I did some experiments [1] and am positively surprised (again) by the performance of the native compression libraries hadoop provides. What do folks think, should we include these also in next nutch release or do we just document how to use them and point users to get a hadoop distribution.
If we plan to include them it would be nice to include also version for windows, anyone capable of building those? -- Sami Siren [1] http://blog.foofactory.fi/