I got it to work with my PowerBook G4 (DVI), D-Link USB Bluetooth adapter (any USB Bluetooth adapter should work), Sony Ericsson T68i, and T-Mobile service.
It wasn't as straightforward as it should have been - however, it might be a settings/firmware problem with the phone. Apparently T-Mobile presets certain settings in the phone which you as an end user can't change. Phones sold by Amazon recently have a problem with WAP access. I also could not use the built in T-Mobile Internet Access Data Account (using CID=2).
I ended up calling T-Mobile's Data Support Service at 1-800-256-9991 and they walked me through using CID=5. I did have to modify Ross Barkman's Ericsson GPRS Script to call using CID=5 instead of 3. Here's the mod:
where it says write "AT+CGDCONT=3", change it to "AT+CGDCONT=5".
I was doing this with the T-Mobile tech on the other side, so I don't know if he had to fiddle with your T-Mobile account settings. Best to call them if you're having difficulties. Everything else was the same as you could easily find in the recipes that other people have posted.
I have also helped a client get it to work with AT&T's mMode service - the toughest part is usually getting the right APN, or knowing to use none.
It was refreshing to be able to contact a service (T-Mobile) and they didn't immediately give up when they learned that you are using a non-Windows platform. AT&T, however, was pretty much "we can't help you since you're not running a supported platform" and even that was narrowly defined.
A PowerEdge 350 is less than half the price (~US$1300ish) of the basic Xserve ($US3000) even after bumping its options to within the same ballpark (1GHz processor, 80GH HDD, 256MB RAM, GB copper ethernet).
Ah... wrong.
The PowerEdge 350 can only take 1gb worth of PC100 RAM (vs. 2gb of DDR on the Xserve), can take only 2 drives (vs. 4 in the Xserve), has no built in firewire, has only 10/100 NICs and one free 32 bit 33Mhz expansion slot (adding a gigabit NIC takes up that slot). The Xserve has dual gigabit NICs and still has a 64 bit 66Mhz slot free. Finally, the 1Ghz G4 with 2gb of L3 cache is much faster than a 1Ghz PIII on most tasks - and really blows it away with Altivec/Velocity Engine aware code. Depending on exactly what you choose (since the two can't be matched exactly), the PowerEdge 350 might be slightly cheaper running RedHat Linux, but would be more expensive running W2k.
The PowerEdge 1650 is closer, but again it mostly comes down to RedHat Linux vs. W2k on the price.
And don't forget these machines are at the bottom end of the server market. If you want more oomph Dell have a whole slew of machines you can move up to. Apple have nothing.
Yes, it would be nice to have a smoother transition to something bigger - switching to another platform with different tools can be a pain.
It's not a bargain. It might be interesting to parties who don't want to employ a Unix admin to look after their workgroup server, but for more mainstream usage, where a real admin will be running the box and the software niceties of OS X are, at best, of little importance, it's a damn expensive piece of hardware.
I respectfully disagree. Even the most advanced, most professional system administrators can benefit from the time savings from the various tools in Mac OS X Server. But it's not for everyone, and not for every case. It is a good value for many situations and is definitely better than the low end Sun and IBM stuff. It is very competitive against the x86 stuff even running Linux (especially against a commercial server version of Linux).
The new top of the line iBooks have a 700Mhz G3 with a 512k full speed on-core L2 cache. That gives a big speed boost over the older generations of iBooks. However, as compared to a PowerBook G4, it sports only a 100Mhz system bus (versus 133Mhz), a Mobile Radeon vs. a Mobile Radeon 7500, and the PowerBook as a 1mb L3 cache not to mention the G3 vs. G4 thing. The PowerBook G4 also has built in gigabit ethernet.
Therefore the PowerBook G4 can be substantially faster than the iBook if you are working with a substantial memory footprint, rely heavily on OpenGL, use applications that can leverage the Velocity Engine/Altivec unit or doing big network transfers over gigabit ethernet. Otherwise, the performance different isn't that big - especially comparing operations that are more disk I/O bound (of course). The iBook may even win a few tests with it's larger full speed L2 cache.
As for your Thinkpad, well, both the iBook and the PowerBook G4 would probably be pretty competitive from a raw perspective. However, the performance is going to probably be very different depending on what you specifically are doing - thus w/o additional information on exactly what you are going to do with the machine as well as what your personal perspective on what is important in terms of performance, it'll be difficult to judge.
The PowerBook G4 really has many additional features, and you have to be comfortable with the sacrifices that are made to make the iBook smaller and cheaper.
I'm very happy with my PowerBook G4 800 DVI and it is a substantial improvement over the PowerBook G4 500Mhz machine I used to use. I'm hoping the next update is more of the same - maybe 933Mhz+ and built in bluetooth. It would be nice if it had better 802.11b range too.
Now this is just insane. I've used A/UX on a Mac IIfx. There was almost _no_ integration between the Mac and the UNIX sides. You were basically running the equivalent of take-over-the-screen Mac OS Classic environment in a relatively hostile and non-standard UNIX host. There was no Cocoa. There was no Carbon. There were no Mac OS GUI applications that actually were UNIX processes. There wasn't any UNIX API exposure to the Mac side. So don't talk about "pretty-much on par" integration.
the G4 is doing much more than the IIfx
Ah, yeah. The two aren't doing the same thing. That's like complaining that Enlightenment doesn't run as fast as twm. Plus, you're also talking about some disk I/O performance related things. Let's see... the high performance SCSI disk (for the time) in your IIfx was how fast? The laptop ATA drive in your G4 PowerBook is how fast? Hmmm... not much difference is there?
In fact, i use the IIfx quite a lot as it provides a cheap and usable X-Terminal with more character and a much quieter fan than my assortment of x86 boxes.
You can build a cheap and nearly silent x86 box. Really.
Sure, A/UX was never a mainstream OS for Apple, but I wonder what might have been had Apple not abandoned the 'UNIX way' all those years ago.
Of all things, A/UX is and was not the beacon of the future.
You can choose the solution that you like best and nobody will know*.
*Except that one of the two results in apps that start faster and can run on an old OS that still has a big following.;)
However, when Carbon apps are busy, the user can't move any of it's windows. Very, very frustrating. Also, Carbon apps not written specifically for Mac OS X have definite disadvantages and are pretty easy to spot. Just look for the most ugly and non-conforming applications except for the pure Java ones.
Plus, it's usually the Carbon apps that don't properly leverage the operating system or contain lots of cruft like 31 character file name limits or have constantly running loops that hog CPU time and consume battery power, even when they're supposed to be totally idle. It's usually the Carbon developers that can't quite "get" daemons and proper modern application architecture. It's a good thing for them that there's a huge installed base out there that are almost as backwards thinking as they are to keep buying their products. A lot of the rest have moved onto Windows or something else.
