Slashdot Mirror
Will MacIntel Hardware Open The Door for Mac OS X CAD?
xcleetusx wonders: "I've been a fan of Apple for years, and with their current strangle-hold on mainstream media my desire to make the switch has been growing ever more, but the same nagging issue that has prevented my switch for years still remains: I'm an engineer, and I simply can't invest in a computer that won't run modeling/simulation software like CATIA and Solidworks. Since this software is available on Unix (which Mac OS X is built on) and also on Windows (Intel hardware), is the Apple switch to Intel-based hardware going to better my chances for a Mac OS X CAD workstation, or will it remain a pipedream?"
Hopefully the switch means increased popularity which will lead to more support from venders. I'm an engineering student (ece) so i don't usually deal with cad and solid works but i deal with other stuff like spice and vhdl. Luckily eagle comes for os x.
Biggest complaint though is that most software that is "ported" uses X11. It's quite nasty.
----
Go canucks, habs, and sens!
Since Mac OS X will use a Darwin (Unixish) on x86, it will not be very hard to port your standard CAD programs to run natively in OSX. Mostly it depends on the demand in the market. If a lot of users start asking the CAD software developers for a port to OSX, it will probably happen. Short of that, your best bet may be Darwine or X11 for OSX. Using one of these may allow you to run standard CAD software without it actually being ported (don't hold your breath for Darwine, though).
What might happen that could help you is that virtual PC programs will be able to run MS Windows at near full speed since it'll be running on the same processor that Windows is written for. So you should be able to run a virtual PC program with Windows and your CAD apps on your Mac.
Slashdot: Failed Car Analogies. Amateur Lawyering. Anecdote Battles.
My best answers at this point are. . .
We can only hope.
-and-
We'll have to wait and see.
I'm a big user of GIS, and while I find GRASS to be a wonderful alternative to ESRI products, it's sometimes too much hassle to fire up GRASS, define a region, import files, etc., if all I want to do is edit a shapefile or query a feature. I do know about QGIS and other alternatives, but sometimes it would be nice if ArcGIS was ported to the Mac. Given the change in landscape over the past couple of years and changes yet to come (Intel, I'm looking at you), I think there's more probability of these sorts of things happening. They are, however, still possibilities. Until a company commits to producing their specialized software (CADD, GIS, etc.) for the Mac, or until there is an increased demand for Macs in such industries, we're still just speculating.
Now, if Apple manages to wedge their way into the server market with a killer Intel-based Xserve coupled with a low-cost Xserve RAID, we may see those pressures come from another side. Time will tell.
Per Square Mile, a blog about density
The only thing that is likely to happen with Intel-Mac is that Windows Emulators - and hence Windows software - will run at nearly native speed.
gopher://cramer.plaintext.cc http://cramer.plaintext.cc:70
MacOS is not going to magically turn into Windows or Linux just because there's Intel Inside. Mac development will be unchanged, with some marginal exceptions.
What I'm listening to now on Pandora...
...it'll run on OS X under X11 with fairly minimal porting effort today.
If the companies haven't made that port available then the (trivial, from an application developer point of view) change from PowerPC to x86 isn't going to change that.
It's all about size of market and differential pricing. Not the CPU that happens to be in the box.
If you can run all your software on unix, you can run it on Mac OS X. Mac OS X has the Unix Window System X11/Free86 built in so all unix programs will run fine. And you can run Windows on Mac OS X, and the software will just work. Microsoft sells Virtual PC with their Pro Office Pack.
Programs run on operating systems, not CPUs. Your best chance is if the new Apple/Intel hardware dual-boots, or if Apple gains enough market share that CAD companies decide to start coding for them.
Dear Slashdot: next time you want to mess with the site, add a rich-text editor for comments.
At least, it will give MacOS X users a chance to use wine to run Windows applications. Applications linked against winelib don't have a problem - if they were built for PPC arch- but you can't use the wine binary itself to start Windows applications on non-Intel arch.
In one word: no.
