[ravenz]

Originally Posted by rboisjoly

Anyway... now its been running the installer for OVER 12 hours!! and it is only at 51%, copying "usrcoina.dll"... I guess it should be done by tomorrow night :-)

You're experiencing what I did - yet, in a recent support email from OpenOSX, they said:

"Something must be wrong as you are the only person complaining about performance."

I'd like someone who is running WinTel on an Intel iMac with excellent performance to step foward.

I'm glad that Jeshua is reading this forum now, as he will get additional feedback on performance.

[baw]

Originally Posted by inkhead

Why would you EVER shut down your computer? I still can't believe people turn off their computers in 2006. Why? It costs you more power to turn it on/off daily than if you let it go to sleep.

I don't know why people love to turn off their computers, it's almost if people can't handle NOT waiting... they have to wait for something.. or it doesn't feel like a computer

Maybe it is because they choose to turn their computers off. Get over it.

[OpenOSX]

Originally Posted by foo2

A 512MB machine to run this test on isn't a valid test - it's a measure of how slow the hard disk is, because that's about all you'll be doing (just hitting the page file 100% of the time).

Yes, exactly, thank you.

Originally Posted by foo2

Get a 2GB machine, or at least a 1GB machine, and redo the test.

Open OSX : Can you post numbers for why you believe you're getting this speedup? It would make things much clearer.

I would suggest that he also wait until he has the correct settings. A small difference in the IPS and VGA update settings can have profound effects on performance.

I am not sure what numbers you are referring to. Basically the Intel binary that ships with WinTel 2.0.0 has zero optimization done to it. In our tests we have seen as much as a 5 fold in performance with optimized binaries, and are quite excited about it. When we released 2.0.0 we wanted to include the optimized binaries, however, we were having some reliability issues that we have to work out first.

The 2.0.1 version will ship next week and most folks would rather get their hands on the product today with a free upgrade in the coming days.

And for those that dream of the FULL performance of Windows, it will NEVER happen under Mac OS X; as Mac OS X itself requires clock cycles - so go buy a DELL!


Sincerely,

Jeshua Lacock
OpenOSX Owner
http://OpenOSX.com

[OpenOSX]

Originally Posted by Icruise

Is the program emulating an x86 processor, or is it using the Intel Mac's actual processor to run x86 code natively? Just because the emulator program itself is "native" does not mean that you are actually running x86 code without emulation, which is what your website leads people to believe. There's no way you're getting "nearly native" performance out of Bochs, I can tell you that right now.

The program of course simulates an entire machine. The x86 calls are native.

It is "native" perfomance as it is running native. We do not state that it is as fast a dual 2 gHz Intel Duo Windows machine (anywhere), nor do we make any spec of float performance claims.

That said, our build of Bochs is achieving around 10 to 20 MIPS, which is definitely Pentium class processor performance. Not the faster Windows machine, but usable.

[OpenOSX]

Originally Posted by rboisjoly

Here's my story...

20 inch iMac Dual Core, 1 Gig of RAM

Tooa few hours (about 8 I'd say to get my download link, then it too less than an hour to get the code to uncompress it. That's fine.

Installing did not work at first, I created my disk image (XP Pro SP2) as per the instructions.

Followed the XP tutorial

Adjusted the values to
512 MB of RAM for the emulator,
20000 VGA Update
3000000 Emulated IPS

Clicking Start 586 Emulator would not do anything (It would ask for my password but nothing happened. After some time though with clicking it again and again, restarting, etc, it finally started-up the Bochs emulator when I deactivated the Network Card and started the Windows XP Installer... I was getting some sort of Authentication error in the Console.

Anyway... now its been running the installer for OVER 12 hours!! and it is only at 51%, copying "usrcoina.dll"... I guess it should be done by tomorrow night :-)

Installing is more than slow...

Now, perhaps tweaking the IPS would help, if I know what that meant! Hopefully we'll get a better system for this in the future.

So, I'm waiting for this to complete to see how fast the emulator actually runs once it is done. It may just be that reading from the CD image is painful and that other things are faster, time will tell.

If you are running anything else on your machine, I fear that you may have given it too much memory. You have to save plenty for Mac OS X. Mac OS X by itself likes GOBS of RAM.

