On 11/11/2007, Aubrey Li <[EMAIL PROTECTED]> wrote: > On Nov 12, 2007 9:06 AM, Shawn Walker <[EMAIL PROTECTED]> wrote: > > On 11/11/2007, Aubrey Li <[EMAIL PROTECTED]> wrote: > > > On Nov 10, 2007 2:26 PM, Shawn Walker <[EMAIL PROTECTED]> wrote: > > > Indiana should install and run the system with the configuration as > > > follows: > > > > > > 1) 256M memory + GUI + X86 + Live CD-based + CD + suitable performance > > > 2) 256M memory + GUI + AMD64 + Live CD-based + CD + suitable performance > > > > > > Justification: > > > That's what we need for a distribution to be doable for our project. > > > > That needs further explanation. What "project" are you referring to? > > > > I should also note that almost every distribution I've found that has > > a native AMD64 environment consistently requires at least 384MiB of > > memory for a LiveCD. This is in part because of the code size being > > significantly larger for that architecture. > > > > As a result, I'm trying to figure out why and how you picked out > > 256MiB as a number. > > We are trying to see if solaris works on the classmate PC. See below > the configuration: > http://www.classmatepc.com/classmatepc-system-hardware.html > > 256M amd64 system may be a hard requirement, but 256M x86 system is > realistic IMO.
Note that the two operating systems they list as running are both 32-bit versions (as far as I can tell). Therefore my comment about running a 64-bit operating system still applies. 256MiB is not a "hard requirement" to support a purely 64-bit environment; 32-bit code runs just as well on and amd64 and with that much memory is a far more practical thing to do. > > > > I should also note that every LiveCD environment I've encountered > > (GNOME-based ones anyway) consistently require between 256-384MiB of > > memory. > > > > Ubuntu, notably, requires 384MiB. > > > > Suitable performance is outside the scope of this particular > > discussion unless you are using it drive the memory requirements, in > > which case, 256MiB is probably erring too low; especially considering > > Indiana is zfs-based. > > > hmm..., I didn't realize the footprint of zfs-based system is bigger > than ufs-based. > If it's not really doable, please ignore my requirement. It may be doable; I'm not qualified to say one way or another. However, I can tell you that zfs-based systems need more memory in some cases than ufs-based systems; or that they will increase the overall system memory pressure and create a experience that is different than a ufs-based system would for low memory systems. -- Shawn Walker, Software and Systems Analyst http://binarycrusader.blogspot.com/ "We don't have enough parallel universes to allow all uses of all junction types--in the absence of quantum computing the combinatorics are not in our favor..." --Larry Wall _______________________________________________ indiana-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/indiana-discuss
