I don't give an ox's behind about whether Apple just stole something from Linux, or if WMP9 has the rating system. All I know is that Apple did a perfect job (unlike WMP9, which IMO and IYO is CRAP) and is easy (unlike most Linuces. Agreed, Mandrake is not too hard. But its not EASY, like OS X).
I don't have any bubbles to burst. Like I mentioned before I'm no Apple salesman. If there is a positive feature about macs, I accept it. If there is a negative feature, I also accept it.
Im just saying, Steve Jobs acts like he invents so many things that he didn't.
Now, he did come up with some really cool innovations, but not every feature of OSX is innovative.
Its just that you talk about feautres that are reasons to switch, and im saying that they are available on other platforms. Though yes, linux is MUCH harder.
I guess I am bias. I usually go by the idea that its better to learn how to do something thouroughly, and to fully understand what it is your doing. I dont want to know how to turn a computer on, I want to know how the actual parts work. Im like this with most things, If I spend the time to learn it, I want to learn it right, or its not really worth learning at all.
Most people aren't such :nerd: though, and I should understand that.
My mom is 77 years old. She has no problem using my PC. She can compose letters in Japanese and even print it vertically! She can also use Excel without any help. She learned this very quickly.
Macs are ok, but i'd rather spend it on something I can easily upgrade/modify/tinker with - that is what is fun for me - also figuring out how things work.
Originally posted by Taha@Jun 24 2003, 04:42 PMI really don't want to sound like a Mac salesman, but here we go:
Internet IS easier on the mac. By internet I mean setting it up, configuring it etc.. coz you don't have to do a THING.
I disagree - getting my dad's Powerbook connected thru my Smoothwall (Linux based router / firewall incase u didn't know 
) was a complete pain.
Connecting my win2k PC and my bro's XP laptop was a breeze - just entered a manual IP, entered the smoothwall IP in the gateway, wait for windows to assign it and voila - easy trouble free broadband access.
The Mac took lots of fiddling to make it actually load any web pages.
OT, but my parents went to look at the college my sister wants to go to today, and when they went in the Graphic Design lab, my dad asked the teacher why they were teaching graphic design using PCs when everyone in the industry uses Macs.... he didn't get an answer <_<
So what DID you do to set up the internet in the Powerbook??
I really don't know why you are all arguing about this! :rant: OK Windows, MacOS, UNIX, BSD, Linux, DOS... All truly great operating systems, both have major innovations...
Windows - Fairly stable, fast, a worthy competitor w/ MacOS
MacOS - Wonderful graphics, powerful, easy to use... the list goes on and on
UNIX/BSB - Fast, very Powerful, old est 1969 i think
Linux - Powerful, stable...
DOS - Very fast!!!! Can do almost anything~
"It's all good"! For god sakes, W/ out Microsoft & IBM we would'nt have PCs, w/out Apple we might not have GUIs like we do, plus this board wouldnt exist! :rant:
Ummm.... none of us are about settling. Unless you'd still rather be using DOS?
Another great unbiased post by Viper-X:
There's a lot of controversy going around about the G5 speed tests and their validity. I would like to state first off that I really don't care if it's the fastest chip on the market. I'm not a mac fanatic, and I own both a PC and a Mac. I'm not a well expierenced programmer, but I believe I know my way around code and compilers for C based languages.
After examining the PDF from Veratest, complaints on Slashdot, and complaints from Haxial I felt that it was time to release my clarification on the results and logical explination.
A: Choice of Compilers
First of all, It's important to recognize that code is created to test a processor, but alone this code does nothing as it has to be compiled. What this process does is convert a complicated readable language like C into machine language. There are more and less efficient ways to communicate with a processor tho, giving you more and less efficient compilers. The primary complaint of the people who wish to disregard these tests is that the compiler, GCC, is not the most efficient compiler for the P4, and thus produced slower machine code for the P4, which skewed the results.
This is not actually the case, as GCC is a more mature compiler on the X86 platform and is most likely more optimized for the P4 than it is for the 970. This is hard to verify unless you are a programmer of GCC on both platforms. Furthermore, it's impossible for a human to go through each bit of machine code and make sure that the optimization produced by the compiler was similar enough for it not to matter. The platforms are completely different processors, and one can only hope that GCC is producing platform similar machine code for each platform. In this case, you are putting your trust in an industry standard compiler for similar operating systems.