I would give Gefen a serious look. Their DVI to ADC Conversion Box is the only one on the market that I know of that claims support for the 23" HD Cinema Display. Apple's, Dr. Bott's, and the Formac converter boxes all disclaim use with the 23" HD display. If nothing else, they should be able to help you with selecting the right graphics card.
One of the biggest problems with Mac OS X is the constant bowing to the classic Mac OS crowd. I think it could have been a much better platform if it had less shackles from an operating environment that has not truly shed it's 512x384, single tasking, single user, system modal panels, non-networked roots.
I could use all my Mac apps in Classic mode in OSX, but I still spend 99% of my time in OS9.
Ah... so you really haven't adjusted yet, huh? I remember with Mac OS X first shipped in March this year, the huge amount of venom from the aged Mac user crowd (never mind that many of the ex-NeXT users were Mac users before using the NeXT). But now, I'm seeing lots of stories of how people really like the Dock now, and are frustrated when they occasionally use Mac OS 9!
The Classic Apple menu is a prime example. How many novice users even know they can modify it? How many know how to modify it? What happens when you turn on Multiple Users and need to change an item globally? Why is there a separation between the active apps menu and the Apple menu - thus you have two menus entries for say, AppleWorks? I could go on.
Mac OS X, with it's preemptive multitasking and rock solid virtual memory system, as well as it's multi-user features and the like should force a re-think of how a user interacts with the system. The problem with Mac OS X is that it the new system isn't finished, and if anything, it should be less like Mac OS 9.
The original idea when Apple went with NeXT was that Apple would ship essentially OPENSTEP/Mach for PowerPC. The early Rhapsody Developer Previews were essentially that, and were available pretty quickly. Apple had to dust off the old NeXT PowerPC port and bring it up to speed and port it to Mac hardware, as it was originally written for the NeXT RISC Workstation that never shipped (I've seen a prototype of the m88k version, but I haven't the PPC version).
The problem with that strategy was that the major ISV's balked at the idea of porting to OPENSTEP API's. They saw it as a lot of time and expense for a platform that might not last out the year. It would not have required a total re-write as some people have suggested, but certainly it would have been a major effort. (I would personally argue that going the Carbon route was also painful and going the Cocoa route would have resulting in a better product). Plus, these ISV's would have to then maintain separate ports for Mac OS X and Mac OS X, and they weren't willing to do that - many of them had already cut out ports to anything but Windows and Mac, and were probably considering dropping the Mac anyways.
So the Rhaspody strategy was abandoned, Steve Jobs took over, Apple re-invested in the traditional Mac OS and got some good releases out the door. They also came up with Carbon, which is a re-tooled Mac Toolbox API that sits native inside of Mac OS X. In doing so, they also re-wrote the graphics layer, removing Display Postscript and replaced it with brand new code called Quartz which is based on PDF. That means re-writing the window manager as well so that it supports simultaneous display of Quicktime, OpenGL, Java2D, QuickDraw, and so on including using underlying hardware support. They also re-wrote the DriverKit layer, replacing it with IOKit which is embedded C++ based and has much broader support. The print system was replaced, the Workspace Manager was tossed and the Finder was re-written in Carbon (IMHO one of the worst parts of the current Mac OS X). Lots and lots of utilities were re-written, the BSD layer was upgraded from BSD 4.3, the kernel was moved from Mach 2.0++ (2.5 and some 3.0 extensions) to Mach 3.0++. The Classic layer was also added so that it can mingle with native apps, Java was added, Mail.app was re-written, and so on and so on. There was a lot of work put into this operating system since OPENSTEP 4.2 for Mach, which basically remained stagnant for years.
In the meantime, Rhapsody did essentially ship as Mac OS X Server 1.0 in 1999. It was basically OPENSTEP 5.5/5.6 with a menu layout that was Mac OS Classic-ish but pretty much everything else was straight from OPENSTEP/Mach.
So... any operating system that Apple might have chosen at the time would have had to go through the wringer in order to get it to support Apple's technologies and what Apple perceives as what their customers require. It would have taken a long time, and BeOS would have been a worst choice in terms of both adapting the technology and the personnel. I think that going with BeOS and C++ would have led back to the Copland and Taligent quicksand pit. As for personnel, if Steve Jobs didn't come back to lead Apple, I'm not sure Apple would have had this resurgence.
Apple's stock is in the tank. <P>So are alot of other high tech companies. Plus, Wall Street isn't known for smarts.
<P><I>Preliminary indications are that Apple users are not particularly interested in the complexity and sluggishness of Apple's latest operating system.</I>
<P>Incorrect. Some are scared, just as many people are scared of the unknown. It's FUD, but they'll get through it.
<P><I>Processor speed is stuck at 500 MHz.</I>
<P>dual 533 is shipping this month, 667's and 733's in March.
<P><I>Alternative architectures and software are killing Apple on features, price, and performance.</I>
<P>Companies and platforms compete. Apple is competing better now, and will compete even better this year and next than it has years ago.
<P><I>There are legions of corporations and individuals who have been disrespected by Apple--from the BeOS community to the Apple clone industry, all of whom comprise a formidable enemies list.</I>
<P>Hmmm... might you be on that list? Did Apple hurt your little feelings? That would explain your use of errors.
<P><I>The main provider of Apple's microprocessor, Motorola, is hurting and hopes to leave the desktop processor business.</I>
<P>Motorola as a whole is hurting, yes.
<P><I>Today Motorola announced 2500 layoffs.</I>
<P>How do layoffs at a cell phone manufacturing plant have anything to do with Apple and the PowerPC?
<P><I>The most reasonable solution would be for Apple to open up. Open up its hardware specs and software so that where now exists little more than a corporate cult, there might exist a vibrant autonomous industry of developers, hackers, and alternative hardware vendors.</I>
<P>Apple is in transition. Darwin, the public source umbrella of Apple's software projects includes the Core OS of Mac OS X. Guess what. You can port Darwin to any PowerPC ABI architecture you want and you'll be able to run Mac OS X on it. Most of all, Apple has to protect itself first. After all, it is a business.
God knows I'd switch if I could afford the @(#%&@)*#$& Mac hardware.
Expensive??? In 1989, as a very poor student, I scraped together $2500 for a 386 system. In 1991, still a poor student, I scraped together $3200 for a NeXTstation (at edu discount) and another $1500 for a 400mb hard drive and 16mb memory upgrade. I saw this as an investment in my future and it has payed off.
