First Intel Macs on track for January - Page 9

Then how do you know installing, running, an using WINE is a pain?

Wine will eventually work over Quartz. It's just a matter of time.

I'm not. I would never direct people to use WINE. I would first try to get them to get Mac versions, if at all possible.

If that wasn't possible, and Crossover was available, I'd suggest that. If they NEEDED Windows (and some do), I'd suggest that they dual boot - if it can be done. We don't know that yet.

If it was an occasional program, and they didn't NEED power, then I'd suggest a VM, if available.

Lastly, I'd suggest that they find someone to install WINE and their program for them.

EDIT:
Just reread your post. Crossover will, hopefully, be running under Aqua on the x86 machines.

I didn't say I never used it. I'm just not a fan of whichever Linux distro is the flavor of the month at any particular time. The people who use Linux are not masochists, as was suggested. They like if for whatever reason they have.

It was interesting to learn, and play with. But if I used it more than just occasionally, I WOULD be a masochist.

That's why I'm here, and not on a Linux site.

Sure, these things 'just happen'...

My Quartz driver for WINE 'works', so long you're only drawing black lines

You'd be surprised how similar Linux and OS X are. Especially if you use GNOME.

You're the only one suggesting dual-boot.


VMs run Win2K very well. I don't understand why you think they lack "power."

The only reason to dual-boot is to run games, which ironically is the only practical reason why you would run WINE


Only to have it break when they update WINE


Bullshit

I've used GNOME, and that's a heaping helpful of fanciful BS.

GNOME...they're trying to be Windows so bad it's sad.

You have no idea what you're talking about.

O RLY?

Taskbar? Check.

MDI interface apps? Check.

Hideous frames/tabs interfaces? Check.

Take a step back and look at that monstrosity!

I'm not suggesting anything that any intelligent person can't see for themselves.

Forget 2K. Think about XP, which is much slower under VM, and then Vista, which will be slower still.

There are reasons to dual boot. I'm happy to leave it up to those who wish to do so.

When they update WINE, if you knew about it, they add API's from Windows. It doesn't break. It's getting those API's to work in the first place that's the problem.

The project leaders have said that this is an objective on their site. A Quote:

" a strong interest in developing the quartzdrv (in order to replace the X11 driver in Wine)"

Jeremy White has also stated that they want Crossover to run in Aqua.

"Bullshit"???

You are a problem child, aren't you? And I thought we were doing so well.

I'll believe it when I see it.

Note that they didn't say Crossover would run without X11. Why? Because it will.

Sure, we all 'want' it to run without X11, but who has actually coded anything to that effect?

(raises hand. Look around to see if anyone else has their hand raised)

Let me translate what that quote means in the real world:

"We're going to do a quick port of our WINE setup tool, I mean Crossover, and if someone were to develop a Quartz driver we would like that. Don't expect us to do it unless we see a huge market in it. After all we're in it for the money, dollar dollar bling y'all."

Are they developing a Quartz driver in complete secrecy as I type this? Perhaps, but I highly doubt it.

I'm not suggesting it takes a braniac to write one. Hell, I wrote and rewrote one. It doesn't take much brains, just lots and lots of toil.

If you want specific technical reasons why I don't think they are doing SHIT towards this effort, I might go on. However, you seem convinced with their talking points.

Actually, I would like to know why you think they aren't, technically, of course, without the enthusiasm.

Frankly I believe you're confusing GNOME with KDE.

GNOME is quite okay Usability-wise, though certainly not quite at the point Mac OS has been for years. Drag&drop support, for one, is lacking. Window management is inferior. Etc.

But still, GNOME is in many ways ahead of Windows in the UI department. KDE, on the other hand, is beyond retarded.

This is why I stay away. Each of the dozens of platforms has its adherents, and the progfams don't travel.

First of all, if they were developing this driver as opposed to giving it lip service, they would have announced it instead of saying it would be a nice and interesting thing to have.

Secondly, if they were actively tinkering around with Darwine they wouldn't keep that a secret either. They certainly don't with WINE on Linux. They're not working on the bomb for crissakes.

Thirdly, there was a lot of 'talk' about this situation and making a true WINE 2D back-end blah blah, but nothing came of it. Why? Nobody is developing it.

As for ditching X11, it is complicated because WINE was developed on X11, and X11 is a (needlessly) complicated beast. A basic GDI driver has to do a crapload of things before you can even draw a line or blit bitmaps. The existing driver tries to do a lot of this in X11, which only complicates things because to do simple things in X11 requires a lot of tricks and instead of doing things in a graphics library agnostic way, WINE was grafted on X11 like an organ. It uses X11 to draw everything and manage simple types like bitmaps.