Using a special compiler from Intel to overinflate the results of the P4, making the test a purely theoretical one, would not create a ballenced playing field, as IBM would be forced to release a processor specific compiler for Darwin which would automatically generate highly optomized code for their chip. Even then, the compiler wars would not end, as it is unknown what level of optimization is possible between the team of programmers at Intel and the one at IBM. All compiler debate ends with the simple fact that you cannot get down to the machine level of the code to verify that the tests are similarly optimized, and must trust the fact that these are the most widely used compilers for both platforms, and that your applications that you use will use these compilers to produce the applications that you run.
B: Optimization Flags
Veritest has been criticized for using special optimization flags for the G5, disableing hyperthreading, and not using SSE2 on the P4. These decisions are said to be specifically designed to heighten the performance of the G5 and cripple the P4.
1: Optimization Flags
When comparing a 32 bit processor to a 64 bit processor, some things are going to be a little hard. General math for example. 32 and 64 bit math use the same pricnipals of binary addition, however the implementation is far different. Veritest has been discredited for using the -fast and -fastf flags in their compiling of the tests because this gives the G5 specific compiler advantages over the P4.
To be fair, both systems were given 5 optimization flags. What this means is that the compiler is capeable of optimizing the code a little better, depending on the programmers choice. There are many options to choose from, however the limit on this test was set to five.
The three most controversial decisions in optimization flags were the use of a special malloc library for the G5, activation of special, processor specific optimizations for the G5, and the disableing of SSE2 in the Pentium 4 models. Before I go into why I believe these decisions were made, let me say that the controversy only exists because Veritest gave no indication as to why they set anything the way they did, therefore anything you read or hear about why is purely speculation.
a. special malloc library
It is unknown to me as to why they chose to use a special malloc library, however here are my guesses. malloc is a method of creating space in RAM for data; memory allocation. If you ask me why this has anything to do with measuring the ability for your computer to count, I couldn't give you a straight answer, because I didn't program the benchmarks. The description by Veritest of this flag is:
"An implementation geered for speed rather than memory efficiency and is single-threaded which makes it unsuitable for many uses. Special provisions were made for very small allocations (less than 4 bytes)"
What this means in laymans terms is this: consider memory like a parking lot with a lot of different sized cars. When a car pulls in that is huge it takes up a lot of space. When that car leaves, 3 or 4 cars of a smaller size can fit in that space and be closer to the entrance than if they passed that empty space and took the next free one. What this also produces is a smaller "stack" of cars. Instead of swiss cheese parking lot, everything is tightly packed together, with no big gaps. This malloc library basically ignores the big space left by the cars that leave, and just grab the next free one down the line because it's faster not to LOOK for a parking space.
Now, if anyone can, please tell me how memory management relates to FPU and INT scores, because I cannot see a reason as to why it would other than storing results data. If the code is well written, most every variable which contains test data should be static, not dynamically malloced as the test goes. In fact, if it IS dynamically allocated, this has got to be the worst processor intensive test created. I do not believe a high speed malloc engine would skew the results.
Futhermore, with the high speed bus on the G5, it is quite possible that the only way to USE the 1Ghz FSB was to use a stripped down version of malloc, for whatever data allocation they DID need to do. Chances are, there isn't an optimized version of malloc specifically for the bus of the G5, at least, not as optimized as the one for the intel bus archatecture. This is mearly speculation however.
b. Activation of special processor optimizations for G5
The flags -fast and -fastf used to compile the tests for the G5 are documented as follows:
"This flag is used with C and C++ and specifically targeted to the G5 and enables G5 specific instruction usage, tuning and 64 bit arithmatic. In addition to enabling the -O3 optimization level, it also enables the use of C99 aliasing rules and relaxed IEEE math operations"
If you replace C and C++ with FORTRAN, that's the description of -fastf as well.