In today's dollars, that much money would probably buy a Sun Ultra 60. Plus, we have today's ludicrous salaries for people with no degrees and no experience, why are there so many people whining about the costs of a Mac? $5000 is what one paid for a decent entry level UNIX workstation, and I gladly paid the extra over a x86 PC for the difference in quality and capability (both hardware and software).
What's really expensive is time. And Macs can often save time and effort, as does quality Sun servers and the like. An hour of time savings == $150 to me. I can't tell you the number of freaking hours I've spent with what is supposed to be high quality x86 hardware (as I gesture toward the database server box based on a 1.1ghz Athlon, 1.5gb PC133 ECC SDRAM, K7T Master motherboard w/ on-board Adaptec 789x U160 SCSI, etc. right next to me for a client - and I've used the name brands too). You would think that starting salaries were under $20k for all the whining about price I've seen on SlashDot...
It's true the Power4 isn't designed for graphic artists or desktop users. The cost is also not in line with that market. However, Apple _could_ use the discards. Power4 chips where one of the CPUs is non-functional. It will still be fast. It will still be expensive. However, it won't be nearly as expensive, and IBM would have tossed those chips anyways. This way IBM recoups more $ per wafer. As long as the non-functional CPU can be disabled and not use any power, I think a 1 way Power4 would be a great personal workstation CPU.
Another approach is to just port Darwin to RS/6000's.
First, it was my impression that IBM had spent a great deal of time on Mach 3.0 as part of OSF/1 as well as other projects like a replacement kernel for OS/2 and a new OS for the RS/6000. There must be a great deal of Mach experience at IBM that you could have drawn from (I say this, not knowing the political situation at IBM).
Since Apple's Darwin is based on OSF's MK7.3 base which includes most of what is considered Mach 4.0. Since Darwin is the core of Mac OS X, and mixed with what IBM has been doing with Mach over the years, you might have had a really nice combination. Linux is just another UNIX.
There are a few places where the the basic Mach kernel is weak for instance no support for drivers as modules, lack of driver support
Apple's IOKit framework helps solve that for you. It's a object framework for writing drivers, and yes, it does support drivers as modules and supports things like USB, IEEE-1394, among other things. It's based on embedded C++ and it's pretty slick.
messaging is pretty slow & used for everything, great for building clusters but overkill otherwise. As mach uses seperate address spaces for everything this improves protection but decreases performance.
Apple's take on this is to keep the semantic separations, but compile BSD into the same address space. Thus, instead of RPC calls into the kernel, they fold into function calls. You keep the abstraction but don't loose nearly as much in performance. Of course, you then couldn't crash the upper BSD/Linux layer and keep the kernel running then.:-)
There are multiple BSD projects, and certainly forking yet another isn't a big deal. However, I think a combination of Mac OS X/Darwin + RS/6000 or better would be a fantastic combination (drool).
I took notes from the Mac OS X Kernel (Darwin) presentation at Apple's World Wide Developer's Conference 2000.
On whose authority? Just because Job's said so? To be fair, the benchmarks jobs was running were crooked Photoshop benchmarks that were solely optimised for the Mac, if he run a standard benchmark then I'd be interested in the performance.
Of course, for many of the Mac power users, Photoshop is *the* application where they need as much performance as they can get. Plus, it really wasn't that crooked, since they ran a "real world" application with a "real world" use of that application. It wasn't a synthetic benchmark, but doing something graphic artists need to do on a daily basis. Now, extrapolating that result to general performance is the problem. But since most of Apple's pro market does graphics of that kind, this is relevant benchmark.
Also, during Job's presentation when he introduced the G4 SMP, he just said "we're going to compare it to a 1ghz P3", great but I'd like to know more details than just "a 1giz P3", which chipset? how much memory? the graphics card? the hard disk type etc? You get the picture; it was hardly a objective test.
It was an off the shelf machine... not a custom built machine. So you can get specs from the OEM. Plus they did state they had outfitted them as close as possible to identical specs.
Also, you have to remember you could buy 3 or 4 Athlons for the price of a G4.
Are you talking CPU's or machines? G4 chips are cheaper than Athlons (in quantity), but Apple does have one of the highest margins in the PC biz. However, 3-4 vs. 1? I think not.
The big difference is that you can go cheap on PC components... decide instead of buying a rock solid Socket A motherboard like the Asus A7V, go buy a Biostar or something like that for 30-40% less. Same goes for components like RAM, hard drives, etc. While Apple may not use the *best* components, they use some very good components (like the latest IBM 7200rpm ATA drives). We could build almost identical (by spec) 1ghz Athlon T-bird boxes, but yet be several hundred dollars or more apart in price just by changing the quality level of some of the components.
BTW, the cheapest dual 500Mhz G4 box you can get from Apple goes for $2449 through Build To Order. At that price, you still get two IEEE 1394 ports, gigabit ethernet, DVD-ROM, and of course, the ability to run Mac OS X (definitely worth the price premium).
I've been wanting to set up a *n*x box at home for some time. I've been looking at all the alternatives: NetBSD or mkLinux on my old Centris, partitioning the drive on my G3, finding/building an AMD box for cheap.
Why not go ahead and partition the drive on your G3? If it is a New World architecture Mac (Blue and White G3 and newer), then the Darwin binary installer will even install into a partition for you (don't need to wipe a drive).
A new/old AMD box is going to cost more time/effort than partitioning your G3.
AppleWorks 6 is a Carbon app. If you have DP3, try running AppleWorks 6 and see what you get. <P>Of course, the merits of AppleWorks 6 is debatable... I don't use it much, but it appears that there are a number of ClarisWorks and AppleWorks 5 users that are not too happy with it. Probably a point release will help quite a bit. <P>Anyways, there is at least one "works" type suite available for Mac OS X on the ship date. I would bet that Apple is working on Carbonizing all their various non-Carbon applications (iMovie, Final Cut Pro). The AppleShare IP and Mac OS X Server functionality will probably also be ported, but most that of that will be to the BSD API's. The admin apps will probably be a mix of Carbon/Cocoa and web based stuff. <P>Various heavy weight 3rd parties in the Mac space have already publically committed to some Carbon ports at least. A whole parade of them were on stage at the last MacWorld Expo in January - Adobe, Macromedia, Microsoft, id Software, Quark, etc. Adobe actually showed a preliminary Carbon port of one of their apps (Photoshop or Illustrator) on stage at function last year. <P>By the end of this year, many people won't need to fire up the Classic environment except for games (whose pressure to port to Carbon seems to be less).
