Spoken like a true democrat. haha!
But seriously though, virtulization is hot stuff right now. And while I've never used Parallels or anything, and probably never will, I hear it's amazing stuff. Are you sure the performance is bad?
-=|Mgkwho
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Mac OS X 10.5 Leopard - Still Waiting
- Thread starter MacRumors
- Start date
- Sort by reaction score
Spoken like a true democrat. haha!
But seriously though, virtulization is hot stuff right now. And while I've never used Parallels or anything, and probably never will, I hear it's amazing stuff. Are you sure the performance is bad?
-=|Mgkwho
I use Parallels on my iMac and the performance sucks. Maybe it's the fact I only have a gig of RAM though. But regardless, the performance will always be worse than if you were running that OS on that machine without virtualization
Oversimplifing here:
1) OS X is comprised of several propritary layers on top of a (quasi) public domain OS: Unix
2) OS X uses the Mach kernal which many say is inferior in capacity/performance to a Linux kernal
3) Linux, like OS X is based on Unix & very similar, externally
4) The proprietary layers of OS X could be made to run on a Linux kernal as well as on the Mach kernal
To do this would:
-- Potentially improve OS X performance in high-performance environments
-- give Apple access to the large install base of Linux users (mainly servers)
-- bring a much-needed consumer-friendly desktop GUI to Linux
--flesh out the Linux environment with drivers, features, applications, etc.
-- potentially exploit several markets: desktop, server, hand-held at a very opportune time
Well for most folks who need Windoze for a few things Parallels is wonderful. Saves buying a PC too
My wife uses Parallels every day on a 1.8 GHz MacBook with 1 GIG of RAM and it is fast enough for IE. She uses it only for Explorer as she is a Realtor and all those bastard MLS sites are Wondoze only (which should be illegal!). For everything else she uses OS X. Great thing about Parallels now is with the new version you don't even have to see the damn Windoze desk top.
Another great thing is I can back up the whole pile of XP crap in a folder and replace it in seconds if it gets screwed up, try that on a PC.
Right. The reality is they are totally separate now and I doubt we will see the sort of intermixing suggested.
The Inquirer knew about it. www.theinq.net
Though I don't know the time frame when they found out... (certainly before the announcement of Apple switching to Intel procs).
Believe me, I know all about how OSX is based on Unix and Unix is similar to Linux yada yada yada....
Lets say I'm a server admin. I have a bunch of servers running one of the many flavors of Linux with all open source software. I haven't paid a dime in licensing fees for software since it's all free. And it's working great.
So, I could spend thousands of dollars on OSX, a ton of time configuring and installing OSX, and then testing on non-production servers. Then, move those servers into production and risk having things break because its a new system. Too much time and money. If it aint broke, don't fix it.
Besides, all OSX is going to provide is a pretty GUI, which is NOT what you need on a server. Servers don't run GUIs because they waste so many precious system resources, everything's done via command line. Since the core of the OS is almost identical to what you've been running, you won't gain anything.
All true but as a non Unix / Linux expert my servers all run OS X and I for one love the GUI interface and I paid for the licences in no time with client fees.
Off the Record:
DigitalClips: I tried to send you a personal message but it was blocked!
Looked at your profile. I am also an Apple user since 07/1978.
Send me an email at dicklacara@mac.com.
Dick Applebaum
P.S. Interesting web site
I do it on a Windows PC all the time
It's very easy to do - copy the folder with the Windows virtual machine, and replace it if trouble occurs.
This has nothing to do with Windows vs OSX - it's a fundamental advantage to using virtualized environments in container files. If the virtual OS gets messed up, make a new copy from a backup of the container.
And "messed up" has nothing to do with OS shortcomings - if you want every student in a class to have the same starting point - give them each a copy of the container. Next period, a whole new set of containers for the next class.
Another great thing is I can back up the whole pile of XP crap in a folder and replace it in seconds if it gets screwed up, try that on a PC.
It's very easy to do - copy the folder with the Windows virtual machine, and replace it if trouble occurs.
This has nothing to do with Windows vs OSX - it's a fundamental advantage to using virtualized environments in container files. If the virtual OS gets messed up, make a new copy from a backup of the container.
And "messed up" has nothing to do with OS shortcomings - if you want every student in a class to have the same starting point - give them each a copy of the container. Next period, a whole new set of containers for the next class.
