Matthew Dillon wrote: > In otherwords, A meta-data journal does not slow things down in the > least. well, see Seltzer et al.'s Usenix 2000 paper for an analysis of this.