Reading the A/UX FAQ, A/UX isn't based on a Mach kernel (which is good or bad depending on your personal viewpoint on microkernels). Also, A/UX wasn't intended to replace Mac OS. Mac OS X (10) is a replacement for Mac OS and therefore the integration with the Mac OS world is far more complete. A/UX was pretty much SVR2 UNIX heavily extended with a Mac OS shell. It doesn't have Cocoa or Carbon apps - Photoshop, Freehand, etc. aren't full blown UNIX tasks in A/UX, but the Carbon versions of these will enjoy the virtual memory and pre-emptive multitasking features right on the Mach/UNIX layers in Mac OS X. This will be a UNIX derivative that is intended for your computer newbie to use.
Apple may have the highest shipping volume desktop UNIX-like system in rather short order. Apple has stated that they intend to completely convert over to Mac OS X by the beginning of next year... all shipping Macs will come with Mac OS X then.
yes, dual power supplies in G4's Re:That's a reaso
on
Rack An iMac
·
· Score: 1
Since then, Apple's been particularly weak in the server space (even more than usual) - throwing OS X Server on a 3 slot G3 or G4 with only one power supply and calling it a server don't make it one.
Actually, the current Apple Power Macintosh G4 Mac OS X Server machines have dual power supplies... and a powerd daemon to go with them. It's not hot swappable, however. It is instant switch-over.
Also, there are many servers out there that don't need more than 3 64bit slots. Matter of fact, many rack mount systems sacrifice slots - leaving only one or two available. With built in 10/100 ethernet, built in firewire (for that brand new firewire SAN to be supported in MacOS X), 2 USB ports, what exactly do you need more than 3 64 bit PCI slots for anyways? Let's see... gigabit ethernet card, dual channel 64 bit PCI Ultra/160 SCSI card, what else? I know that some people do need them, but not usually as servers and certainly not in this price category (under $10k). Plus, PCI expansion chassis's also exist.
In general, Macs do make nice solid quality machines compared to most x86 machines - not as solid as real Sun boxes (not the Ultra 5's), but then the price is a whole lot cheaper. It takes a lot of effort (which can mean dollars which is often discounted) to identify, specify, and purchase quality x86 systems. It is possible, but with vendors changing components all the time, it can be difficult.
DMDx86 wrote: Yep. This is exactly how Apple pulled off the "No G4 Until we say so" stunt. I was quite close to buying a Mac for Linux purposes, until I heard about this. If I can't upgrade my machine as I please, I do not want it. Now I am buying an Alpha instead..
The "no G4 until we say so" was implemented in firmware - there isn't a Mac OS ROM on a chip anymore - the MacOS ROM is loaded from disk on all recent machines. The G4 firmware block is suspect... but on the balance, not as bad as it initially sounds. The Blue and White G3 motherboard and firmware wasn't designed to work with a G4. Apple doesn't produce a G4 upgrade for those machines, doesn't plan to, nor touts that machine as upgradable. 3rd party manufacturers have already worked around the issue and so you can get a G4 upgrade now, but now it is obvious to all involved that Apple has no responsibility for the results. After all, if there is a data corruption or stability issue, Apple doesn't want to be on the hook for it.
Personally, I wouldn't upgrade to a G4 in a Blue and White G3. If you want to upgrade, buy the latest G3 chip or sell the machine and buy a G4. The new UMA chipset G4's have much higher memory bandwidth - it doesn't make sense to drop it into a Blue and White machine which still uses a 60x protocol. Plus the G4 upgrade pricing is too expensive (compared to other PC's, not Alpha's or UltraSPARCs).
IMHO, this issue is a poor one to use when deciding between a Mac or an Alpha for Linux. There are tons of other technical and price issues to work with first which should decide the point. How much are you spending? What kind of support are you expecting? What applications to you expect to use? Do you need to run commercial applications or do you have source for everything you want to run? What kind of upgrades? What level and kind of performance are you looking for? What peripherals do you need to get? The Mac and the Alpha don't really compete against each other since the answers to the above questions can be very different.
A/UX and Mac OS X are on opposite ends of the UNIX spectrum. Yes, both look and act like UNIX at a certain level. After that, things are different.
A/UX doesn't have: Cocoa/WebObjects Quartz/Aqua/OpenGL Carbon transparent (rootless) Classic Mac OS support Darwin (open source underpinnings) PowerPC support Mach transparency to UNIX (Mac OS X won't require UNIX knowledge at all) and all the various components built upon the above.
In A/UX, there was no bridge between the Mac OS environment and the UNIX layer. There was no future migration strategy or technology to get off of classic MacOS. It was basically a UNIX for 68k Macs that happen to run Mac OS in a virtual machine in take-over screen mode. Kinda like what the Mac On Linux project would get when it done on LinuxPPC. Interesting for its time, yes. A future for Mac OS, no.
RFC959 wrote: I don't know what graphics system OSX uses, but I suspect it's either X-like or MacOS-like. We're talking about multi-user machines here, remember? X is what you wanted to get rid of, and MacOS is built with the assumption that you've got one machine, one framebuffer, and one user (and one GUI!) in your "computing environment", none of which is necessarily true anymore, making it an unsuitable starting point.
You're right. You don't know what graphics system Mac OS X uses. It uses a lightweight window server (which runs as it own process) with multiple rendering engines, including Display PDF, Quickdraw, and OpenGL. This is not like the graphics system on most other OS's. It relies on high speed IPC to implement a client/server graphics solution.
This system has some real tangible end user benefits. For example, even when an app is busy waiting, an end user can still move that app's windows and panels and the refresh of the window still happens (unlike Windows where you start getting lots and lots of white space). Since the lightweight window server maintains the backing store, you get great UI performance even with sluggish apps. This model may have performance benefits compared to the overhead of lots of threads... it's a different approach as compared to the BeOS approach... both have their strengths and weaknesses and both are superior to most other graphics systems available.
As for age, this system is an evolution of the Display Postscript window server model introduced in NeXTstep in 1988. It's not new, but it is significantly enhanced and competitors still haven't achieved the 2D graphics user experience given in the original NeXTstep 1.0
As soon as I do some more testing, I'll get a Mac OS X Server.pkg package up on next-ftp.peak.org. My Samba 2.0.4b and 2.0.5a packages are already up there, including instructions for compiling it yourself.
Another place to look for software for Mac OS X Server is Stepwise's Softrak service.
If you are in a hurry, the just copy the config.guess and the config.sub from/usr/libexec to replace the ones given in the Samba distribution. Configuring the smb.conf and other Samba issues are pretty much the same as other platforms.
Why pay for a BSD clone when you could just run BSD?