If you over allocate memory, it can be worse than under allocating it.

Also, your IPS settings are most likely too high, try around 15,000,000.

[OpenOSX]

Originally Posted by Icruise

I wasn't even addressing ravenz's experiences. Bochs (which is what this is) is a DOG in terms of performance (a lot slower that Virtual PC, which itself was unbearably slow for many things). Even if this version runs a lot faster than previous versions, it's still nowhere near "native" performance.

Bochs running on an Intel will SMOKE Bochs running on a G5.

The performance on the G5s has been lacking, finally with Intels we think that the speed is finally acceptable.

Don't believe me? Run VPC on the Intel (under Rosetta) and benchmark it against Bochs running on Intel...

And as long as the binaries are native (whatever the program - however slow) it would be running native, wouldn't it?

[Simon]

Originally Posted by OpenOSX

Don't believe me? Run VPC on the Intel (under Rosetta) and benchmark it against Bochs running on Intel...

Actually, I'm starting to wonder how much you really know about Macs. There are probably two dozen thirteen year olds on this board who could have told you that VPC will not run under Rosetta.

http://www.macfixit.com/article.php?...60110120028762

[OpenOSX]

Originally Posted by ravenz

They falsely advertise about "nearly native" performance on their website.

Yes, I have 512MB of RAM. However, I dedicated 256MB to WinTel before booting up Bochs. I get another 1GB of RAM on Tuesday and will try the same test with 1GB allocated to Bochs, but I don't think that is the problem.

That is FAR worse than giving it 64 MB of RAM!!!! As explained previously, if you gave it 64 MB of RAM you would only have about 12 MB of free physical RAM for Mac OS X. If you gave it 256 MB, you would be in the hole, minus 244 MB!!! If you don't believe, start your setup as described and monitor your memory with something like top.

Sorry, but can't give something more than you have and expect it to create the RAM from the air!

Originally Posted by ravenz

Bochs is consuming 99.9% of the CPU. It's CPU-bound.

This is meaningless.

Because even when it is idle, it consumes 90-95% Don't believe me? Boot in to FreeDOS at watch your CPU meter while it sits there at the DOS prompt.

[OpenOSX]

Originally Posted by Simon

Actually, I'm starting to wonder how much you really know about Macs. There are probably two dozen thirteen year olds on this board who could have told you that VPC will not run under Rosetta.

http://www.macfixit.com/article.php?...60110120028762

Uh yah - thats my point. So admit-ably, Bochs runs a just little faster on the Intel than VPC, right?

Bochs = Usable VPC = DOA

And even if VPC run under emulation Bochs with native Intel calls would run circles around it. And Bochs is not much slower than VPC when emulating the Intel calls...

[Simon]

Riiiight.

[OpenOSX]

Originally Posted by ravenz

They falsely advertise about "nearly native" performance on their website.

Yes, I have 512MB of RAM. However, I dedicated 256MB to WinTel before booting up Bochs. I get another 1GB of RAM on Tuesday and will try the same test with 1GB allocated to Bochs, but I don't think that is the problem. Bochs is consuming 99.9% of the CPU. It's CPU-bound.

It is a native application, therefore it IS native performance.

Giving a 1 GB of RAM will not help! You will end up starving you system of vital memory again.

[toddtmw]

Originally Posted by OpenOSX

It is a native application, therefore it IS native performance.

Giving a 1 GB of RAM will not help! You will end up starving you system of vital memory again.

Man, I hope you stop posting here and answer my support e-mail about not being able to boot at all...

-Todd

[baw]

I'll quote someone from digg.com;

"This app is pure crap. This 'OpenOSX' company is infamous for bad business practices, taking people's money and not sending product, and producing a pure crap product to begin with. All their products are generally available for free elsewhere, they just repackage and ship (i.e. 'if' they ship).

Just check out this link where costumers hate this product. http://www.versiontracker.com/dyn/mo...amp;vid=114956"

[baw]

http://www.xlr8yourmac.com/OSX/openo..._emulator.html

[baw]

http://fink.sourceforge.net/pr/openosx.php

Anyone need more proof that OpenOSX is shady?

