I have no idea how much the disk/network access impacts the boot speed.
Having it on a local 5 MB/s drive might well speed it up significantly.
Also I'm sure it's not having to transfer the entirety of the image, but
just the bits its using, so maybe not.
As for SoftPC, there were several 68k versions. I have one myself called
SoftPC Classic that emulates an XT-class machine on a Mac Classic. I've
actually run Windows 3.0 in Real Mode on a Mac Classic. It's...
something to see. If you really, really needed access to a DOS or early
Windows application it'd be usable but not much fun. It'll run on up to
a 68030 at least with the proper application of Wish I Were, but crashes
on a 68040. I suspect it would also run all the way back to a Mac Plus,
but it requires a Superdrive to emulate the PC floppy properly.
There were later versions of SoftPC for 68k that went up the chain of
PCs, including some variety of 286. Haven't tried those. As for CISC vs.
RISC, I'm not sure that really enters into it much when you're emulating
a chip that started at 233mhz and did something like 20-25 MFlops
(according to a quick Googling) on a 33mhz chip doing 4.6 MFlops (also
only according to a quick Googling).
My next goal is to recompile X from the ground up on the thing to see if
it helps with the crashes. I think the event pool is filling up and
overflowing the buffer, killing it, so it might take something more like
a kernel recompile.
Scott
On 1/10/2014 8:06 PM, Derek Morton wrote:
On Jan 10, 2014, at 6:28 PM, Scott Holder wrote:
Oh, there's no point at all beyond "Because I wanted to try". Just part of the
silliness I like to do.
The image is about 5GB. Not very much of it is used, I could probably slim it
down quite a lot. I'm mostly lazy. The drive I'm using is only 3GB total, and I
have 2GB devoted to Linux and 1GB to Mac OS.
As for the network speed, given the SCSI speed it might actually have been
faster over the network than locally. I haven't benchmarked the network speed
but I only get about 1 - 1.5 MB/sec out of it. I'm not sure if this is a factor
of the drive or the SCSI drivers in Linux since I believe it's supposed to be
able to get 5MB/s.
I don't believe network speed on an LC 475 is ever going to achieve 1 MB/sec...
That is around it's theoretical maximum (10 Mb/sec) and realistically it
probably reaches 400-450 KB/sec maximum... Maybe a bit more if you are going
downhill and you have a stiff tailwind. Even if you have one of the Dayna
BlueStreak 10/100 cards you aren't likely to hit anything near the 5+ MB/sec
speed of the built-in SCSI. If memory serves, the onboard SCSI can achieve
close to 6 MB/sec with a fast enough drive. I have certainly exceeded 5 MB/sec
using a Quadra 950's built-in SCSI. It's all pretty much a moot point if there
isn't too much data being sent though. That's the reason I asked about the
size of the image. Though 5 GB is nothing to sneeze at, the difference between
.5 MB/sec and 5 MB/sec is not significant if you are looking at 3 days to boot.
Now if we were talking about 500 GB...
When talking about speed, keep in mind we're talking about emulated code. Fast
optimized 68040 code was often faster than poorly written PPC code in the very
early 601 era. For this experiment it's having to translate the PPC
instructions to 68k before executing them - this is always orders of magnitude
slower than native code. It's just usually you're emulating a slower, obsolete
architecture on a much faster newer architecture (think Rosetta on Intel Mac OS
X running PPC code) rather than the other way around.
A big part of the emulation problem, as I understand it, is that you are trying
to emulate a RISC processor on a CISC processor. Every RISC op-code will take
numerous (slow and ponderous) CISC op-codes to emulate where the reverse (one
slow and ponderous CISC turning into several quick and efficient RISC) works
much better. But in any case, emulation is a snail-like process... I still
shudder at the original SoftPC / Virtual PC offerings... And I think those
were PPC only (no 68K option).
Having an activity monitor auto-start is a good idea. I'll check on that.
Scott
Good luck, and let us know how it's going... Sometime next week I would guess.
:-)
Derek
On 1/10/2014 3:57 PM, Derek Morton wrote:
Ahhhh... Danamania... I wonder what she is up to these days?
I am not sure I see the point in this exercise, though a great deal of my ideas
are likely considered fruitless by most.
How large is the image you are running it from? You mentioned possibly
tweaking some of the settings - Perhaps having activity monitor autostart might
help you see if there is anything going on (and cpu usage). Do you think much
of the slow execution was due to network speed? Perhaps on a 950 with a nice
fast hard drive... Big display... Over-clocked...
I know our current machines are faster than older ones, but I really never
thought the PPCs were THAT much faster than the 68Ks... 3 days to boot?!?! LOL
Derek
On Jan 8, 2014, at 10:44 PM, Scott Holder wrote:
Picking up where Danamania left off many years ago, I finally got around to
trying to boot Mac OS X on my 68k Mac. I've created a page about it so far at
http://iamscott.net/68kosx/ . So far I seem to have gotten farther than she
did! I can't say it's exactly usable, per se, but we'll see what comes of it. I
also really, really need to get a decent camera and some studio lights...
Scott H.
--
--
--
-----
You received this message because you are a member of the Vintage Macs group.
The list FAQ is at http://lowendmac.com/lists/vintagemacs.shtml and our
netiquette guide is at http://www.lowendmac.com/lists/netiquette.shtml
To post to this group, send email to [email protected]
To leave this group, send email to [email protected]
For more options, visit this group at
http://groups.google.com/group/vintage-macs
Support for older Macs: http://lowendmac.com/services/
---
You received this message because you are subscribed to the Google Groups "Vintage Macs" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/groups/opt_out.
