From: C. Scott Ananian Date: Wed, 20 Apr 2005 15:52:02 GMT Subject: Re: [PATCH] write-tree performance problems Message-ID: In-Reply-To: On Wed, 20 Apr 2005, Linus Torvalds wrote: >> I was considering using a chunked representation for *all* files (not just >> blobs), which would avoid the original 'trees must reference other trees >> or they become too large' issue -- and maybe the performance issue you're >> referring to, as well? > No. The most common index file operation is reading, and that's the one > that has to be _fast_. And it is - it's a single "mmap" and some parsing. OK, sure. But how 'bout chunking trees? Are you grown happy with the new trees-reference-other-trees paradigm, or is there a deep longing in your heart for the simplicity of 'trees-reference-blobs-period'? I'm fairly certain that chunking could get you the space-savings you need without multi-level trees, if the simplicity of that is still appealing. Not necessarily for rev.1 of the chunking code, but I'm curious as to whether it's still of interest at all. I don't know exactly how far ingrained multilevel trees have become since they were adopted. --scott Japan explosion BLUEBIRD Honduras jihad D5 SLBM Diplomat overthrow JMTIDE CABOUNCE AMTHUG ESODIC Kennedy AVBRANDY CLOWER mail drop PHOENIX ( http://cscott.net/ )