It doesn't matter if you put a bowtie on a pig. It's still a pig.

I don;t know, but the developer post I saw were pretty convincing in their evidence that CherryOS was really PearPC under the hood...

I guess we'll have to wait & see what PearPC sez...

John

what Apple will think about this. I doubt they'd have any say in the product, since it's only an emulator, but still, I'm sure they won't care for PC users skipping out on switching to a Mac if they can just run an emulation on their Windows box.
Of course, nothing is as good as the real thing, as users of VPC know.

In the end though, one of the major complaints against PC's is Windows. If users can run Mac OS X on their PC, I wonder what they'll do.

This software is merely PearPC packaged with a clunky amateurish interface. It uses GPL'd code illegally.

This software is merely PearPC packaged with a clunky amateurish interface. It uses GPL'd code illegally.

There are a lot of difficulties you have to overcome when making a any emulator, particularly a PPC (and OS X) emulator on the x86. For example:

- the PPC has 32 architectural registers vs. 8 on the x86;
this means you have to use memory-memory addressing modes,
which don't exist for all instructions, or register-memory
addressing modes, which will double or quadruple the base
instruction count

- to get good performance, you'll have to use the MMU; so
either you are sharing the MMU with a native OS (painful)
or you can't run a native OS at the same time

- the PPC and x86 have opposite byte orders, so whenever you
load or store you need an extra byte-swap instruction

- to get good graphics performance, you need to have a path
to native OpenGL - this is feasible but is definitely
non-trivial to implement

- you should really emulate a G3 rather than a G4 -
otherwise the G4's vector instructions will cause a major
*slowdown* as you try to emulate them or map them to
x86 vector instructions

- the only way to get good performance in an emulator is
to use translation; but this translation is costly -
even if you use off-line and cached translations, they
have to be looked up; if you use dynamic translation,
then it cuts your dynamic performance

- fortunately Unix is pretty good about keeping data and
code in different pages; but if you ever have them in
the same page, you're going to have to invalidate any
translations in the modified page

- the PPC tends to do more per clock cycle than the x86;
this will increase your clock-per-clock slowdown

Based on these considerations, I would be impressed if they managed to get 20-25% of native performance, e.g. something like an 850 Mhz G3 emulated on a 3.4 GHz P4. That would be usable. Now 80%, that would be miraculous.....

Anyway, the only way to evaluate this emulator is to look at the actual performance by running actual software.

So, is anyone going to step up to the plate and run some benchmarks?

A more likely estimate would be a 50-100 Mhz G3.

Usable, maybe, but not much fun.

What your saying makes perfect sense...

isn't a 700 Mhz G4 equivalent to a P4 of almost twice that speed anyway? hehe... if a 1.6 Mhz P4 or AMD (i know there are differences, don't need to point it out) performs at the same pace as a 600 or 700 Mhz G4, it should run all OS X equivallent programs just dandy... unless, of course, they want to have a snowball's chance in hell of recording or editing anything...

how's that for pretzel logic?

iLemmings good news, you don't have to pay for that expensive Mac hardware. Finally a fast low cost machine with your favorite OS. Apple's not going to like this.

Doesn't the EULA for OS X specify that it has to be run on Apple hardware, or am I mistaken about that?

"This License allows you to install and use one copy of the Apple Software on a single Apple-labeled computer at a time."
(from the Mac OS X Software License Agreement)

So, I'm guessing it ain't legal to run OS X on peecees.

Quote

Guest wrote:
"This License allows you to install and use one copy of the Apple Software on a single Apple-labeled computer at a time."
(from the Mac OS X Software License Agreement)

So, I'm guessing it ain't legal to run OS X on peecees.

I dunno, I got some extra Apple stickers. If I slap one on a PC case, does that make it "Apple labeled"?

Is this really a "TMO Report"? Or is it just a reworded press release? It certainly feels like the latter.

"Finally a fast low
cost machine with your favorite OS. Apple's not going to like this"

Except the emulation is filled with compatability problems and runs at an awful speed. Finally, a fast low-cost machine that runs my favourite OS horribly! What a deal. Apple must be sweating.

a cheap pc? you mean the calculators they seel for 500$ and claim it's a computer. look at a comparable deel to the top of the line powermac.
doesn't look cheaper to me. wonder why? cause 500$ don't get you $#!% unles it's a mac mini. and as for calling us iLemmings. Are you mad steve called pc users lemming 21 years ago and you still haven't gotten it?