[foo2]

Originally Posted by OpenOSX

I am not sure what numbers you are referring to. Basically the Intel binary that ships with WinTel 2.0.0 has zero optimization done to it. In our tests we have seen as much as a 5 fold in performance with optimized binaries, and are quite excited about it. When we released 2.0.0 we wanted to include the optimized binaries, however, we were having some reliability issues that we have to work out first.
http://OpenOSX.com

I'm simply asking you to post applications that have a 5x speedup, and I'm asking that you post benchmarks of common speeds people should expect. Where do you get your claims numbers?

[icruise]

Originally Posted by OpenOSX

It is a native application, therefore it IS native performance.

I think this statement tells us everything we have to know about their speed claims.

[baw]

Originally Posted by Icruise

I think this statement tells us everything we have to know about their speed claims.

Yeah, that statement is sig-worthy.

[Pierre B.]

Originally Posted by Icruise

I think this statement tells us everything we have to know about their speed claims.

Indeed, I am surprised.

[Chuckit]

This thread makes me giggle.

"Of course the binaries are native, so like it or not, it is native performance."

"The performance on the G5s has been lacking, finally with Intels we think that the speed is finally acceptable. Don't believe me? Run VPC on the Intel (under Rosetta) and benchmark it against Bochs running on Intel..." Then, after somebody points out VPC won't run on the Intel. "So admit-ably, Bochs runs a just little faster on the Intel than VPC, right?"

"That said, our build of Bochs is achieving around 10 to 20 MIPS, which is definitely Pentium class processor performance." Meaning it runs as well as a 60 MHz processor (though according to several sources I've seen, this is about a third of what that 60 MHz processor could do).

[GregAlex]

Originally Posted by OpenOSX

It is "native" perfomance as it is running native. We do not state that it is as fast a dual 2 gHz Intel Duo Windows machine (anywhere), nor do we make any spec of float performance claims.

That said, our build of Bochs is achieving around 10 to 20 MIPS, which is definitely Pentium class processor performance. Not the faster Windows machine, but usable.

That answer worries me!

My brother bought a "100% chicken breast" the other day and found it was a reprocessed pattie with veges and whatever else they put in it. He complained that he wanted a genuine chicken breast and that they'd falsely advertised and they said "but the chicken IN the pattie is 100% chicken breast!"... which is technically correct.

Anyway, which Pentium class of performance? Pentium 1 @ 100Mhz? Pentium 4 @ 2 Ghz?

Originally Posted by OpenOSX

And as long as the binaries are native (whatever the program - however slow) it would be running native, wouldn't it?

Not necessarily. Take these 4 options

1) PPC code running through Rosetta, converting Windows-Intel calls to native PPC calls (which Rosetta converts back to Intel)
2) Recompile #1 for Intel -> now it's Intel code, converting Windows-Intel calls to PPC calls (which Rosetta converts back)
3) Rewrite #1 emulation method -> now it's Intel code, converting WIndows-Intel calls to other Intel calls.
4) Or what Xen claims - Intel code passing through Windows-Intel calls.

Originally Posted by OpenOSX

As explained previously, if you gave it 64 MB of RAM you would only have about 12 MB of free physical RAM for Mac OS X. If you gave it 256 MB, you would be in the hole, minus 244 MB!!! If you don't believe, start your setup as described and monitor your memory with something like top.

I assume you're saying OSX requires 436MB - so if you start with 512 and give OpenOSX 256MB you're left with 256MB actual ram (the rest uses virtual memory on the hard disk?). Please note that many functions of an OS are not accessed and are fine to be swapped out - though of course more RAM is going to have a good effect!

[icruise]

Originally Posted by OpenOSX

And as long as the binaries are native (whatever the program - however slow) it would be running native, wouldn't it?

The word "native" is being thrown around an awful lot here. You seem to be saying that you are using "native" to mean that the emulator is running natively on the Intel Macs (meaning that the emulator program is not running in Rosetta). And if you said something like "our version of Bochs works much faster than previous versions because it's Intel native" then maybe there wouldn't be any confusion. But your web page has the following sentence on it:

Our WinTel product built for the new Intel based Macs runs Windows(R) within Mac OS X offering performance we have always dreamed of: nearly native.