What this describes, to someone looking for a scapegoat, is processor specific optimization for the G5, built into GCC by Apple or IBM to create a program more efficient than a stock version of GCC. However, it is unknown what "G5 specific instruction usage" it enabled. What I am reading here is quite simply that the G5 is a 64 bit processor, and to enable the optimization flags -O3, a very common flag for optimization which is also used on the P4 list, you had to switch the compiler into 64 bit arithmatic. Basically, it told the compiler to send special assembly commands which would enable the G5 to use it's special 64 bit arithmatic, which is a stock piece of the G5 chip. It simply has to be enabled. I would like to know what the difference is in -fast and -O3 in the machine code, but it is impossible to really know at this point. I believe -fast is mearely a shortcut to -O3 and -C99 optimization flags for a new, beta hardware, 64-bit processor. I see no hard evidence that -O3 on the P4 does more or less hardware specific optimization for the P4, a widely popular chip that GCC is certain to optimize it's compiler for.
c. Enableing SSE, but not SSE2
This controversial decision threw me for a loop at first, until I realized the thing that most everyone has not mentioned up until now. I will get to that later.
SSE and SSE2 are special coprocessors built into the chip which give P4 faster integer performance. The difference is that SSE2 is also a double precision co-processor and gives a huge bonus to floating point calculations. This is evident in the SPEC scores that everyone refers to online when trying to debunk Apple's work. The optimization flag used was
"-mfpmath=sse
Generate floating point arithmetics for sse. Use scalar floating point instructions present in the SSE instruction set. This instruction set is supported by Pentium3 and newer chips, in the AMD line by Athlon-4, Athlon-XP and Athlon-MP chips. The earlier version of SSE instruciton set supports only single precision artithmetics, thus the double and extended precision arithmetics is still done using 387. Later version, present only in Pentium 4 and future AMD x86-64 chips supports double precision arithmetics too."
A single precision number is a float in C code and a double precision number is a double. Here are the IEEE standards on this designation. So, the question remains: why did Veritest choose to disable the SSE2 double precision coprocessor in the Pentium4 when they were comparing ONLY Pentium4s with the G5.
The answer: Atli-vec. However there are two sides to this tale.
I have seen no evidence that the code used on the G5 was optimized for Alti-vec in any way. What this means is that if the Pentium4 was allowed to use their special co-processor for not only single, but double-precision calculations, it would be unfair to the G5, since the test it's self was not designed to take advantage of the 970's coprocessor, Alti-vec.
However, SSE2 is capeable of calculating double precision floating point variables. This can also be seen as Apple trying to skew the facts simply because their co-processor isn't capeable of using SIMD (a very fast way of calculating floating points) at 64-bits. The G5 already has two dedicated double precision floating point processors in it's line, why does it need the advantage of taking out the double precision floats on the P4? Did the single precision numbers on the G5 use Altivec? These are the questions that have to be answered to know what is being skewed and what isn't. I'm going to do more research of SPECfp and see if I can find out if it is altivec enhanced. IF IT IS: then Apple will say they were leveling the playing field by disabling an unfair co-processor advantage over the G5, and Intel will say that disableing SSE2 is their way of skewing the facts.
C: System Setup
There has also been a stir over Hyperthreading in the Pentium4 Dual Xeon configuration. Veritest chose to enable hyperthreading while the machine was in single processor mode, but disable it when the both of processors were both running.
Hyperthreading is a "software" emulation of dual processors that runs on the fly inside the motherboard and processor of a Pentium4 3.0 and 3.06Ghz. This emulation allows the processor to more effectivly use threading to maximize the potential of a single processor by by using the concepts of a dual processor system. Basically when a process or thread is busy accessing the hard drive, querying the OS or other such things, the processor is idle and could be used more effectivly by another thread. Hyperthreading is a compilcated thing that I have not studied front to back, however I feel I have enough information to make the following statments.
Hyperthreading was invented for the Pentium4 for the sole purpose of making a single processor emulate dual processors in order to more effectively use threads. This concept may work well for a single processor trying to emulate a dual processor, but it is not the same for a dual processor to emulate a quad. Standard SMP processes built into the motherboards and compilers should be able to more effecitvely use dual processor processing time than hyperthreading enabled processors in a dual configuration. In other words, the overhead involved in hyperthreading would compound the overhead in SMP, and performance would be lost, IMHO. If you are already running a dual processor system, it makes no sense to me that Hyperthreading would add any boost of speed. If there is clear and strong evidence to the contrary, I have not seen it.
The best debate I could find on this topic is here.