 Print This Page


Dell Precision Workstation 670
Date & Time: March 10, 2005  12:06:24 AM
SYSTEM COMPONENTS
------------------------------------------------------------------------
Dell Precision Workstation 670 Qty    1
Intel® Xeon™ Processor 3.60GHz, 2MB L2 Cache, Microsoft® Windows® XP Professional, SP2 with Media Unit Price   $3,976.00
Save $200 on select Dell Precisionâ„¢ 670 workstations through the Small Business division.
Special offer           - $200.00
Catalog Number:  4
 Module  Description  Product Code  Sku  Id
Dell Precision Workstation 670 Intel® Xeon™ Processor 3.60GHz, 2MB L2 Cache 6T362 [221-7978] 1
Operating System Microsoft® Windows® XP Professional, SP2 with Media XPP2E [420-4860] 11
2nd Processor (Must match speed selection above) Intel® Xeon™ Processor 3.60GHz, 2MB L2 cache PR362 [311-4853] 2
Memory 512MB, DDR2 SDRAM Memory,400MHz, ECC (2 DIMMS) 512E2 [310-4983] 3
Keyboard Entry Level, USB, No Hot Keys U [320-3811] 4
Monitor No Monitor Option N [320-3316] 5
Graphic Cards 128MB PCIe x16 ATI FireGL V3100, Dual VGA or DVI + VGA Capable ATI128 [320-3958] 6
First Hard Drive 160GB SATA, 7200 RPM Hard Drive with DataBurst Cacheâ„¢ without RAID 160SAT [341-1139] 8
Hard Drive Configuration C1- All SATA drives, Non-RAID, 1 drive total configuration SATA12 [341-1113] 9
Floppy Drive Options No Floppy Drive NFD [341-1436] 10
Mouse Dell USB 2-Button Mechanical Mouse with Scroll ELD [310-5203] 12
File System NTFS File System NTFS [420-3699] 13
CD-ROM, DVD, and Read-Write Devices 16X DVD+/-RW w/ Sonic RecordNow! Deluxe,Sonic DVDit! SE,CyberLink PowerDVD DVRM16S [313-3026] 16
Speakers No Speaker option N [313-2663] 18
Resource CD Resources CD contains Diagnostics and Driver for Precision Systems RCD [310-5419] 21
Hardware Support Services 3 Year Basic Plan W3YOS [980-7000][980-7002] 29
Onsite System Setup No Onsite System Setup NOINSTL [900-9987] 32
Quick Reference Guide Quick Reference Guide REF [310-5420] 40
Intel Hyper-Threading Hyper-Threading feature preset to "ON." Can be disabled/enabled in BIOS. HYPER [460-7163] 41
Energy Star Energy Starâ„¢ ES [310-4821] 50
Mail- In Rebate None NONE [463-1832] 81
TOTAL: $3,776.00
------------------------------------------------------------------------

oh yeah thats comparable to base powermac which is 3,000 not 3,700 like a comparable Dull

Assuming, as is fairly apparent, that CherryOS is based on PearPC, I would expect it to have similar performance.

The best reported XBench results for PearPC are 5.26 (unknown system - presumably a fast PC as the results for an Athlon 3200+ were 3.93),
which is still significantly slower than the slowest Mac system rated, the PowerBook G3 which came in at 26.34.)

For comparison, the Power Mac G5 results are around 180.

http://ladd.dyndns.org/xbench/

If you consider the G5 to be comparable in performance to a fast PC, then you are looking at a 34x slowdown, or about 3% of native performance.

Quote

RealityCheck wrote:
iLemmings good news, you don't have to pay for that expensive Mac hardware. Finally a fast low cost machine with your favorite OS. Apple's not going to like this.

LOL oh man this is one of the funniest posts I have seen from anyone is a long time. Someone's lack of understanding of fundamental computing concepts is showing!

^^

Looks like you're having your own problems with computer fundimentals, guest. Put your comments OUTSIDE the quote box.

Quote

Small White Car wrote:
^^

Looks like you're having your own problems with computer fundimentals, guest. Put your comments OUTSIDE the quote box. Oh, wait a minute, I get it, he is rewording people's comments. What a beautiful format!

The pearpc mailing lists describe the processes needed to compare strings embedded within each application.

Strings such as function names. Some of which you would expect to overlap with similar (but rarely identical) sounding names. If you had two completely independent emulators for ppc code, you could expect symbols within the executable to have some commonalities, yes. You might see 'ppc' and 'x86' and 'altivec' referenced as the coders name the pieces of the emulator for their functions. You would see some references to the mac PRAM and PROM, and to firmware. symbols would overlap in that they may contain various permutations of these.

However, the downloadable version of CherryOS contains vast swaths of hundreds of symbols that are word for word, spelling for spelling, misspelling for misspelling, capitalisation, grammar and punctuation IDENTICAL to those that have been in pearpc from the start, from before CherryOS's existence.

There is more chance of winning the lottery 50 times in a row than there is of symbols matching to this extent.

Maui X Stream couldn't be lying any more if they said "2+2=5"

The following 3 files that come with CherryOS
are identical to the file "video.x" that comes with PearPC
and MacOnLinux:

"Cherry IDE.dat", "OSMAC.AXM" und "CASE.EXM"

Copy them together and you have a 100% identical copy
of video.x.... Video.x was written by the MacOnLinux
project,a driver for a virtual graphics card, and it is
under the GPL. Easiest proof that CherryOS STOLE SOMETHING.