If you are indeed using the word "native" in the sense I explained above, this sentence does not make any sense. Either the emulator is native or it isn't. It can't be "nearly native." Unless, of course, you are using "native" to refer to the performance you would get from Windows running directly on an Intel processor. As you can see, your argument does not hold up. Your web site is making claims that are clearly false.

[rboisjoly]

Let us assume here that there is a language issue and communication is difficult. Let us assume this is an ok product and that we can get it running. The OpenOSX guys are at least responding to this thread on a public board, which is a good sign.

Back to business...

I've changed my settings to 256 MB of RAM and 150000 for the other value (can anyone explain what that second number means?)

It seems somewhat faster to install... only 5 hours so far and its at 29% (yck)

Now, I'd like to explain this whole "native" thing...

Here are three examples of code running on an Intel Mac:

1. A PowerPC binary will run under Rosetta which emulates PowerPC calls on the Core Duo chip
2. A Universal binary will compile the original program to run to call the API's using Intel commands, but they are the same calls as the PowerPC code, only, they are not interpreted by Rosetta.
3. A Universal Binary could be created which calls Intel calls directly, but this is a lot more work.

I have a feeling that bochs in this situation is using example number 2, where they have recompiled the code to NOT use Rosetta by calling the API's using Intel code, but it is still emulating an x86 processor. Basically, it is emulating a pentium on a Core Duo, Intel calls are interpreted and executed by an emulator. Its native code, but its still an emulator, not passing-through (like VMWare would do for example).

A real improvement would be to re-write Bochs not to interpret the calls but process the Intel commands directly by passing them on to the Core Duo.

Not sure if this is clear, I think it is emulating a Pentium still, not passing on calls to the Core Duo directly, which is why it is faster than the PowerPC, but slower than a real PC.

Unless they completely stripped the emulation code out of Bochs and are just passing on the calls to the Core Duo... but you wouldn't need bochs at that point, I think it would require something completely different...

If you were to run Bochs on a Dell machine, it would still be emulating a Pentium, right? It would still be slow to a certain point right?

Anyhow, we'll see how fast this is once I get Windows installed... by mid next week... :-)

Installing XP is a lot faster on Guest PC on an iMac G5, I can't wait to get that emulator or VPC for the Intel machines, but it will take some time. Perhaps changing the settings will make this better still...

[baw]

Originally Posted by rboisjoly

The OpenOSX guys are at least responding to this thread on a public board, which is a good sign.

You need to look at the links I posted. This company is a fraud.

[MORT A POTTY]

Originally Posted by baw

You need to look at the links I posted. This company is a fraud.

and if that wasn't painfully obvious from the moment this jerkoff opened his mouth here, but then again, you should have known it was shady anyway from his abuse of the open source software anyway.

this "company" is a fraud at best.

"Jeshua" I'll see you in hell.

[tadd]

My wife and I received our Mactel iMac 20" 1GB RAM about a week ago. It took us several days to get the files moved from her former G5 17" first edition iMac and before she got comfortable with the new machine.

Before I talk about my review of the machine I want to answer somebody's question on why you'd turn off your iMac instead of sleeping it. When the computer is asleep it is still susceptible to power loss. Off is safe from power loss. I always thought that turning off a computer, even a sleeping one, was bad for your file systems. Correct me if I'm wrong??

My wife mostly uses Apple applications. I use a few non-Apple apps but haven't tried them on the new computer yet.

The computer is amazingly fast compared to the G5 iMac. iPhoto is still slow. Safari and web pages viewed are quick. All OS GUI functions seem very fast. Quicktime is very responsive even with 1080 HD movies (which are still more pixels than the 20" display can show).

The computer is very quiet. Noise was one of my biggest complaints about the G5 imac. The heat off the top seems less than the G5 imac as well.

We had no end of HD corruption problems on the G5 iMac. I bought the new computer partially because of the geek in me but also because it may take the bad HD issues away from my wife's computing experience.

The camera works very well. It is aligned well for pointing at the user as well -- better even than when the iSight was sitting on top of the G5 imac. We eventually bought a gooseneck stand for the old iSight.