The fact that it is debateable mearely means that speed testers won't use it. If it's unclear wether this option will skew the results, don't use it. I believe they did the right thing in having Hyperthreading enabled in a single processor, which can give to to a 10% boost in performance, and off in Dual Processors, which having a second processor generally gives you an 85 - 90% boost (do they really think dual hyperthreading is going to make up the 15% lost in doing SMP anyway? I doubt it... seriously.)
So: the question remains. Is the G5 faster than the P4 Dual Xeon 3.06. My answer, it will be faster far more often than it is slower and where it is faster it will be FAR faster. I sincerely hope this helps anyone floundering around in all this debate.
Viper-X-
vi is for hippies, old hippies 
whats up
with the release of apple's new G5 processor all the "gotta get me a mac" wannabe's are happy and i do got to admit that panther looks really nice and that iSight has a really innovative stand but from there i really didn't think there was anything great.
let me play devils advocate for a sec :evil:
for example i saw the specs in which a dual g5 blows away a dell 3.0 ghz by 32 percent "wow" i'm sure that's great but considering u can get a dell for 3000 and a dual G5 cost's i dunno more than 4000, if u consider it this way 32 percent isn't all that great.
(allright i'm on a roll)
next everybody trashes microsoft for being a monopoly but isn't apple a little monopoly in its onright, i mean at least for PC's u can mix and match parts and put together a nice computer but that would be pretty hard to do on a mac.
sure mac looks great and is really stable but XP is not bad as far as stability is concerned, more fortune 500 companies run microsoft software than anyother software company and we have aquasoft and aquaxp to get a real nice look on XP.
ok thats all (i hope after this apple bashing i don't get :goku's job: )
Apple is, and I bet always will be, one of the most innovative companies. They use to put behind its products amazing features and state-of-the-art style and look and feel.
In the other hand, seems that a well known history with the G4's benchmarks is raising up again with the new G5:
http://www.extremetech.com/article2/0,3973...,1136018,00.asp
(I've found a couple links more but I lost them, sorry)
Iceman.
I'd love the keynote especially the comparison Panther vs Longhorn...
for example i saw the specs in which a dual g5 blows away a dell 3.0 ghz by 32 percent "wow" i'm sure that's great but considering u can get a dell for 3000 and a dual G5 cost's i dunno more than 4000, if u consider it this way 32 percent isn't all that great.
They made a comparison on that same keynote with a 'similar' highend Dell and the Dell is $1000 more expensive... yes you can get cheaper Dell's but also cheaper Apples?!!
next everybody trashes microsoft for being a monopoly but isn't apple a little monopoly in its onright, i mean at least for PC's u can mix and match parts and put together a nice computer but that would be pretty hard to do on a mac.
I agree that a monopoly is sick, but what is wrong with inhouse monopoly, Nokia has it, Mercedes has it, and Apple has it, why cause it works. They create each and every component hardware and software themselve with one thing in mind, stable and fast.
About the mix and matching it think it is for amateurs on a budget (ok ok not everyone), i mean you don't take a SGI Workstation apart to mix and match things, same counts for Apple in my opinion. Those machines were meant to last. And dont rely on a upgrade path driven by Intel's need for marketing speed.
You buy one and configure it just right.. ( man just open the box.. how clean...)you can upgrade the videocard, the ram and even the disks in the future..
If Microsoft did exactly that same thing people wouldn't be switching, but since M$ is a software company without any hardware but a good position, it relies on thirdparties and that's where it gets ugly, very ugly.
My two cents.
The great thing about Apple maufacturing their own hardware AND software is that since they know every detail about the hardware that their software will be used on, they can tailor it to take advantage of every single aspect of their CPU's. That is why each version of OS X gets faster. Case in point, several old 233 mhz iMacs (i.e., the first iMac line) can run Jaguar quite well, as long as the iMac has adequate RAM. Do you think Windows XP runs well on a 233 mhz P1 or P2? No, it doesn't. And these same iMacs will also be able to run Panther. Why? Because Apple is making their OS take advantage of their hardware and tailoring the OS around it. Windows can't do that for the exact reason you stated- people mix and match. Windows runs on SO many different system configurations that it is impossible to make the OS fit each machine's specific needs. The result? Windows hogging memory and processor cycles....just let XP run for a couple days without running a RAM enhancement program and you will find that Windows slowly munches away at your RAM until there is less and less open for other apps, slowing down the entire system.
Now I'm not bashing Windows, it's just the way it goes...if Microsoft produced it's own hardware, you can bet they'd be building Windows around that hardware.