In a few more words: only if Mactel effectively replaces Wintel as the "default" office platform (which I don't see happen any time soon). Yes, the office platform. Why do you think the migration of CAD software from commercial unices to Wintel happened? Not because Wintel is cheaper in and by itself, but because Wintel is cheaper than a SPARCstation to run your CATIA and a Wintel to run a certain office suite. Oh, and because schools today are teaching Windows under the guise of teaching computers, and your average graduate would be totally lost if placed in front of a SPARCstation. That's the two reasons, mainly. Mactel will not change either of them.
Happy now?
'm an engineer, and I simply can't invest in a computer that won't run modeling/simulation software like CATIA and Solidworks.
You do realize that Solidworks is available for OS X, right?
rooooar
MacOS is not going to magically turn into Windows or Linux just because there's Intel Inside. Mac development will be unchanged, with some marginal exceptions.
I dunno, maybe this falls into the "marginal" category, but "scientific" [or "mathematical"] programming is really REALLY REALLY difficult.
For instance, take a gander at the list of FFTs catalogued at BenchFFT:
Then look at their relative performances for speed and accuracy: In many instances, the software and hardware engineers at the companies that build the chips [Intel and AMD] can't write FFTs that are as efficient as third-party vendors [or hobbyists], which should be a pretty good indication that something as ostensibly straightforward as writing an FFT routine is just fantastically complicated in practice.So if you're a company with a lot of low-level proprietary software that's tuned for the x86/x86-64 instructions sets, or for classical PCI bus communications [apparently PCIe is very backwards compatible], or for the nVidia/ATI/Oxygen instruction sets, then your porting job just got a heckuva lot easier if you don't have to deal with PPC RISC, Altivec, etc etc etc.
VectorWorks and TurboCAD are already here - and have been for many years. After several conversations with the AutoDesk AutoCAD procuct manager, it seems they think the Mac is dead, or dying. Anyhow, VectorWorks does handle AutoCAD files nicely, as long as you can get the M$ AutoCAD users (wiennies) to use standards-based file formats. CAD was born on the Mac... and I believe its still better with the available stuff today; AutoCAD is very much a M$ product - too much of everything, and nothing worth using day-to-day
As many people have mentioned above, the fact that OS X runs X11 means that if there was a market, there would be a version of your desired UNIX programs on the Mac right now. A good example is MATLAB. For Mac, it's basically the UNIX version with (I'm assuming) minimal changes since it runs in X11.
As for Intel chips, I agree that there will be no magic change that would all of a sudden allow your programs. Macs will still be Macs.
I'm a big user of GIS, and while I find GRASS to be a wonderful alternative to ESRI products, it's sometimes too much hassle to fire up GRASS, define a region, import files, etc., if all I want to do is edit a shapefile or query a feature.
Well personally, I find grass to be a wonderful alternative to sobriety. It's never too much hassle for me to fire up grass.
Something more interesting will be the ability to run x86 emulators (VMware, Xen, bochs) and have a full windows running at native speed. That should be much better than VirtualPC, and more flexible.
#include "coucou.h"
This is a viewer application, it's on Linux too. One can see data created via Solidworks on a win32 PC, that's about it. Well, measure, etc...
Solidworks is about as closely married to the win32 API as one can get. They stated this goal early in their development process and have not deviated one iota.
The whole integrated deal will keep a lot of MCAD off of Mac and Linux for a very long time to come yet. Microsoft is very aggressive in this area, working with vendors closely to interlink CAD with Office. The combination is good, however it will remain win32 as well. (Sigh...)
The older cad packages, that still have UNIX versions, are more likely ports. We have PTC on Linux today, it's not too much of a stretch to see OSX --provided there is demand. That's what all the vendors what to see. Tell them and tell your friends to tell them.
Better: Inquire about their software, get them to do a demo, then tell 'em you need it on the Mac. As they walk out the door, know they will be carrying that information with them to their technical marketing people. --Those are the folks that need to be sold.
Don't say Mac up front either... just keep the discussion about CAD and needs. The assumption will be win32. (It always is) Then drop the Mac bomb on them.
Blogging because I can...
The switch to x86 doesn't change the API of MacOS X and hence won't magically give you Intel PC software. And if that software had been cross-API-compatible (via Qt, wxwidgets etc.), it could have been released for PPC-MacOS already.
True.