The little remote is cute and works well. That may be a useless feature for us but I can see that Apple is heading for a TV viewing environment that isn't unreasonable. I want to see them make it possible to select a TV show from ITMS and have it play, all from the remote. Can they do this now? I haven't tried really.

I don't like the location of the power switch or the lack of USB on the front panel for the iPod shuffle to plug into. This is pretty much the same as the old G5 iMac.

The volume output of the audio port seems lower than the G5 imac. It's still enough to drive the creature speakers though.

I have a question about partitioning. From my MSWindows days I had a habit of making a separate user partition. This allowed me to reformat the boot partition to reinstall without worrying about my user data. i did this on the G5 iMac as well. Is this a problem? I had all sorts of HD corruption problems with the G5 iMac. With the Mactell iMac we did not partition. I just copied the user data over to the Desktop and home folder after logging in the first time. Could the corruption issues have been aggravated by partitioning?

[JKT]

Originally Posted by tadd

<snip>
I have a question about partitioning. From my MSWindows days I had a habit of making a separate user partition. This allowed me to reformat the boot partition to reinstall without worrying about my user data. i did this on the G5 iMac as well. Is this a problem? I had all sorts of HD corruption problems with the G5 iMac. With the Mactell iMac we did not partition. I just copied the user data over to the Desktop and home folder after logging in the first time. Could the corruption issues have been aggravated by partitioning?

If your partition wasn't large enough to accommodate the OS and the virtual memory created while using the system then, yes, partitioning may have been the problem. What size was the partition, how much RAM did you have and what version of OS X was it?

You don't need to partition to be able to nuke the old system and re-install with Mac OS. You can do something called an Archive and Install - this will rename the old system (so that its contents are preserved if you need to copy something from it) and replace it with a freshly installed system. Partitioning is generally a complete waste of time and disk space unless you have very specific requirements (such as the need to boot different versions of the OS if you are a developer).

[ViktorCode]

Originally Posted by rboisjoly

Here are three examples of code running on an Intel Mac:

1. A PowerPC binary will run under Rosetta which emulates PowerPC calls on the Core Duo chip
2. A Universal binary will compile the original program to run to call the API's using Intel commands, but they are the same calls as the PowerPC code, only, they are not interpreted by Rosetta.
3. A Universal Binary could be created which calls Intel calls directly, but this is a lot more work.

I would like to clear the matter of native and non-native code on Intel Macs. There are only to cases:

1. The code is compiled for PowerPC line of processors. That kind of code is non-native to Intel, so it will run under Rosetta (unless there is a compatibility issue, see below), which is a translator replacing blocks of program code on the fly (as opposed to emulator, which is emulating another CPU and enviroment to run the original code).

2. The code is compiled for Intel x86 architecture. That kind code is native, it just runs.

Universal binary contains both versions of code, as its name suggests.

What the code do by itself is a different story. For example, let's look at some Intel x86 emulator for PowerPC architecture. It works by replacing each instruction call of x86 application with the PowerPC instruction call, which allows x86 application to run on your Mac, albeit slowly. What will happen if we recompile this emulator as universal binary? (it is simply a matter of checking the "Intel" checkbox in XCode) The emulator still be an emulator, but it will run natively doing now unnecessary work: it is still replacing each x86 instruction call of application it runs into corresponding PowerPC call... which will run under Rosetta on intel Mac. To run code natively the emulator has to be rewritten completely. In other words, it will no longer need CPU emulationg part. It still need to emulate virtual address space and so on, but it will run much faster.

Now, how do you think, which kind of emulator OpenOSX is?

A few words about Rosetta: Apple strongly resisted when someone called Rosetta an emulator, and they have the right to do so. Rosseta runs rewritten blocks of code which are translated as application making calls to them. It uses caching to boost code reuse and other optimization methods. According to ADC documentation Rosetta will not run the following:

Applications built for any version of the Mac OS earlier than Mac OS X — that means Mac OS 9, Mac OS 8, Mac OS 7, and so forth

The Classic environment

Screensavers written for the PowerPC architecture

Code that inserts preferences in the System Preferences pane

Applications that require a G5 processor

Applications that depend on one or more PowerPC-only kernel extensions

Kernel extensions

Java applications with JNI libraries

Java applets in applications that Rosetta can translate; that means a web browser that Rosetta can run translated will not be able to load Java applets.

