From mboxrd@z Thu Jan 1 00:00:00 1970 From: Lucio De Re To: 9fans@cse.psu.edu Subject: Re: [9fans] changes in 9load Message-ID: <20030521194706.E7647@cackle.proxima.alt.za> References: <20030521192502.C7647@cackle.proxima.alt.za> <90b20611bddbc6da6a3a70d5d136263a@plan9.bell-labs.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <90b20611bddbc6da6a3a70d5d136263a@plan9.bell-labs.com>; from Russ Cox on Wed, May 21, 2003 at 01:35:17PM -0400 Date: Wed, 21 May 2003 19:47:06 +0200 Topicbox-Message-UUID: b6aaf7e6-eacb-11e9-9e20-41e7f4b1d025 On Wed, May 21, 2003 at 01:35:17PM -0400, Russ Cox wrote: > > What would a rw /boot mean? Changing it certainly > wouldn't change the kernel image (wherever it may > have been picked up from), so you might as well use > ramfs if you want an in-memory rw file system. > In an embedded system, for example, you could release the memory allocated to modules you no longer need. You'd certainly not want that reflected in the permanent storage image, but it could make a significant difference where memory is at a premium. That's really the only purpose I could imagine. Being able to change data in /boot sounds like great complication with very little gain. ++L PS: Often the argument that memory and computing cycles are cheap resources is raised here and elsewhere. My take is that this is certainly the case, but that making efficient use of resources is a better investment than relying on low cost. The difference becomes significant when utilisation soares and the inefficiencies bring a system to its knees. IMHO reliance on low cost leads to environmental waste, specially in an economic sense. Not that I don't understand the danger of premature optimisation :-)