The only thing that is likely to happen with Intel-Mac is that Windows Emulators - and hence Windows software - will run at nearly native speed.
Not strictly true. Everybody is concentrating so hard on the whole Windows emulator possibility that they're completely missing another benefit to x86 Macs that I'm personally looking forward to: Linux binary compatibility.
FreeBSD has had rock-solid Linux binary compatibility for years. Almost any executable compiled on and for Linux will run perfectly well on FreeBSD. Porting the Linux compatibility layer to Darwin is probably something that a skilled dev can do on a rainy weekend. And that's if it hasn't been done already. For x86 Mac users, this immediately opens the door to almost all programs built for Linux, both open and closed.
I say to the fellow who wants his CAD software on Mac: You'll probably waste your time pestering the vendor to release a native OSX version of the application. And WINE is unreliable at best, which x86 OSX won't change. What you want is to be able to run the Unix version of the app natively on your Mac and that's what Linux binary compatibility will do.
Florian is correct. The differences between Mac OS X and Windows is probably a bigger factor than the differences in PPC and x86 CPUs.
Last I heard from the Intel OSX leaks a while back, you could just install Windows on the systems with no problems. I don't imagine there would even be any legal issues with this, as long as your Windows copy is legit.
So yeah, I'm guessing you'll be able to buy a Mac, buy a copy of Windows, and with a bit of fiddling, install it so you can choose to boot to Windows.
Even if this isn't possible, I'm sure programs like WINE will be running a LOT faster with the Intel Macs, so you could probably run your Windows programs through those.
But the apps compiled for Linux on MacOSX will need X to run; no X, no GUI. That will be a major stumbling block; likely if X is already available then the App vendor can just recompile the app for Mac OSX already.
better not tell the people at Architosh.com that there is not a MacOS CAD workstation
> And WINE is unreliable at best, which x86 OSX won't change.
x86 OSX will roughly double the wine user base, dunno how much difference that would make for wine development, but i'm sure it's more than nothing.
and another thing: companies that make applications for windows are used to target multiple (sometimes quite different) versions of that OS. agreed, "system requirements: windows 2000/XP/vista or wine vX.Y.Z or newer" does sound utopic, but not _that_ utopic.
[i have an opinion and i am not afraid to use it]
The differences between Mac OS X and Windows can be handled by an API layer like wine, the differences in PPC and x86 CPUs can be handled by CPU emulation like virtualPC. Both are complex tasks and likely to never reach perfection, but there is a huge performance in performance hit.
[i have an opinion and i am not afraid to use it]
You mean, like this?
Except that it is, and they don't.
"Since this software is available on UNIX (which Mac OS X is built on) and also on Windows (Intel hardware), is the Apple switch to Intel-based hardware going to better my chances for a MacOS CAD workstation, or will it remain a pipedream?"
If the software is available on UNIX, and is not available on the Mac right now, then whatever is holding it back is unrelated to the processor the Mac is using. Either the vendor does not consider the Mac market large enough (which is odd, since by this time the majority of workstations capable of running UNIX software are Macs), or they consider even a port to another UNIX platform unreasonably difficult, or they don't realise that Mac OS X runs ordinary UNIX applications very well.
These are not problems that will be solved by switching to a new processor, case design, color scheme, mouse, keyboard, monitor, or pizza topping.
That's *exactly* the sort of thing that I meant by "marginal".
I Am Not A CAD Developer [IANACD], and, for that matter, I Am Not Even A CAD User [IANEACU].
On the other hand, I know a fair amount about LabVIEW, which shares a great deal in common with CAD environments, and I know that National Instruments has a budding problem on their hands because their graphics package depends on OpenGL/MESA, and Microsoft looks to be deprecating support for it.
But it would be interesting to hear from some CAD developers - do you write low-level stuff in-house, e.g. do you write triangles directly to the GPU, or do you purchase development environments [PIXAR, DirectX, OpenGL/Mesa, VRML, whatever] that perform the low-level translations for you?
Programs run on operating systems, not CPUs.
Nope, they run on CPUs also. Operating systems do to. Operating systems and programs are both software. They both run on CPUs. The operating system schedules what programs get to run when, and when the OS itself runs, but everything happens on the CPU.