Rosetta does not support precise exceptions. Any application that relies on register states being accurate in exception handlers or signal handlers will not function properly running with Rosetta.

[rboisjoly]

Yes, your two cases are what I meant to describe, #3 is basically what you say after, you need to rewrite a program (as a universal binary of course, so it was useless for me to specify that). You need to bypass all "emulation" of the processor itself and only emulate the rest.

I do not think Bochs does this at all... it will happen eventually with other emulators though.

[kcmac]

Originally Posted by rboisjoly

Let us assume here that there is a language issue and communication is difficult. Let us assume this is an ok product and that we can get it running. The OpenOSX guys are at least responding to this thread on a public board, which is a good sign..

Look. A lot of us here, if not all of us, have used VPC. The word NATIVE refers to speed when talking about VPC. Always has, always will.

OpenOSX knows this and for marketing reasons has chosen to use this term rather loosely.

A long drawn out discussion on the word native is pointless. We have a saying where I work. How do you know when a marketing person is not being truthful? When their lips are moving.

I do however hope to see more responses from OpenOsX to those that are currently trying to use this app. They can only be helpful and are appreciated.

[tadd]

Originally Posted by JKT

If your partition wasn't large enough to accommodate the OS and the virtual memory created while using the system then, yes, partitioning may have been the problem. What size was the partition, how much RAM did you have and what version of OS X was it?

You don't need to partition to be able to nuke the old system and re-install with Mac OS. You can do something called an Archive and Install - this will rename the old system (so that its contents are preserved if you need to copy something from it) and replace it with a freshly installed system. Partitioning is generally a complete waste of time and disk space unless you have very specific requirements (such as the need to boot different versions of the OS if you are a developer).

Thanks JKT for the questions.
I'm sorry to the rest of the thread for this getting off topic.

The computer had a 256MB DIMM from Apple and a 512MB DIMM from Crucial. It used OSX10.3 through 10.4.3 during the time it had corruption issues. The partition for boot was 50GB and 200GB for the user space. The reason I like partitioning is so that disk repair may be completed. You can repair the partition that is hurting by booting off the other partition. in addition the abuse of the user partition doesn't fragment the boot partition (I realise that OSX defragments on the fly).

The boot partition wasn't getting corrupted as often as the user partition. We had 3 non-repairable failures (using Apple's DIsk Utility and 3 other repairable failures on the user partition. Only one repairable failure and one non repairable on the boot parition. This is in a period of about 14 months (I think). No SMART errors were reported. The drive is a 250GB SATA drive that I put in the machine the day it arrived.

The 2nd to last event coincided with the original Apple 256MB DRAM failing. THe failure was that the machine wouldn't boot any more and made interesting sounding chimes. I removed that stick and the machine worked fine. I acquired a RAM test and decided that the remaining stick was intact. After that I erased both partitions, installed OSX 10.4, copied files from a backup into the user parition and ran for about a month before I had another issue. This time Disk Utility couldn't fix the problem (again) and I acquired DiskWarrior.

This machine had another issue where the display went out. This was repaired by Apple.

Right now the machine is on my desk displaying some nice widgets. i suspect it will end up in our living room as an "extra" computer. I am considering next moves.
1. put the original factory 80GB HD back into the unit -- restore it to a more or less default condition though with modern software.
2. if another corruption issue happens on the factory HD, then take it to the local genius bar for consult.

Any ideas?

[ravenz]

Originally Posted by OpenOSX

[Our free WinTel 2.0.1 release will be free and available soon with increased performance.

OpenOSX released the 2.0.1 updater today, but I don't see how to download the update.

[JKT]

Originally Posted by tadd

Thanks JKT for the questions.
Right now the machine is on my desk displaying some nice widgets. i suspect it will end up in our living room as an "extra" computer. I am considering next moves.
1. put the original factory 80GB HD back into the unit -- restore it to a more or less default condition though with modern software.
2. if another corruption issue happens on the factory HD, then take it to the local genius bar for consult.

Any ideas?