For example let's say you want to implement Win32 bitmaps. There is a back-end library GDI which uses the front-end X11 driver. However, because the ONLY front-end anyone is using is X11, the back-end is designed to make the front-end a simple(r) implementation. The back-end also assumes the only front-end supporting stuff like OpenGL/D3D is X11/GLX. The end result is a quartz-specific back and front-end will have to be written. Nobody is (I hope) going to do this without involving a larger effort to clean up both ends so the same code runs on both platforms.

Add to this the problem that you can't draw a lot of these primitives in Quartz and a bunch of other problems...

I was in the process of doing this on my own, but I got too frustrated with the lack of documentation (both Win32 and WINE). I kept having to simulate the behavior of the X11 driver in terms of handling bitmaps, and the internal behavior for Win32 and WINE are undocumented (and may not even agree). Not being a Win32 or X11 expert also led to frustration. Having some help would have been nice.

Eventually I'll get a head of steam and try it again, but I don't have much incentive to try that hard.

I'm not. I'm just not enamored with anything GNOME, although I am amazed how much memory GNOME apps consume.


Yea, you can thank the ICCCM for everything lacking and not working.
(Inter Client Communication Conventions Manual)


My point is GNOME keeps trying to achieve the pinnacle of interface disasters which is Windows.

Ok, you have some good arguments there. But doing this in isolation as you say you have, is different than working a team with a monetary goal in mind. I've seen more difficult projects come to fruition.

If you are saying that Crossover isn't being developed on the Mac, you'd be wrong. That was a public announcement.

http://www.codeweavers.com/about/gen...s/?id=20050622

Jeremy Did state that they want to have it running under Aqua, though not in this statement. It was in an answer to a question. Will that come soon? Not likely. But that doesn't mean that it won't come at all. Remember that Crossover is a commercial product. If running it under Aqua gets them more customers, both OEM, and retail, that's an incentive.

You should go to the Darwine team, and contribute to that. I'm sure that they would be very happy to get your help (seriously).

I cannot agree with that at all.

Like I said before, difficulty isn't the issue.


That press release says nothing of the sort. Quote the part where they say they are developing on Mac OS X.


If an "OEM" customer wants to port their windows app, they would likely strip out the GDI if they didn't want to use X11. This means they don't need a Quartz driver. This is especially true of any apps or games with OpenGL or D3D interfaces, or just a simple interface which can be written in any other API.


Gee, why didn't I think of that

They are doing their work on Apple's developers kit. I suppose that's not enough? Even though they will initially have this under X11, it's still being done through OS X, evin if they Start by supporting Darwin through X11. From a June letter they sent out:

"and waiting for our dev kit to arrive )."

I suppose you'll figure out a way to say that it means nothing.


That's true, if they did that they wouldn't need a quartz driver. But the Darwine team is working on one anyway.


Narcissum perhaps?

There, fixed that for you. Hope the cluemobile didn't clip you as it sped over your head.

T, FTFY.

Learn a new acronym every day.

Forget one too.

/LANAED?

T, FTFY.

Your cluemobile hit a bump and went off the road.

A lot of people claim a lot of things.

Ok, enough of the sparring.

If true, this is a big one:

http://www.osx86project.org/index.ph...id=68&Itemid=2

It could be the reason why Apple MIGHT be able to intro a machine in January.

downloading it now will PM you kids if i ever get it up and running on my overclocked venice as i am in the far flung reaches of civilisation, it'll take about a week of bittorrent to get tha goods.

Better do it quick. Bit Torrent just agreed to take down ALL pointers to, uh, less than squeeky clean material.

Haven't they just agreed to remove movie-related material?

Yes, they have. I was putting it succinctly.


http://news.yahoo.com/s/ap/20051123/...NlYwN5bmNhdA--

I'm surprised Apple went the extra step to include G4 emulation. I would expect most CPU intensive apps to be properly ported.

AltiVec emulation of some operators will be slow (like permute) so I hope there is an option to revert to G3 emulation.

They agreed to not provide search functions for movies, and OSXonX86 does not qualify.

Here's a link to that article that will actually work:

clicky

It's a silly article anyway. I mean this is complete bullshit:

"... to remove Web links to pirated versions of movies from his Web site, bittorrent.com, effectively frustrating people who search for illegal copies of films."

Sure, very frustrating, because I always go to bittorrent.com--oh wait--I've NEVER gone to bittorrent.com!

Stupid journalist.

I pulled the link directly from their page, strange.

Well, millions do.

OK, maybe I'm using the terminology wrong or I'm misunderstanding something, but...

Windows running in a VM on a dual-core Intel machine is a very different beast from Virtual PC running on a G3/G4/G5 Mac.