You could do some of your own research. See Apple's Mac OS X Server Product Page. Now, you might not want to run it, but you should be able to see why some other people might want to run it. Just in case, here are some keywords:
Apple File Sharing
Macintosh Manager
WebObjects
Cocoa OO Frameworks
Apple Hardware Support
Supported Quicktime Streaming
NetBoot Server for MacOS
ISV app support (especially in comparison to *BSD or Linux on PowerPC)
MacOS compatibility environment (w/o CPU emulation)
These are all above and beyond the fact that it has a Mach kernel with BSD 4.4isms and for the most part plays nicely with other UNIX variants.
The header files that come with OSX server are seriously f**ked.
F**ked in what way? Samba 2.0.6 compiles cleanly... the only issue is that MacOS X Server isn't recognized by the config.guess and the config.sub as distributed in Samba. That's easily remedied, especially by someone of your prowess by copying the ones that Apple supplies in/usr/libexec.
Slashdot Mirror
I got it to work with my PowerBook G4 (DVI), D-Link USB Bluetooth adapter (any USB Bluetooth adapter should work), Sony Ericsson T68i, and T-Mobile service.
It wasn't as straightforward as it should have been - however, it might be a settings/firmware problem with the phone. Apparently T-Mobile presets certain settings in the phone which you as an end user can't change. Phones sold by Amazon recently have a problem with WAP access. I also could not use the built in T-Mobile Internet Access Data Account (using CID=2).
I ended up calling T-Mobile's Data Support Service at 1-800-256-9991 and they walked me through using CID=5. I did have to modify Ross Barkman's Ericsson GPRS Script to call using CID=5 instead of 3. Here's the mod:
I was doing this with the T-Mobile tech on the other side, so I don't know if he had to fiddle with your T-Mobile account settings. Best to call them if you're having difficulties. Everything else was the same as you could easily find in the recipes that other people have posted.
I have also helped a client get it to work with AT&T's mMode service - the toughest part is usually getting the right APN, or knowing to use none.
It was refreshing to be able to contact a service (T-Mobile) and they didn't immediately give up when they learned that you are using a non-Windows platform. AT&T, however, was pretty much "we can't help you since you're not running a supported platform" and even that was narrowly defined.
Ah... wrong.
The PowerEdge 350 can only take 1gb worth of PC100 RAM (vs. 2gb of DDR on the Xserve), can take only 2 drives (vs. 4 in the Xserve), has no built in firewire, has only 10/100 NICs and one free 32 bit 33Mhz expansion slot (adding a gigabit NIC takes up that slot). The Xserve has dual gigabit NICs and still has a 64 bit 66Mhz slot free. Finally, the 1Ghz G4 with 2gb of L3 cache is much faster than a 1Ghz PIII on most tasks - and really blows it away with Altivec/Velocity Engine aware code. Depending on exactly what you choose (since the two can't be matched exactly), the PowerEdge 350 might be slightly cheaper running RedHat Linux, but would be more expensive running W2k.
The PowerEdge 1650 is closer, but again it mostly comes down to RedHat Linux vs. W2k on the price.
Yes, it would be nice to have a smoother transition to something bigger - switching to another platform with different tools can be a pain.
I respectfully disagree. Even the most advanced, most professional system administrators can benefit from the time savings from the various tools in Mac OS X Server. But it's not for everyone, and not for every case. It is a good value for many situations and is definitely better than the low end Sun and IBM stuff. It is very competitive against the x86 stuff even running Linux (especially against a commercial server version of Linux).
The new top of the line iBooks have a 700Mhz G3 with a 512k full speed on-core L2 cache. That gives a big speed boost over the older generations of iBooks. However, as compared to a PowerBook G4, it sports only a 100Mhz system bus (versus 133Mhz), a Mobile Radeon vs. a Mobile Radeon 7500, and the PowerBook as a 1mb L3 cache not to mention the G3 vs. G4 thing. The PowerBook G4 also has built in gigabit ethernet.
Therefore the PowerBook G4 can be substantially faster than the iBook if you are working with a substantial memory footprint, rely heavily on OpenGL, use applications that can leverage the Velocity Engine/Altivec unit or doing big network transfers over gigabit ethernet. Otherwise, the performance different isn't that big - especially comparing operations that are more disk I/O bound (of course). The iBook may even win a few tests with it's larger full speed L2 cache.
As for your Thinkpad, well, both the iBook and the PowerBook G4 would probably be pretty competitive from a raw perspective. However, the performance is going to probably be very different depending on what you specifically are doing - thus w/o additional information on exactly what you are going to do with the machine as well as what your personal perspective on what is important in terms of performance, it'll be difficult to judge.
The PowerBook G4 really has many additional features, and you have to be comfortable with the sacrifices that are made to make the iBook smaller and cheaper.
I'm very happy with my PowerBook G4 800 DVI and it is a substantial improvement over the PowerBook G4 500Mhz machine I used to use. I'm hoping the next update is more of the same - maybe 933Mhz+ and built in bluetooth. It would be nice if it had better 802.11b range too.
Now this is just insane. I've used A/UX on a Mac IIfx. There was almost _no_ integration between the Mac and the UNIX sides. You were basically running the equivalent of take-over-the-screen Mac OS Classic environment in a relatively hostile and non-standard UNIX host. There was no Cocoa. There was no Carbon. There were no Mac OS GUI applications that actually were UNIX processes. There wasn't any UNIX API exposure to the Mac side. So don't talk about "pretty-much on par" integration.
the G4 is doing much more than the IIfx
Ah, yeah. The two aren't doing the same thing. That's like complaining that Enlightenment doesn't run as fast as twm. Plus, you're also talking about some disk I/O performance related things. Let's see... the high performance SCSI disk (for the time) in your IIfx was how fast? The laptop ATA drive in your G4 PowerBook is how fast? Hmmm... not much difference is there?
In fact, i use the IIfx quite a lot as it provides a cheap and usable X-Terminal with more character and a much quieter fan than my assortment of x86 boxes.
You can build a cheap and nearly silent x86 box. Really.
Sure, A/UX was never a mainstream OS for Apple, but I wonder what might have been had Apple not abandoned the 'UNIX way' all those years ago.
Of all things, A/UX is and was not the beacon of the future.
However, when Carbon apps are busy, the user can't move any of it's windows. Very, very frustrating. Also, Carbon apps not written specifically for Mac OS X have definite disadvantages and are pretty easy to spot. Just look for the most ugly and non-conforming applications except for the pure Java ones.
Plus, it's usually the Carbon apps that don't properly leverage the operating system or contain lots of cruft like 31 character file name limits or have constantly running loops that hog CPU time and consume battery power, even when they're supposed to be totally idle. It's usually the Carbon developers that can't quite "get" daemons and proper modern application architecture. It's a good thing for them that there's a huge installed base out there that are almost as backwards thinking as they are to keep buying their products. A lot of the rest have moved onto Windows or something else.
