v1.0.11

Changed

• Fixed support for bad outStatistics params in the URL

v1.0.10

Changed

• using mapnik-pool for all tiles now
• refactored tiles for better vector tile buffer/edge support in the mapnik
• PBF tiles are now deflated via zlib, which makes them smaller
  • providers need to set the header content-encoding: deflate

v1.0.9

Changed

• splitting objectIds for feature service queries
  • must be a string, not an array now (as per the geoservices rest spec)

v1.0.8

Changed

• fixed an issue with object id fields in feature service queries for Ids with filters
  • when using returnIdsOnly the objectID was not getting set, which return false rows.

v1.0.7

Changed

• lib/Tiles.js now passes along an optional name param to the tile generator. Vector tile use this name to store features.
• fixed the tests, they are passing but jshint is not.
• changed the way the lib/BaseController.js sends back its statuses.

v1.0.6

Added

• a flag to the lib/Exporter.js class that will lock export jobs to prevent duplicate jobs stepping on each other in the queue.
  • This means that one job per dataset can be enqueued at a time.
  • Down the road this may cause problems for when we want to pre-cache several formats for the same dataset at the same time. That bridge will be crossed at that time.

v1.0.5

Added

• a flag to the lib/PostGIS.js cache too ignore the selection limit unless a provider wants to use it.
  • since this feature is used by only a couple providers it seemed better to make it an opt-in option for the few, rather than an opt-out option across every provider.

v1.0.4

Added

• adding an ability to override koop's data limit of 2k.

v1.0.3

Changed

• Fixed a bug in the ExportWorker where shps were being created as a directory with an extension .shp; this fixes some formatting errors that were related.

v1.0.2

Changed

• Koop exports now force OGR to create shp file dirs instead of files; This adds more consistency so the code can maintain a single way for creating zip exports
• The query support for outStatistics now removes any passed in slashes so it stops choking on parsing semi-bad inputs
• The large data limit was moved from 10k features to 2k. This helps koop when its deployed with workers in that more work is handed down to the workers, which is good.