In a dual-core Intel machine, one core could be assigned to OS X and one core could be assigned to Windows. There is no hardware instruction translation, since both OSes run *natively* on Intel. Both run essentially at full speed on their individual cores.

OTOH, with Virtual PC, Windows is running on an *emulated* Intel processor with an *emulated* video card, etc. There is a *tremendous* slowdown doing it this way.

So I just don't see the need for dual booting on a dual-core Intel with OS X Intel. You'd just have a separate VM for Windows. It would run at near-real speed, and you wouldn't have to reboot just to run QuickBooks or a game.

Am I misunderstanding something?

Correct.



You'd need some sort of a hypervisor (e.g. Xen), though. For purely compute-bound tasks, you'd essentially see native speed; however, there is some overhead for access to virtualized I/O and such.



When you say "VM", are you talking about somethig like VMWare, or something like Xen?

VMWare (and Virtual PC on x86) only differ from VPC/Mac in that there is no need to translate code from PPC to x86; the guest systems still use emulated graphics, sound and other I/O devices.

But even for a true hypervisor, you can't just let multiple VMs use the same video card (or any other shared I/O device) simultaneously. Graphics drivers for OS X and Windows are designed to take full, exclusive control of the graphics card, i.e. they expect to be able to access the entire VRAM, the cards config registers, etc. without any external restrictions. Allowing multiple drivers to access the hardware at will simultaneously would result in chaos, so you'd either have to provide a dedicated graphics card for each VM, or some kind of virtualization would have to take place (resulting in additional overhead).



I guess that would depend on whether a sufficiently sophisticated virtualization solution will be available for the Intel-based Macs. VMWare would certainly not be a very good choice for games.

I find it interesting that this all could have been done back when NeXT became os X . It's painfully obvious that os X is a good deal NeXT (home icon, window sidebar etc. etc.)... although it is color. (yes, I tried one of the cludgy betas- quite an interesting trip to see OpenStep on an apple)

The thing that intrigues me is back when everyone felt this should mean a move towards OS multiplatform compatibility, Apple wasn't biting. And now, they find themselves having to backpeddle and rework code, when this could have been done from the beginning when Jobs came in with *X (and a few rumors say it was "working" on intel in house). Mind you, I understand the hardware/software link... let's face it, Apple is a hardware company... they make their money there, and for the most part, their hardware is beautiful (minus the cluster of dead pixels in the left hand center of my 15" streaky high-res powerbook just purchased). So, they would have lost a lot by going multi-platform (we all remember the PPC compatibles, eh?)... This os could be running on every darned box on the market, if Apple wished it to be so. Obviously Apple doesn't, so it's not. However, they now have more work cut out for themselves that they could have been working on since X was first implemented.

Although, perhaps it's not important to have Apple on every piece of hardware imaginable to mankind: Where's the keyboard?

One intriguing question I have: how will the new intel DRM be utilized what with the thumb-screws being applied to Jobs as of late. [Kudos for him for making this work, but now he must fight back the zombies with sticks if he's going to keep it moving strong... for an almost cost-free distribution system, they should be happy with $1/song... especially with the DRM and lack of physical artwork.]

Alright, I rambled... I'm sorry... _Ducking from throwing stones here_

The fact still remains- from the old Apple ][e, I'm still here... guess I prefer the light flogging to the heavy flogging in the PC world.

RazzFazz is correct. But beyond that is the fact that you will still be running two OS's at the same time. As OS X, Windows, and at least most Linux distro's are symmetrical in their use of cpu's the OS will span both cpu's.

OS X works much better with two slower cpu's than it does with one faster one. It also splits services.

Now this is not inherently a bad thing when running a virtual service such as VMware or Xen (which I've never used). But you are not getting the entire computing power of the machine for either of the OS's in use. Even if you are not running any OS X user based program at the time, the OS is still doing it's own work. If you are not sure of what that would mean for usage rates, check the Activity monitor. Remember you are still using the services of X to run VM ware or Xen. That is overhead not incurred when running a dual booted OS.

Again, though, we don't know for sure whether Apple will let us dual boot. I think they will, from what it says in their patent. But, it might not be possible, so this may be moot.

Crossover, when it works, will not have this problem to the same extent, and WINE is truly a pain.

I've not played with VMware, Xen, WINE or Crossover so perhaps someone could answer this.

The advantage I see with running VirtualPC over dual booting, apart from the obvious ability to run two OSs at the same time, is that you can drag and drop files between Explorer and Finder and cut and paste between the two OSs. That IME is worth the price of VirtualPC.

For me it's secondary advantage is I've got three VPC sessions, each with a different browser installed. It's the best way to test IE compatibility.