On the manner of supposed "internal builds" at Apple:
Apple almost certainly has internal builds. Nightly builds, even. Probably builds which happen every couple of hours, or constantly. So internal builds will happen, period.
Also, if I've read right, the same bugs appear to be in each of the seeds Apple has given out. These should have been easy enough for Apple to fix if they've wanted (they have a whole team working on Leopard, have they been out partying?), so really, we must wonder, why hasn't Apple fixed them?
I'm NOT saying that Leopard is simply missing key components... after all, such things may be too easy to trace. Though possible, I think two other possibilities more likely:
1. Apple has an internal fork of Leopard in which they've fixed most of the bugs which have been presented (not necessarily all) and actively watch for new bug reports, while simultaneously adding or completing new features. The problem here is that this internal fork has "secret" features, which if they were not careful about merging with the developer's preview, would leave traces pointing right to the features. Thus, they haven't merged very much at all. I consider this situation most likely.
2. Apple knows about the bugs, but doesn't care for the time being, considering them low-priority and easy-to-fix. Instead, Apple wants to work on implementing/completing other parts of the system. Once they are relaxing, congratulating themselves on a job well done, they will go back and fix those bugs which, while they may cause crashes, may be "technically" minor. I consider this less likely than #1 however, because personally as a programmer I have a hard time not fixing a bug when I see it.
EDIT: Since everyone's giving predictions: I'd say late April to early June sounds about right. Even the WWDC sounds about right. In fact, how about this: release date announced or extra features released at that NAB thing ... and the actual OS released at WWDC.
Apple almost certainly has internal builds. Nightly builds, even. Probably builds which happen every couple of hours, or constantly. So internal builds will happen, period.
Also, if I've read right, the same bugs appear to be in each of the seeds Apple has given out. These should have been easy enough for Apple to fix if they've wanted (they have a whole team working on Leopard, have they been out partying?), so really, we must wonder, why hasn't Apple fixed them?
I'm NOT saying that Leopard is simply missing key components... after all, such things may be too easy to trace. Though possible, I think two other possibilities more likely:
1. Apple has an internal fork of Leopard in which they've fixed most of the bugs which have been presented (not necessarily all) and actively watch for new bug reports, while simultaneously adding or completing new features. The problem here is that this internal fork has "secret" features, which if they were not careful about merging with the developer's preview, would leave traces pointing right to the features. Thus, they haven't merged very much at all. I consider this situation most likely.
2. Apple knows about the bugs, but doesn't care for the time being, considering them low-priority and easy-to-fix. Instead, Apple wants to work on implementing/completing other parts of the system. Once they are relaxing, congratulating themselves on a job well done, they will go back and fix those bugs which, while they may cause crashes, may be "technically" minor. I consider this less likely than #1 however, because personally as a programmer I have a hard time not fixing a bug when I see it.
EDIT: Since everyone's giving predictions: I'd say late April to early June sounds about right. Even the WWDC sounds about right. In fact, how about this: release date announced or extra features released at that NAB thing
... and the actual OS released at WWDC.You're talking crap. Have you actually used virtualization software like VMWare or Parallels? The performance speed is near native, you just need a lot of memory (which stands to reason, you're running two operating systems in memory). I run Parallels on my 2Gb RAM MacBook and the performance I get running Windows apps is exellent. Far from "absolute crap", I believe you are getting this impression from emulation, not running an OS in a virtual machine.
EDIT: Since everyone's giving predictions: I'd say late April to early June sounds about right. Even the WWDC sounds about right. In fact, how about this: release date announced or extra features released at that NAB thing
Another vote for this one.
Apple has to give the developers builds with the "secret" features, just to make sure that the secret sauce hasn't caused problems with base components that the developers are using.
Anyone who has worked with a software project knows that sometimes a change in subsystem "A" causes a problem in component "B" - even though the connection between the two is unknown.
Well, if you read the thread, you'd see that I have used Parallels, but I only have 1 gig of RAM which I admit, is probably the problem. However, there's no way performance speed is native. If I'm giving 1 gig of RAM to the host OS, and 1 gig to the virtual OS, it's going to run worse than running the virtual OS on its own and it getting 2 gigs.
He did say he had only 1 GiB...
You mean "instruction set emulation" - the fact is that VMware, Parallels, Virtual PC and every other "Intel on Intel" virtualization product emulates a hardware PC for the virtual machine that differs from the actual physical PC (different network or graphics card, different memory size, ...).