50GB is more than enough for the system etc. Either you got a lemon hard drive, you have other failing hardware, or perhaps your iMac is overheating? Sticking the original HD back in will tell you whether it is the former so I would try that for a while then go to the "geniuses" if it continues to fail.

[Blackcrayon]

Straight from the Bochs faq:

1.9.5. Tell me about performance when running Bochs.

Because Bochs emulates every x86 instruction and all the devices in a PC system, it does not reach high emulation speeds. Kevin reported approximately 1.5MIPS using Bochs on a 400MHz PII Linux machine. Users who have an x86 processor and want the highest emulation speeds may want to consider PC virtualization software such as plex86 (free software) or VMware (proprietary and commercial).

1.5 MIPS on a 400 MHz Pentium II, you're talking 10 Mhz 286 speeds... Even if you multiply that by ten it's

If the Bochs guys themselves say their emulator is slow on x86, there's nothing OpenOSX can do

[icruise]

Originally Posted by Blackcrayon

If the Bochs guys themselves say their emulator is slow on x86, there's nothing OpenOSX can do

You know what they could do? Not advertise the product as if you were going to get native Intel speeds out of it.

[Eriamjh]

I just realized maybe the Intel Macs aren't coming in as nearly twice as fast as the G5s in iMovie and Quicktime because they are not optimized for two processors? Until now, now consumer Mac had dual CPUs. Now that they do, maybe Apple will compile them to take adavantage of them? (I just had to get off the Wintel stuff for a post or two).

[Pierre B.]

Originally Posted by Eriamjh

I just realized maybe the Intel Macs aren't coming in as nearly twice as fast as the G5s in iMovie and Quicktime because they are not optimized for two processors?

Of course they are not properly optimised (iMovie does not even use two processors, not sure about QuickTime, but I think it does). Look here (near the end of the page) for example what happens with better optimised software:


I have to say after a few hours with my 20" Intel iMac I love it. I just sold a 2.0Ghz 20" G5 and the Intel iMac feels much snappier. The biggest difference I've noted so far is in handbrake. H.264 encoding on the G5 was slow while the Intel iMac is basically encoding in real time. From a FPS perspective the Intel iMac is ~4-5X faster encoding @ an average of ~25 FPS while my G5's average was ~5-7 FPS.

Originally Posted by Eriamjh

Until now, now consumer Mac had dual CPUs. Now that they do, maybe Apple will compile them to take adavantage of them? (I just had to get off the Wintel stuff for a post or two).

You can bet that they will do it. The question is if this will happen in this year or next year with iLife 07.

[t_hah]

A friend of mine sent me the following webpage, that compares the Intel iMac, iMac G5 and the Power Mac G5 Quad in various tests.

http://web.mac.com/handras/iWeb/Site/intelimac1.html

There is kind of a cool QuickTime presentation available also with only the numbers. The presentation seems to be done with the new Keynote.

http://web.mac.com/handras/iWeb/Site/intelimac.html

[Anand]

Originally Posted by t_hah

A friend of mine sent me the following webpage, that compares the Intel iMac, iMac G5 and the Power Mac G5 Quad in various tests.

http://web.mac.com/handras/iWeb/Site/intelimac1.html

There is kind of a cool QuickTime presentation available also with only the numbers. The presentation seems to be done with the new Keynote.

http://web.mac.com/handras/iWeb/Site/intelimac.html

That is really well done. Can't read it (don't even know the language!) but it looks cool!

[t_hah]

The blog in in Hungraian. The graphs are in English though for easy. From what I heard, an English version of the blog is in the works....

The numbers looked great I thought.

[Eug Wanker]

It is a native application, therefore it IS native performance.

I run VPC natively on my G5 Mac, because my VPC is PowerPC native.

The blog in in Hungraian. The graphs are in English though for easy. From what I heard, an English version of the blog is in the works....

Nice benchmark page. A few of the graphs say "Time in seconds, longer bars are better" though, which doesn't quite make sense.

[Duck57]

I transferred AW from my old G4 under 10.3.9 and it works fine on my Inel iMac G5.

Originally Posted by GreatMuppetCaper

This is exciting... seems things are gonna go much smoother than the OS 9 to OS X transition!
I am concerned that AppleWorks is not included... that's a real bummer. But if one doesn't want to pay for Office I wonder if paying $79 for AW is even an option.

