Startup from custom-burned OS X CD |
September, 03, 2002 11:56 AM |
cmliljes |
Hello Ryan (Anyone else with suggestions please chime in, too)! I prefer installing from a custom CD (or custom partition, which can make for a very speedy installation process, but see below***). But now, from the release notes for 2.2 (August 23, 2002), we read: > The changes since version 2.2b17 are as follows: > ... > Only allows the "install" button to be used with the Install CD > (not "restart"). 1) Has this truly been disabled? (haven't tried 2.2) 2) If, so, why? 3) If so, when might it come back? :-) ------- *** Also (slightly OT), one feature that has never been implemented is the enabling of the G3 or G4 L2 cache during installation. Any hope for this? (My Supermac S900's soldered-on 512K L2 cache recently went kablooey, thanks to my own less-than-expert soldering "skills" while attempting the "J38 Jumper Dance." So I now have NO L2 cache at all until OS X is installed and up and running and your L2 Cache Enabler [or PowerLogix's utility] kicks in.) |
. |
RE: Startup from custom-burned OS X CD |
September, 07, 2002 4:31 PM |
OSXGuru |
. |
It has truly been disabled. The reason is that if XPostFacto can't write to the startup volume, then it can't ensure that the correct kernel extensions are installed, or that the custom BootX will be used. (Actually, I suppose it could check to see whether the correct kernel extensions are installed, and I may implement that for version 3.0 of XPF). I suppose the most important reason I did it was this: in most cases, using the "Install" button with a CD will result in a kernel panic, and that will result in a support question, which will result in me having to take time to figure it out and/or explain it. So I guess I'm being lazy :-) In your case, when you are installing from a custom partition, you will still be able to use "Restart". The reason is that the custom partition is writeable, and that is what XPostFacto cares about. When you are installing from a custom CD, you will need to use "Install". But it shouldn't make much difference as far as the speed of the install goes. On the "L2 cache during install" issue, I am interested in doing that. The complication is this: I either need to pass in the correct L2CR value as a parameter in NVRAM, or I need to determine it on the fly. Determining it on the fly is possible, and I have some ideas about how to do it, but it will be time-consuming to code properly. (There may be a short-cut possible there, but we will see). Passing the correct value in via NVRAM is problematic because space in NVRAM is tight for the install in the first place. I hope to be able to report some progress on that eventually. |
|
|