Emulating an Intel PC on an Intel processor, of course, doesn't require emulation of all instructions - although many privileged instructions are in fact emulated in software (even on systems with Intel VT support).
Its Possible...and very likely
I've had it up to here *draws line 50 ft above my head* with people who refuse to accept that Apple has two seperate code bases, one for external seeds and one for internal.
Why? Because Microsoft has done it; and Microsoft is ALOT less secret than Apple.
When did they do this? Travel back to 2001 when XP was coming out. The new UI (can't remember for sure, was it code named Luna?) was not introduced until RC1 if I remember correctly (I was an official beta tester).
RC1!! Thats late in the process. Luna was a major internals change for UI rendering, yet they specifically pulled it out of prior builds to keep people in the dark.
Now think about this: People are complaining about how slow the fixes are coming in the seeds for "known issues". Ever stop to think they are already fixed? Maybe the newer versions of libraries with those fixes are DEPENDANT on libraries they aren't willing to introduce yet. Apple wouldn't take the time to back-port a fix to a developer seed when they are focused on moving forward.
Recently I read an article about Windows's code check-in process....it took a month on average for simple UI change code to get checked into the official build of windows for changing shut down buttons....because it goes thru localized builds, then team builds, and then finally gets scheduled for a merge with the full system. You have no idea how many things are fixed or features are added to Leopard that haven't made it up the check-in chain of command.
So yes, its possible and likely.
I've had it up to here *draws line 50 ft above my head* with people who refuse to accept that Apple has two seperate code bases, one for external seeds and one for internal.
Why? Because Microsoft has done it; and Microsoft is ALOT less secret than Apple.
When did they do this? Travel back to 2001 when XP was coming out. The new UI (can't remember for sure, was it code named Luna?) was not introduced until RC1 if I remember correctly (I was an official beta tester).
RC1!! Thats late in the process. Luna was a major internals change for UI rendering, yet they specifically pulled it out of prior builds to keep people in the dark.
Now think about this: People are complaining about how slow the fixes are coming in the seeds for "known issues". Ever stop to think they are already fixed? Maybe the newer versions of libraries with those fixes are DEPENDANT on libraries they aren't willing to introduce yet. Apple wouldn't take the time to back-port a fix to a developer seed when they are focused on moving forward.
Recently I read an article about Windows's code check-in process....it took a month on average for simple UI change code to get checked into the official build of windows for changing shut down buttons....because it goes thru localized builds, then team builds, and then finally gets scheduled for a merge with the full system. You have no idea how many things are fixed or features are added to Leopard that haven't made it up the check-in chain of command.
So yes, its possible and likely.
OK, and that's fine. But if OSX was built on top of a Linux kernel, rather than a Unix kernel, you'd gain nothing. They're so similar, there would be absolutely no reason to do it.
And large companies with enterprise-grade systems (who OSX server is really marketed to), have system admins who can configure a server in command line with their hands tied behind their back and blindfolded, and would have no need for a GUI. And if they really did want an OSX GUI, then there's nothing wrong with the Darwin-based OSX we've always had
the way you make it sound, who knows if we'll even see it in June. I predict we see it next macworld. Leopard 2008!!! no joke. oh wait, that was a joke.
i honestly think it's a distinct possibility that leopard might be announced or go GM at WWDC and released later that month. apple has a history of missing ship dates. though they've been better of late
tv was a reminder to those of us who've been around since the beginning.
Sorry to be all lawyery(not a real word) but rumors are speculation so if you are not a fan of speculation go work for the Apple.com offcial website because only upon events happening does the speculation stop. Good Luck in you new line of work. Have fun.
Speculation = Roumor = Speculation
I very much disagree... see the link posted
http://normalkid.com/2007/03/16/the-difference-between-rumor-and-speculation/
Rumor != Speculation
arn
Thats certainly an exaggeration...
Come on. Servers don't run GUIs and everything is done by command line? As someone who has gone through the pain of installing Oracle 10gR2 on Debian, I can tell you a GUI is mandatory.
Come on. Servers don't run GUIs and everything is done by command line? As someone who has gone through the pain of installing Oracle 10gR2 on Debian, I can tell you a GUI is mandatory.
i honestly think it's a distinct possibility that leopard might be announced or go GM at WWDC and released later that month. apple has a history of missing ship dates. though they've been better of late
That would be horrible. But again, as long as it's pretty bug free I guess I can stand the wait. Doesn't make me any happier though. lol