metronius

I dont think that you need so much pages - most of 150B pages will be never ever shown displayed at SERP.

If you would crawled web using Foreach(links) { get(link)} you would end up crawling useless automatically generated webpages.

I think so, spending one year in big company is really good experience. You will learn how to work in large teams.

Yes, i think that the debate makes no sense too - for me, Spark is no Hadoop competitor its rather complement.

Spark does not need Hadoop - you can run it also with Mesos or in local mode..

I mean M/R by M/R ;) http://vision.cloudera.com/mapreduce-spark/

1. Cloudera left M/R for Spark, Mahout left M/R for Spark. Spark community will be huge soon.

2. Yes, Spark was/is buggy.

3. For me Spark is really paradigm shift, next generation framework compared to M/R

What change the biggest difference in performance between Spark and MapReduce?

Do you mean SIMR or something another ?

I was considering backing but for me its pity that game is not for Mobile(Android tablet). I dont play games on PC (except Portal ;) )

"The dataset for this model was acquired within mere 45 minutes by a Pteryx UAV with maximum TOW of 5kg, flying 150m above ground level. The drone carried a consumer grade compact camera (Canon PowerShot S90) and took a total of 541 nadir oriented images. The resultant true 3D model demonstrates Melown Maps ability to capture façades even in the absence of oblique aerial imagery."

The paragraph is about another dataset, not the Brno city dataset.

from Pteryx UAV

https://www.google.cz/search?client=ubuntu&channel=fs&q=Pter...

If you want to go below nlog(n) during generic sorting you need to choose high information gain algorithm, its easy. This is no way.