offtopic: anyone the least bit interested in software development should watch this video. It's 1 hr, but even in the first 10 minutes you'll see things to blow your mind and excite you about could/should be possible for building real things in real time: flowingdata.com/2012/02/20/…

re: relative performance of qgis and arcgis (gis.stackexchange.com/questions/20403/…)

Today I merged 30mb of polygon shapefiles in 351 files into one shapefile in each of ogr2ogr, qgis, and arcgis (arccatalog Merge toolbox).

I was expecting ogr2ogr to be 3-5x faster than arcgis (v10), and was wrong. Here are the numbers:

the difference and sample set is too small to derive meaningful conclusions. Maybe someone has the time to do better(?) If so, here is an easy source of public shapefile data to use: ftp2.cits.nrcan.gc.ca/pub/canvec/50k_shp