Can someone try loading AppleWorks and see if it runs under Rosetta?
AppleWorks may be old and outdated, yada yada yada, but atleast I could do headers and footers with page numbering. Plus I don't like that clunky font panel in TextEdit. Since when was that more convenient than going to a menu!?

I don't know if getting iWork is really an option for me either- when I used the last version it didn't take long to realize that it sucks for writing papers- more of a page layout tool than a good word processor.

Thanks for all the info on these iMacs!

[baw]

Originally Posted by Duck57

I transferred AW from my old G4 under 10.3.9 and it works fine on my Inel iMac G5.

[ciparis]

Originally Posted by OpenOSX

And for those that dream of the FULL performance of Windows, it will NEVER happen under Mac OS X; as Mac OS X itself requires clock cycles - so go buy a DELL!


Sincerely,

Jeshua Lacock
OpenOSX Owner

That's probably the [self-edit]-est thing I've seen you say yet. Which is quite a feat -- there's alot of competition for that title.

[Mediaman_12]

Ok so the thread has swayed way from what the tittle suggests but. I have Just got my Intel iMac, and didn't realise that I would need new drivers for my Logitech MX1000 mouse, the thumb buttons ect no longer working So in there any word from Logitech on updated drivers, or on any third party drivers (like USB Overdrive).

[Eriamjh]

Originally Posted by Duck57

I transferred AW from my old G4 under 10.3.9 and it works fine on my Inel iMac G5.

Yeah. What the hell is an Intel iMac G5?

[toddtmw]

Originally Posted by Mediaman_12

Ok so the thread has swayed way from what the tittle suggests but. I have Just got my Intel iMac, and didn't realise that I would need new drivers for my Logitech MX1000 mouse, the thumb buttons ect no longer working So in there any word from Logitech on updated drivers, or on any third party drivers (like USB Overdrive).

Yes, the Logitech people were very helpful when I asked about new drivers. This is what they had to say:

Dear Todd,

Thank you for your recent inquiry about your Keyboards.

I understand that you are experiencing difficulties in using the LCC software on your Intel iMacs.

Thank you for your interest in Logitech products. Unfortunately, Customer Support cannot provide information on products that have not yet been released. Since Customer Support's knowledge is limited to new and current products, we do not have any information on the product that you have mentioned. For information on the newest products, please visit the Logitech web site at the following link:

http://www.logitech.com/index.cfm?pa...p;languageid=1

When I rpelied back that this was a useless unacceptable answer and asked if my question could possibly be escalated to someone who has a clue about anything, I got no response.

However, I knew that Rosetta did not support (among other things) anything that loads a preference pane. So, go to your old computer, open preferences. See all the icons on the bottom? All of those will have to be upgraded to work on Intel. On my old Mac, I had the Logitech Control Center down there. All I really wanted to know from Logitech was that they were aware of the issue and that they were working on it. But they appear to be clueless. (Of course, their web site also says their software only supports up to 10.4.2...)

-Todd

[toddtmw]

Originally Posted by ravenz

OpenOSX released the 2.0.1 updater today, but I don't see how to download the update.

Did you get the update yet? I got an e-mail this morning with the link, but didn't have time to do much with it before I had to go to work. I started the win98 installer and it seemed to start out at a good speed. Then it stopped, saying I needed 16 MB to install Windows. (I had it set to 64, so I'm not sure what the problem was...) I had to go to work and didn't get to play with it anymore.

I have 2 GB of RAM coming Friday, but was hoping to play with it more tonight with the 512 MB.

Like I said, when the installer first started up, it felt like the speed wasn't bad, but I didn't really get very far. Has anyone else tried 2.0.1 to see if it is any better?

I'll post more on my experiences tonight.

-Todd

[Chuckit]

Maybe the new one will make the binary even more native than the last release.

[handras]

Originally Posted by Anand

That is really well done. Can't read it (don't even know the language!) but it looks cool!

Thanks.
I've done the English translation of the test. Enjoy it. When I get a MacBook Pros, a 20-inch iMac, I will make it with them, too, maybe complete with a Dual Core PowerMac G5.