Any questions left?

Read this blog entry here; has links to the keygen for CherryOS (download a copy from the official website and then visit this site for the keygen:

http://davidstechblog.blogspot.com/2005/03/run-mac-os-x-on-your-pc-easily-and.html

You're a retard, you know nothing of how PowerPCs work when compared to x86 processors. In now way would it run better on a P4. Go to school idiot.

Actually, the G4 emulation used in PearPC isn't included in the main branch now, because it's horribly under-developed. (And I will guarentee you that it's the same way with CherryOS)

There are only a handful of instructions that are dynamically compiled, and most are just calling their generic cousin (which causes significant overhead)

Now, as for all the slowdowns that have been mentioned, 32 registers to 8, blah, blah, blah. Yes, they exist, yes, they're slowing us down, but PearPC is still a useable platform.

Now, you can debate all you want about if it's worth it or not, but the idea hasn't been to make a product with PearPC, it's been to make a useable emulator for the PowerPC, which hasn't really existed so far (I am aware of QEMU, and SheepShaver, but none of these have been able to run OSX yet)

We (the PearPC developers) are looking to make an accurate, and usable emulator of the PowerPC platform. This means, we're not trying to hack together something that will just run OSX software fast. There are a number of ways to do that way better than what PearPC is doing now. We want to be able to run PowerPC operating systems, regardless of their nature.

CherryOS has so obviously stolen PearPC's code, that it's ridiculous. Not just variable names, which can be argued to have similar names, and formats. But error messages, that are inconsistant from AltiVec portion to core PearPC. This is because I have my own way of wanting to see error messages, which is different from how PearPC is doing them. Since my code is still highly under development, I don't have to comply or be consistant with the core PearPC reporting system, I can use my own. And I have. And CherryOS has all my error messages word-for-word in their program.

I just don't even know what to think of a person, who would steal code so poorly, barely try to cover it up, and then deny openly to the public that they have not stolen our code. I would have to likely say that such person would be an idiot.

even if this is "not" a rip of PearPC, it most definitely contains a modified, and renamed version of HFV Explorer. As there is no mention of the use of this program on the CherryOS site, nor in any documentation in the CherryOS program, it's probably used without permission.

Quote

Guest wrote:
Is this really a "TMO Report"? Or is it just a reworded press release? It certainly feels like the latter.

I agree. Were the forum posts reporting good performance read by the TMO staff itself or alleged by Kartes? If the former, I wish the article would have quoted some of the posts, or at least specified the forums, so that I could check it out myself.

Hope your code can help you stop those shysters Dan

>We here at the Mac Observer are mindless Mac minions and have >the t-shirts to prove it.

Finally the truth comes out

Actually I ran pearpc, runs very good, especially when they compiled it for athlon. Don't have XP so couldn't do networking, but I think someone needs to sue Maui Xstream from the sound of it.

Quote

Guest wrote:

Actually I ran pearpc, runs very good, especially when they compiled it for athlon. Don't have XP so couldn't do networking, but I think someone needs to sue Maui Xstream from the sound of it.

I'm running pearPC on a Windows 2000 box and network (Internet to be exactly) works just fine. Set-up like described in the How-to's (for XP). Only difference: the TAP-Win32 adapter gets the IP 192.168.0.1 and this can't be changed - if used for ICS.

greetz Tom

I would like to correct the article on a couple of points:

1) CherryOS does NOT have FireWire support
2) CherryOS does NOT have USB support. I think it was 'promised' at one time but strangely has not materialised.
3) Exactly what hardware does it detect?

This article is little more than a puff piece giving further market penetration to a 'product' whose future is on less-than-solid grounds. Please do your research a little more diligently, for the sake of any poor folk who may actually pay money for this software.

I can install Cherry OS, but OS-X 10.2.3 does not install. I've tried on two separate machines, both running Windows XP "latest version" as per the install instructions.

Nope doesn't work, doesn't do what is says on the tin. What a disappointment. I'll try a different version of OS-X when I have time, but my experience is BAD so far...

DO NOT BUY THIS PRODUCT!!!
I downloaded the trial version and tried at least 10 times to get OS X to load with no success. I've got as far as the initial grey apple screen with the spinning timer, but nothing past that. If you really want to run Windows and OS X side by side on the same machine, buy an Apple computer and load Virtual PC. It's more money, but it works.
This company is just looking to rip people off for $50. At least they have the decency to allow a free trial period to see what a POS this software really is.

dont buy this bag of wank

lol

Isn't the pearPC code under GPL? Why can't this foolish company rebrand the box and use the code how they choose, including taking credit for it and selling it? Isn't that one of the rights of the GPL?

I use windows. i hate it. i want a mac. im saving for an ibook.

i dont see how it can be any simpler.

Well, people can now get the same results by directly installing OS X on their PC without CherryOS or PearPC. They get the same incompatibility issues, but at least have it running at native speed and save the 50$ or so that go to the emulatuion software.