On Sep 24, 2011, at 12:34 05, Phil Ashby wrote: > I recently read through the code for rdimport (for different reasons!), > and was wondering if this use of temporary disk space could be avoided by > pipe-lining the intermediate conversion? As a by-product it would also run > faster on a multi-core machine :) It's adding thread complexity however...
In theory yes, but only if no normalization is being employed. My experience has been that normalization is used the overwhelming majority of the time, so I'd be reluctant to muck up the overall design of the import/export system (which is already quite complex enough) for the sake of what amounts to an edge case. Cheers! |-------------------------------------------------------------------------| | Frederick F. Gleason, Jr. | Chief Developer | | | Paravel Systems | |-------------------------------------------------------------------------| | The real problem with hunting elephants is carrying the decoys. | | -- Anonymous | |-------------------------------------------------------------------------| _______________________________________________ Rivendell-dev mailing list [email protected] http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
