yes it goes to OS with one exception, I was trying to explain :) The exception is that file system is mapped with memory, and so the writes go to the memory fast, but they go to the file system as well however slowly. The trick is that the data goes to both memory and IO, they can be read immediately, and meanwhile you can lazily compute new data for new writing. This means the operations which compute data are blocked on the speed level of memory acccess and not the physical speed of file system. Therefore if you do it with chunks data, then your algorithm does not slow down. You don't need two Threads (one for computing and another for writing to IO), therefore single thread and ByteBuffer from NIO package does not need to be thread-safe. This is useful only if data computing really takes some time, and then the benefit is that all r/w are on memory speed. Then your application should have DMA access to that memory this way.
----- BR, tibor17 -- View this message in context: http://maven.40175.n5.nabble.com/Preview-release-Plexus-Archiver-multithreaded-Zip-edition-tp5822942p5823268.html Sent from the Maven Developers mailing list archive at Nabble.com. --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org For additional commands, e-mail: dev-h...@maven.apache.org