Now what you may have been trying to say is that programs are built to be run with certain operating systems, which would be correct.
You may be right, in that the change of processor will not mean a magical porting of an application that was not previously ported to Cocoa under PPC. I agree with that. However, a Mac with an x86 processor may increase the probabilities of a user using it as a dual-boot system. So, you get to experience the wonders of OS X Tiger, and you keep to use your good ol' engineering software under Windows or Linux.
uhmm Shakespeare coined over 700 of those "old sayings" & words out of whole cloth, he just stuck the damn things in there and brazened it out. The fact that many of them are still around in our language attests to his creativity and chutzpah :)
The switch to x86 doesn't change the API of MacOS X and hence won't magically give you Intel PC software.
Right. But...
Virtual PC will run a whole lot faster.
"I am the king of the Romans, and am superior to rules of grammar!"
-Sigismund, Holy Roman Emperor (1368-1437)
This is not true of all sofwtare. For software that does a lot of math calculations, it may use code aoptimized for SSE2 etc. They may ahve an unix version for x86 linux/bsd, but not PPC. This version may even use a crospplatform AIP
like QT. Porting to macOS PPC would take a lot of rewriting. Getting it to run on xwindows on intel macOS might be trivial, as it's the SSE2 parts that are nonportable.
You assume that the software is windows. If it's unix for intel, theer should be little difficulty porting it.
RJL
There's a saying, perhaps you've heard it ... "God helps those whom help themselves." Translations: quit your bitching and write it yourself.
Do you even lift?
These aren't the 'roids you're looking for.
Hopefully the switch means increased popularity which will lead to more support from venders.
The switch will mean OS X will be easily pirated. Apple's whole plan is predicated on something Microsoft has known for years: piracy = marketshare. No matter how you slice it, people who otherwise wouldn't have bought an apple machine will download and install this on some machine or another, even just "because they can". Apple knows this.
When they release OS X for x86, you can expect a huge jump in marketshare from the current ~2%, simply because people will be torrenting this thing like crazy. (as if they aren't already)
The theory of relativity doesn't work right in Arkansas.
Don't know about SolidWorks, but as far as CATIA is concerned, it runs on different processor architectures, as well as different operating systems (Basically Windows and Unix-based). I do not see how a switch of processor might have anything to do with paving the way for a Mac port.
It is likely that if they really wanted to do so, they would already have done it already.
What makes you think the switch will prompt Dassault Systemes to port CATIA to OSX?
:\), and it's one of those pieces of software that keeps me from switching to OSX (Valve's Source is the other big one).
They already support AIX on POWER and PowerPC. Given that they haven't ported a program that runs on AIX/PPC to OSX/PPC, what makes you think they will port a Win/x86 program to OSX/x86?
It's not about obscurity. CATIA runs on platforms with tiny marketshares like HP-UX (on PA-RISC?), Solaris on Sparc, and IRIX on Rx000. The software is obviously very portable, DS just has no interest in an OSX port.
I'm a huge fan of CATIA (just reupped my license a week ago
Versacad was available on the Apple II, and followed Apple to the Mac and was available as early as the Mac Plus, I think. I used it many years on a Mac II and later machines. Then it disappeared for awhile, but is now back. Its available from a company run (owned?) by the original designer/programmer, at www.archwaysystems.com. I transitioned to something else, and no longer use it, but its good for 2D for sure.
Not yet, anyways.
MOL works because the G3/4/5 are fully virtualizable. You can trap any instruction you want, allowing you to properly virtual an OS within another OS.
x86 is not the same way. VMWare solves this by doing some really nasty tricky stuff. It's not an easy problem by far, and the performance would suck balls because you have to check all the code before it's executed, slowing everything a whole lot.
--
Internet Explorer (n): Another bug -- that is, a feature that can't be turned off -- in Windows.
