site_archiver@lists.apple.com Delivered-To: darwin-dev@lists.apple.com Dkim-signature: v=1; a=rsa-sha256; c=simple/simple; d=cs.duke.edu; s=mail; t=1265389356; bh=tKsieGK9SaXYaoAoYy+vdzqYZwHlg5nXzen2Ee3LMBU=; h=Message-ID:Date:From:MIME-Version:To:CC:Subject:References: In-Reply-To:Content-Type:Content-Transfer-Encoding; b=XH/XsikzEnRzYdNzCNseymeC6H7SsaXrNlnFfINCWbPF1BA+q72ejt+WNxEuMwG1g ocXC9ModRrdF2mdO0tFn8PVE+eGh9s3lD5HG2TAx0EDiJO1vbSvQAppHkUs2l3Y/IF 76lqutU6H+9Ot/LboxLtdOxPReb3iW2ZcpAfCgKI= User-agent: Thunderbird 2.0.0.23 (X11/20090817) I think the research moved in a different direction, and I don't remember there ever being an MMB paper. I did all the networking aspects, so I have no real knowledge of MMB other than what's in the cheating paper. There's probably more recent research, but this conversation made me remember MMB. Drew _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev@lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.appl... Jens Alfke wrote: So, apropos of my last post, I'm curious whether there's been any work/research/proposal on ways to enhance memory-mapping to address some of the problems Michael brought up. (I mean in any OS, not just Darwin.) We did something like this in a research group I worked for 12 years ago. See MMB in section 4.2.1 of "Cheating the I/O Bottleneck: Network Storage with Trapeze/Myrinet" (http://www.cs.duke.edu/ari/publications/cheating.pdf) This email sent to site_archiver@lists.apple.com