# QGIS - good support for large coverage (overviews, irregularly tiled coverage), import export. What need to be enhanced?
 
# QGIS - good support for large coverage (overviews, irregularly tiled coverage), import export. What need to be enhanced?
 
# ST_CreateOverview()
 
# ST_CreateOverview()

Revision as of 15:27, 4 March 2013

Meeting times and locations can be found: Boston Code Sprint 2013

MapServer

GDAL

PostGIS / pgRouting

General Stuff

  1. When are we going to release 2.1. Steps remaining before we can
  2. SpGist -- when is that going to happen will make cut for 2.1 or push to 2.2
  3. How best to integrate Oslandia 3D work, concerns etc.
  4. Google SoC 2013
  5. Regression suites (performance, valgrind?)

Geocoding specific stuff

  1. Integrating Steve Woodbridge - address normalizer into PostGIS
  2. Packaging as part of windows builds at very least

pgRouting

  1. Packaging pgRouting for windows to be availabe in stack builder along side PostGIS
  2. Discussion of how pgRouting / PostGIS can better work together and melding the two projects
  3. Side note both Oslandia 3D work and pgRouting both use CGAL, possible opportunity/concern there ?
    1. I am looking into the possibility of drop CGAL and GAUL dependency to simplify builds
  4. pgRouting 2.0 release (working on building that plan now)

Raster

  1. Raster seems in darn good shape for 2.1 so least worried about raster. Great job :)
  2. Rework of MapAlgebra for 2.1 or push to 2.2
    1. Performance of expression-based mapalgebra
    2. Performance of ST_Clip
  3. GDAL raster driver - good support for overviews, irregularly tiled coverage. Some nice todo improvements:
    1.  Add write support
    2. Rewrite tests (gdal_autotest)
    3. More tests with Mapserver
    4.  Tests with QGIS plugin
    5. Deal with really BIG raster data in IRasterIO (recursion? split in several calls?)
    6.  Take default user, host, password
    7.  Review SQL strings (avoid SQL injection)
    8. Take care of locale (I needed to export LC_ALL=C to make it work with QGIS)
    9.  Allow user to specify resolution strategy when dealing with different pixel size sources (maybe an option in connection string)
    10.  Allow user to specify the desired srid when dealing with different srids
  4. QGIS - good support for large coverage (overviews, irregularly tiled coverage), import export. What need to be enhanced?
  5. ST_CreateOverview()
  6. Test performance of ST_Union() when number of pixel increase
  7. Define another (faster?) way to rasterize a vector coverage with ST_MakeEmptyTiledCovererage(), ST_BurnToRaster(geometry, raster) & ST_UnionToRaster()
  8. Discuss a more flexible way to rasterize a vector coverage
  9. Raster using pthreads in the future?
  10. Rewrite parts of raster2pgsql
  11. pgsql2raster prototype?
  12. Discuss "PostGIS Extra PL/pgSQL Function Set"
  13. Discuss interpolation - from some point, from millions of points
  14. Test regular blocking constraint
  15. Discuss viewshed
  16. Discuss integration of raster2pgsql into the GUI

PDAL