CAD on the Mac depends a lot on what you are doing- for Architecture Vector Works is quite viable, for solid modeling it's very weak. I've been using Ashlar's Cobalt http://www.ashlar.com/ under OSX and Windows for about 18 months, and for what I do, tending towards industrial Design, Injection molded plastic parts, prototyping, etc. I like it much more than SolidWorks. You get both Windows and Mac versions for one price, and the license allows you to switch back and forth, which I do frequently as in order to export Parasolid files for SolidWorks users you have to be under Windows. It runs at least as well on the Mac- better really. They offer a 2 week fully functional trial- after using it for about two hours I was scrounging up the $4000 for the full package. I only open SolidWorks now when I absolutely need to check an exported file for someone. Pros: much better interface (Mac or Windows), more fun to use, better tools for creating complex organic shapes Cons: fewer tools for very complex assemblies, less than fully realized parametric tools, lack of many of the tie ins to specialized software like FEA, etc. The Rutan / Scaled Composites/ Spaceship one supposedly group used Cobalt for the airframe design, SolidWorks for the landing gear- not sure if this is entirely true, but it does summarize the relative strengths of the systems). There are persistent rumors of ProE coming over to the Mac platform- That would be a very good thing. RonR
What you want is to be able to run the Unix version of the app natively on your Mac and that's what Linux binary compatibility will do.
Yes, users want to run apps natively, but running from the command line or X Windows is not good enough.
To blend in with the system, it must look and feel like a Mac app , not only superficially. Users will expect such things as:
In short: users want a Mac app. If they were happy running a Linux app, they would be running Linux already.
The chips Apple uses don't matter. Commercial vendors want to know if there's money to be made by adding a new platform. If the move to Intel expands Apple's marketshare, then your chances for getting your CAD programs goes up.
-- Slashdot: When Public Access TV Says "No"
http://www.architosh.com/features/2005/features/a
... is the Apple switch to Intel-based hardware going to better my chances for a Mac OS X CAD workstation, or will it remain a pipedream?
My wife lost her ability to see the future, and my friend who can read the minds of distant CEOs an product managers isn't here right now, so I guess I'll have to resort to my magic 8-ball:
"Better not tell you now"
"Alcohol, Tobacco, Firearms, and Explosives" should be a convenience store, not a government agency.
I don't know if any nice packages will be ported to OS X for Mactel, but since the dev boards are basically standard x86 mbs, it is likely that any OS that will run on a Dell will run on the macs (although OS X seems to be copy protected to only run on the macs). So even if CAD software won't run on OS X, you can still dual boot into Windows or Linux and run it there.
It's been my understanding that X11 has been working on OSX since day one or earlier. No, it might not come with the default OSX installation, but it is available.
dual boot please
The movie titles were old enough to be public domain, so that dosen't really matter either.
I have freaks! I did something right...
I have to say this:
I don't mean to brag, I don't mean to boast, but I like hot butter on my breakfast toast.
Y'know, the old school kind.
I too would love to see better GIS on the Mac. GRASS is good, and the price is right, but when you need to get down and work sometimes you want what you paid for to run on your pretty workstation...
EOM
For years I too have wished for a proper CAD system that could run on Mac hardware. However, at this point I'd settle for being able to teach everyone out there that 2D CAD is a pale, shallow, crippled, painful excuse for CAD as compared to any mainstream parametric 3D CAD system.
OK, I'll concede that there is CAD for Macs, but it's a glorified electric drawing board, not a useful engineering tool. THAT is what I want to run on Macs... The day that Pro/ENGINEER, or SolidWorks, Alibre Design, or any of those ilk (and quality) can run on a Mac, THEN I will be one happy design engineer. Until then, there is no CAD for Macs.
Mr. Ska
Some years ago I actually dumped my ROMs from my Ebay Mac IIcx before I scrapped the thing. I cut off the ROM chips with a dremel tool and have them in my desk drawer even now. I have fooled around with the Basilisk ][ macintosh emulator using my ROM dump and I must say its pretty nifty. I actually used it to read a Mac CDROM in my PC and copy a datafile over to the hard drive.
I'd gladly install OSX on my PC if I could purchase it at a reasonable price. I've always admired Apple software, its simply a matter of economics in my case.
Clickety Click
it makes you woder what CAD software macs are made on...solidworks on a PC platform?.. wonder why they dont ask themselves this question: should we not make a parametric solid modeller for mac osx. One that actually matches the industry standart.