I would give Gefen a serious look. Their DVI to ADC Conversion Box is the only one on the market that I know of that claims support for the 23" HD Cinema Display. Apple's, Dr. Bott's, and the Formac converter boxes all disclaim use with the 23" HD display. If nothing else, they should be able to help you with selecting the right graphics card.
One of the biggest problems with Mac OS X is the constant bowing to the classic Mac OS crowd. I think it could have been a much better platform if it had less shackles from an operating environment that has not truly shed it's 512x384, single tasking, single user, system modal panels, non-networked roots.
Ah... so you really haven't adjusted yet, huh? I remember with Mac OS X first shipped in March this year, the huge amount of venom from the aged Mac user crowd (never mind that many of the ex-NeXT users were Mac users before using the NeXT). But now, I'm seeing lots of stories of how people really like the Dock now, and are frustrated when they occasionally use Mac OS 9!
The Classic Apple menu is a prime example. How many novice users even know they can modify it? How many know how to modify it? What happens when you turn on Multiple Users and need to change an item globally? Why is there a separation between the active apps menu and the Apple menu - thus you have two menus entries for say, AppleWorks? I could go on.
Mac OS X, with it's preemptive multitasking and rock solid virtual memory system, as well as it's multi-user features and the like should force a re-think of how a user interacts with the system. The problem with Mac OS X is that it the new system isn't finished, and if anything, it should be less like Mac OS 9.
The original idea when Apple went with NeXT was that Apple would ship essentially OPENSTEP/Mach for PowerPC. The early Rhapsody Developer Previews were essentially that, and were available pretty quickly. Apple had to dust off the old NeXT PowerPC port and bring it up to speed and port it to Mac hardware, as it was originally written for the NeXT RISC Workstation that never shipped (I've seen a prototype of the m88k version, but I haven't the PPC version).
The problem with that strategy was that the major ISV's balked at the idea of porting to OPENSTEP API's. They saw it as a lot of time and expense for a platform that might not last out the year. It would not have required a total re-write as some people have suggested, but certainly it would have been a major effort. (I would personally argue that going the Carbon route was also painful and going the Cocoa route would have resulting in a better product). Plus, these ISV's would have to then maintain separate ports for Mac OS X and Mac OS X, and they weren't willing to do that - many of them had already cut out ports to anything but Windows and Mac, and were probably considering dropping the Mac anyways.
So the Rhaspody strategy was abandoned, Steve Jobs took over, Apple re-invested in the traditional Mac OS and got some good releases out the door. They also came up with Carbon, which is a re-tooled Mac Toolbox API that sits native inside of Mac OS X. In doing so, they also re-wrote the graphics layer, removing Display Postscript and replaced it with brand new code called Quartz which is based on PDF. That means re-writing the window manager as well so that it supports simultaneous display of Quicktime, OpenGL, Java2D, QuickDraw, and so on including using underlying hardware support. They also re-wrote the DriverKit layer, replacing it with IOKit which is embedded C++ based and has much broader support. The print system was replaced, the Workspace Manager was tossed and the Finder was re-written in Carbon (IMHO one of the worst parts of the current Mac OS X). Lots and lots of utilities were re-written, the BSD layer was upgraded from BSD 4.3, the kernel was moved from Mach 2.0++ (2.5 and some 3.0 extensions) to Mach 3.0++. The Classic layer was also added so that it can mingle with native apps, Java was added, Mail.app was re-written, and so on and so on. There was a lot of work put into this operating system since OPENSTEP 4.2 for Mach, which basically remained stagnant for years.
In the meantime, Rhapsody did essentially ship as Mac OS X Server 1.0 in 1999. It was basically OPENSTEP 5.5/5.6 with a menu layout that was Mac OS Classic-ish but pretty much everything else was straight from OPENSTEP/Mach.
So... any operating system that Apple might have chosen at the time would have had to go through the wringer in order to get it to support Apple's technologies and what Apple perceives as what their customers require. It would have taken a long time, and BeOS would have been a worst choice in terms of both adapting the technology and the personnel. I think that going with BeOS and C++ would have led back to the Copland and Taligent quicksand pit. As for personnel, if Steve Jobs didn't come back to lead Apple, I'm not sure Apple would have had this resurgence.
Apple's stock is in the tank.
<P>So are alot of other high tech companies. Plus, Wall Street isn't known for smarts.
<P><I>Preliminary indications are that Apple users are not particularly interested in the complexity and sluggishness of Apple's latest operating system.</I>
<P>Incorrect. Some are scared, just as many people are scared of the unknown. It's FUD, but they'll get through it.
<P><I>Processor speed is stuck at 500 MHz.</I>
<P>dual 533 is shipping this month, 667's and 733's in March.
<P><I>Alternative architectures and software are killing Apple on features, price, and performance.</I>
<P>Companies and platforms compete. Apple is competing better now, and will compete even better this year and next than it has years ago.
<P><I>There are legions of corporations and individuals who have been disrespected by Apple--from the BeOS community to the Apple clone industry, all of whom comprise a formidable enemies list.</I>
<P>Hmmm... might you be on that list? Did Apple hurt your little feelings? That would explain your use of errors.
<P><I>The main provider of Apple's microprocessor, Motorola, is hurting and hopes to leave the desktop processor business.</I>
<P>Motorola as a whole is hurting, yes.
<P><I>Today Motorola announced 2500 layoffs.</I>
<P>How do layoffs at a cell phone manufacturing plant have anything to do with Apple and the PowerPC?
<P><I>The most reasonable solution would be for Apple to open up. Open up its hardware specs and software so that where now exists little more than a corporate cult, there might exist a vibrant autonomous industry of developers, hackers, and alternative hardware vendors.</I>
<P>Apple is in transition. Darwin, the public source umbrella of Apple's software projects includes the Core OS of Mac OS X. Guess what. You can port Darwin to any PowerPC ABI architecture you want and you'll be able to run Mac OS X on it. Most of all, Apple has to protect itself first. After all, it is a business.
God knows I'd switch if I could afford the @(#%&@)*#$& Mac hardware.
Expensive??? In 1989, as a very poor student, I scraped together $2500 for a 386 system. In 1991, still a poor student, I scraped together $3200 for a NeXTstation (at edu discount) and another $1500 for a 400mb hard drive and 16mb memory upgrade. I saw this as an investment in my future and it has payed off.
In today's dollars, that much money would probably buy a Sun Ultra 60. Plus, we have today's ludicrous salaries for people with no degrees and no experience, why are there so many people whining about the costs of a Mac? $5000 is what one paid for a decent entry level UNIX workstation, and I gladly paid the extra over a x86 PC for the difference in quality and capability (both hardware and software).