-godpunk
"Now I'm not bashing Windows, it's just the way it goes...if Microsoft produced it's own hardware, you can bet they'd be building Windows around that hardware."
I KNOW we will see the day when MS will build (Like Dell) its own computers. Slowly there getting into the hardware side of things. Ive used there wireless networking devices. There pretty easy to setup. Just like when you buy Apple stuff. ie: Airport
Originally posted by MetalMusicAddict@Jun 24 2003, 06:34 PMI KNOW we will see the day when MS will build (Like Dell) its own computers. Slowly there getting into the hardware side of things.
did someone just say D-Day?
D-day? not till the real design folks are stepping over to the 'darkside'
Maybe those x86 folks should start hiring people from Apple just like Intel tried with that Apple Marketing dude dont see jack from him... still same old blabla..
I always kept wondering why doesnt someone for instance Dell stand up and builds x86 pc's around chosen hardware with a very good and userfriendly design, And make it a direct Apple contender. Look at the laptops for godsake nobody is even close, look at everything.. Apple has unleashed... no x86 clone is close...
I remember SGI doing it a few years ago with the 540 series it had Xeons inside but a SGI graphic system that was a first good attempt...
The oem's don't consider it a profitable venture to branch off a line of cool-sexy designed pc's,...why?:
Problem 1: Most use similar providers: (Dell & Hewlett Packard laptops: Made by Compal, etc)
Problem 2. The oem's dont consider it a profitable venture because companies that do design cool-sexy pc's (at considerable design & production costs, e.g. Msi Cube PC) do not make as much profit.
I don't know, I'm no expert, I'm sure somebody will come and argue about what I just said.
Lol, I was talking to my friend the other day about how much thinner my iBook was than her Dell...I told her that they would cut down a lot on thickness if Dell (and almost all other PC makers for that matter) would just LET GO OF THE FLOPPY DRIVE! This isn't 1992, after all. No one uses them anymore, CD's are just as cheap these days and hold ridiculously more info than floppies do. Still, PC's insist on having them, making the laptop substantially heavier and thicker.
Now that I think of it, I can't think of an instance where I've used a floppy since I started college (2 years ago!) 
-godpunk
or those USB memory sticks... you could use those..
@Threedrives
Very true, they make more money with those crappy mainstream non-design...products but if Dell was known for their good ui, sexy design, and even more important real innovation on things (even only on a highend line like Sony is offering in a few months... Qualia anyone...)
I think it will sell, make a product cool, market it right, do a Dell-icous store like a Apple store etc etc. the Apple concept works.. why not for PC....
Originally posted by Eric Rabesandratana@Jun 24 2003, 09:41 AMJust to add a bit of perspective to the situation, take a look here.
As always, somewhere between this article and Apple's line is the truth, but if the raw data can provide such shockingly different conclusions to the one Apple drew, it's not looking good.
Yeah, I saw that link before too. I'm not surprised that Apple fudged their numbers. Even ATI and Nvidia pull those stunts. I love Apple and everything and I don't take every word they say as gospel but, the G5 is alot better than what they had going with the G4. I'll give Apple props for that. Could you blame Apple fans from being so exuberant these last 2 days? Give it a month or two after the G5s are out in the general public. Then we can start seeing some real world testing in those video editing and photoshop magazines.
I wonder how much higher Apple's profit margin is going to be with these new G5's? I imagine these cases don't cost as much to make as the older ones and I read somewhere that IBM gave Apple a huge discount on the G5s. But leave it to Apple to not pass on the savings to the users and try to sell more units. I guess they're happy being "niche" players. (uh-oh, I hope I don't get flamed for calling Apple "niche" players)
Originally posted by GameBoyzGB@Jun 24 2003, 04:37 AMI'm amazed at the design. Look Ma... No Wires!!!
Left = PC ..................................... Right=G5
I like that minimalist look. I'm not sure if I have this right but, are they only equipped to handle 1 CD drive? I'd hate to be limited to one drive. I prefer having 2 just in case I need to do CD to CD copying.
Originally posted by Aries326+Jun 24 2003, 04:37 PM-->QUOTE(Aries326 @ Jun 24 2003, 04:37 PM)
speaking of floppies...
Germans like to keep things close to them selves... i guess
Aqua-Soft Forums