What's really expensive is time. And Macs can often save time and effort, as does quality Sun servers and the like. An hour of time savings == $150 to me. I can't tell you the number of freaking hours I've spent with what is supposed to be high quality x86 hardware (as I gesture toward the database server box based on a 1.1ghz Athlon, 1.5gb PC133 ECC SDRAM, K7T Master motherboard w/ on-board Adaptec 789x U160 SCSI, etc. right next to me for a client - and I've used the name brands too). You would think that starting salaries were under $20k for all the whining about price I've seen on SlashDot...
It's true the Power4 isn't designed for graphic artists or desktop users. The cost is also not in line with that market. However, Apple _could_ use the discards. Power4 chips where one of the CPUs is non-functional. It will still be fast. It will still be expensive. However, it won't be nearly as expensive, and IBM would have tossed those chips anyways. This way IBM recoups more $ per wafer. As long as the non-functional CPU can be disabled and not use any power, I think a 1 way Power4 would be a great personal workstation CPU.
Another approach is to just port Darwin to RS/6000's.
First, it was my impression that IBM had spent a great deal of time on Mach 3.0 as part of OSF/1 as well as other projects like a replacement kernel for OS/2 and a new OS for the RS/6000. There must be a great deal of Mach experience at IBM that you could have drawn from (I say this, not knowing the political situation at IBM).
Since Apple's Darwin is based on OSF's MK7.3 base which includes most of what is considered Mach 4.0. Since Darwin is the core of Mac OS X, and mixed with what IBM has been doing with Mach over the years, you might have had a really nice combination. Linux is just another UNIX.
There are a few places where the the basic Mach kernel is weak for instance no support for drivers as modules, lack of driver support
Apple's IOKit framework helps solve that for you. It's a object framework for writing drivers, and yes, it does support drivers as modules and supports things like USB, IEEE-1394, among other things. It's based on embedded C++ and it's pretty slick.
messaging is pretty slow & used for everything, great for building clusters but overkill otherwise. As mach uses seperate address spaces for everything this improves protection but decreases performance.
Apple's take on this is to keep the semantic separations, but compile BSD into the same address space. Thus, instead of RPC calls into the kernel, they fold into function calls. You keep the abstraction but don't loose nearly as much in performance. Of course, you then couldn't crash the upper BSD/Linux layer and keep the kernel running then. :-)
There are multiple BSD projects, and certainly forking yet another isn't a big deal. However, I think a combination of Mac OS X/Darwin + RS/6000 or better would be a fantastic combination (drool).
I took notes from the Mac OS X Kernel (Darwin) presentation at Apple's World Wide Developer's Conference 2000.
On whose authority? Just because Job's said so? To be fair, the benchmarks jobs was running were crooked Photoshop benchmarks that were solely optimised for the Mac, if he run a standard benchmark then I'd be interested in the performance.
Of course, for many of the Mac power users, Photoshop is *the* application where they need as much performance as they can get. Plus, it really wasn't that crooked, since they ran a "real world" application with a "real world" use of that application. It wasn't a synthetic benchmark, but doing something graphic artists need to do on a daily basis. Now, extrapolating that result to general performance is the problem. But since most of Apple's pro market does graphics of that kind, this is relevant benchmark.
Also, during Job's presentation when he introduced the G4 SMP, he just said "we're going to compare it to a 1ghz P3", great but I'd like to know more details than just "a 1giz P3", which chipset? how much memory? the graphics card? the hard disk type etc? You get the picture; it was hardly a objective test.
It was an off the shelf machine... not a custom built machine. So you can get specs from the OEM. Plus they did state they had outfitted them as close as possible to identical specs.
Also, you have to remember you could buy 3 or 4 Athlons for the price of a G4.
Are you talking CPU's or machines? G4 chips are cheaper than Athlons (in quantity), but Apple does have one of the highest margins in the PC biz. However, 3-4 vs. 1? I think not.
The big difference is that you can go cheap on PC components... decide instead of buying a rock solid Socket A motherboard like the Asus A7V, go buy a Biostar or something like that for 30-40% less. Same goes for components like RAM, hard drives, etc. While Apple may not use the *best* components, they use some very good components (like the latest IBM 7200rpm ATA drives). We could build almost identical (by spec) 1ghz Athlon T-bird boxes, but yet be several hundred dollars or more apart in price just by changing the quality level of some of the components.
BTW, the cheapest dual 500Mhz G4 box you can get from Apple goes for $2449 through Build To Order. At that price, you still get two IEEE 1394 ports, gigabit ethernet, DVD-ROM, and of course, the ability to run Mac OS X (definitely worth the price premium).
ChristTrekker wrote:
Why not go ahead and partition the drive on your G3? If it is a New World architecture Mac (Blue and White G3 and newer), then the Darwin binary installer will even install into a partition for you (don't need to wipe a drive).
A new/old AMD box is going to cost more time/effort than partitioning your G3.
AppleWorks 6 is a Carbon app. If you have DP3, try running AppleWorks 6 and see what you get.
<P>Of course, the merits of AppleWorks 6 is debatable... I don't use it much, but it appears that there are a number of ClarisWorks and AppleWorks 5 users that are not too happy with it. Probably a point release will help quite a bit.
<P>Anyways, there is at least one "works" type suite available for Mac OS X on the ship date. I would bet that Apple is working on Carbonizing all their various non-Carbon applications (iMovie, Final Cut Pro). The AppleShare IP and Mac OS X Server functionality will probably also be ported, but most that of that will be to the BSD API's. The admin apps will probably be a mix of Carbon/Cocoa and web based stuff.
<P>Various heavy weight 3rd parties in the Mac space have already publically committed to some Carbon ports at least. A whole parade of them were on stage at the last MacWorld Expo in January - Adobe, Macromedia, Microsoft, id Software, Quark, etc. Adobe actually showed a preliminary Carbon port of one of their apps (Photoshop or Illustrator) on stage at function last year.
<P>By the end of this year, many people won't need to fire up the Classic environment except for games (whose pressure to port to Carbon seems to be less).
Reading the A/UX FAQ, A/UX isn't based on a Mach kernel (which is good or bad depending on your personal viewpoint on microkernels). Also, A/UX wasn't intended to replace Mac OS. Mac OS X (10) is a replacement for Mac OS and therefore the integration with the Mac OS world is far more complete. A/UX was pretty much SVR2 UNIX heavily extended with a Mac OS shell. It doesn't have Cocoa or Carbon apps - Photoshop, Freehand, etc. aren't full blown UNIX tasks in A/UX, but the Carbon versions of these will enjoy the virtual memory and pre-emptive multitasking features right on the Mach/UNIX layers in Mac OS X. This will be a UNIX derivative that is intended for your computer newbie to use.
Apple may have the highest shipping volume desktop UNIX-like system in rather short order. Apple has stated that they intend to completely convert over to Mac OS X by the beginning of next year... all shipping Macs will come with Mac OS X then.
Actually, the current Apple Power Macintosh G4 Mac OS X Server machines have dual power supplies... and a powerd daemon to go with them. It's not hot swappable, however. It is instant switch-over.
Also, there are many servers out there that don't need more than 3 64bit slots. Matter of fact, many rack mount systems sacrifice slots - leaving only one or two available. With built in 10/100 ethernet, built in firewire (for that brand new firewire SAN to be supported in MacOS X), 2 USB ports, what exactly do you need more than 3 64 bit PCI slots for anyways? Let's see... gigabit ethernet card, dual channel 64 bit PCI Ultra/160 SCSI card, what else? I know that some people do need them, but not usually as servers and certainly not in this price category (under $10k). Plus, PCI expansion chassis's also exist.
In general, Macs do make nice solid quality machines compared to most x86 machines - not as solid as real Sun boxes (not the Ultra 5's), but then the price is a whole lot cheaper. It takes a lot of effort (which can mean dollars which is often discounted) to identify, specify, and purchase quality x86 systems. It is possible, but with vendors changing components all the time, it can be difficult.
DMDx86 wrote:
Yep. This is exactly how Apple pulled off the "No G4 Until we say so" stunt. I was quite close to buying a Mac for Linux purposes, until I heard about this. If I can't upgrade my machine as I please, I do not want it. Now I am buying an Alpha instead..
The "no G4 until we say so" was implemented in firmware - there isn't a Mac OS ROM on a chip anymore - the MacOS ROM is loaded from disk on all recent machines. The G4 firmware block is suspect... but on the balance, not as bad as it initially sounds. The Blue and White G3 motherboard and firmware wasn't designed to work with a G4. Apple doesn't produce a G4 upgrade for those machines, doesn't plan to, nor touts that machine as upgradable. 3rd party manufacturers have already worked around the issue and so you can get a G4 upgrade now, but now it is obvious to all involved that Apple has no responsibility for the results. After all, if there is a data corruption or stability issue, Apple doesn't want to be on the hook for it.
Personally, I wouldn't upgrade to a G4 in a Blue and White G3. If you want to upgrade, buy the latest G3 chip or sell the machine and buy a G4. The new UMA chipset G4's have much higher memory bandwidth - it doesn't make sense to drop it into a Blue and White machine which still uses a 60x protocol. Plus the G4 upgrade pricing is too expensive (compared to other PC's, not Alpha's or UltraSPARCs).
IMHO, this issue is a poor one to use when deciding between a Mac or an Alpha for Linux. There are tons of other technical and price issues to work with first which should decide the point. How much are you spending? What kind of support are you expecting? What applications to you expect to use? Do you need to run commercial applications or do you have source for everything you want to run? What kind of upgrades? What level and kind of performance are you looking for? What peripherals do you need to get? The Mac and the Alpha don't really compete against each other since the answers to the above questions can be very different.
A/UX and Mac OS X are on opposite ends of the UNIX spectrum. Yes, both look and act like UNIX at a certain level. After that, things are different.
A/UX doesn't have:
Cocoa/WebObjects
Quartz/Aqua/OpenGL
Carbon
transparent (rootless) Classic Mac OS support
Darwin (open source underpinnings)
PowerPC support
Mach
transparency to UNIX (Mac OS X won't require UNIX knowledge at all)
and all the various components built upon the above.
In A/UX, there was no bridge between the Mac OS environment and the UNIX layer. There was no future migration strategy or technology to get off of classic MacOS. It was basically a UNIX for 68k Macs that happen to run Mac OS in a virtual machine in take-over screen mode. Kinda like what the Mac On Linux project would get when it done on LinuxPPC. Interesting for its time, yes. A future for Mac OS, no.
RFC959 wrote: I don't know what graphics system OSX uses, but I suspect it's either X-like or MacOS-like. We're talking about multi-user machines here, remember? X is what you wanted to get rid of, and MacOS is built with the assumption that you've got one machine, one framebuffer, and one user (and one GUI!) in your "computing environment", none of which is necessarily true anymore, making it an unsuitable starting point.
You're right. You don't know what graphics system Mac OS X uses. It uses a lightweight window server (which runs as it own process) with multiple rendering engines, including Display PDF, Quickdraw, and OpenGL. This is not like the graphics system on most other OS's. It relies on high speed IPC to implement a client/server graphics solution.
See Mac OS X Graphics for a newbie and shallow overview, and a technical overview from Stepwise's WWDC '99 Graphics Coverage
This system has some real tangible end user benefits. For example, even when an app is busy waiting, an end user can still move that app's windows and panels and the refresh of the window still happens (unlike Windows where you start getting lots and lots of white space). Since the lightweight window server maintains the backing store, you get great UI performance even with sluggish apps. This model may have performance benefits compared to the overhead of lots of threads... it's a different approach as compared to the BeOS approach... both have their strengths and weaknesses and both are superior to most other graphics systems available.
As for age, this system is an evolution of the Display Postscript window server model introduced in NeXTstep in 1988. It's not new, but it is significantly enhanced and competitors still haven't achieved the 2D graphics user experience given in the original NeXTstep 1.0
As soon as I do some more testing, I'll get a Mac OS X Server .pkg package up on next-ftp.peak.org. My Samba 2.0.4b and 2.0.5a packages are already up there, including instructions for compiling it yourself.
Another place to look for software for Mac OS X Server is Stepwise's Softrak service.
If you are in a hurry, the just copy the config.guess and the config.sub from /usr/libexec to replace the ones given in the Samba distribution. Configuring the smb.conf and other Samba issues are pretty much the same as other platforms.
Why pay for a BSD clone when you could just run BSD?
You could do some of your own research. See Apple's Mac OS X Server Product Page. Now, you might not want to run it, but you should be able to see why some other people might want to run it. Just in case, here are some keywords:
These are all above and beyond the fact that it has a Mach kernel with BSD 4.4isms and for the most part plays nicely with other UNIX variants.
The header files that come with OSX server are seriously f**ked.
F**ked in what way? Samba 2.0.6 compiles cleanly... the only issue is that MacOS X Server isn't recognized by the config.guess and the config.sub as distributed in Samba. That's easily remedied, especially by someone of your prowess by copying the ones that Apple supplies in /usr/libexec.