Wisest advice I've ever heard--it was in my machine org and assembly textbook.
*Any* cross-platform benchmark should be taken with a shaker full of salt--they simply do not represent real world performance.
SPEC, for all of its nice points, also falls into this same category. In the end, when all is said and done, people prefer to confuse the model with reality--they think that real world performance follows SPEC scores.
Part of the problem is that Rowling has to balance both the game in her world for the audiences in her world and for/her/ audience (i.e., us).
She indicated that games of Quidditch often take hours and can take days (if memory serves she said one Quidditch World Cup game lasted 4 *days*). This would give a little more opportunity for either or both sides to get a 150 point lead (perhaps it is still not enough, but as you said: Rowling is a storyteller, not a game theorist). The length of time before the snitch is caught can help balance out the game.
Then she has to balance out what we are willing to read through. She wants to describe the action in Quidditch and give us a fast-paced feel for it, but she can't really do that if the game lasts too long. She tries to describe Harry's thoughts and follow the character through the game, but how long would we be willing to put up with that if the game lasted 12 hours?
So its somewhat likely that there are multiple issues involved in trying to balance the entire mess.
'compare The Two Towers where there are three simultaneous stories- and they're relentlessly literal where they could be surreal. "
Actually, this can be said to be part of litterary style--neither method is necessarily better than another anymore than you can make the unilateral statement "acrylic is better than oil in painting" or "oil is better than watercolor".
I took a class about a year ago in Writing Fiction with the teacher Joanne Greenberg (The Running of the Dear, I never promised you a rose garden, and many others). She explained that writing a good story with multiple points of view is more difficult than a singular point of view simply because every time you switched your point of view you had a tremendous potential for losing power with your audience. You can also amplify it, but it is signficiantly more difficult to do so.
Which one you choose is a matter of preference--it doesn't make your writing automatically "better" if you have 16 points of view which you follow throughout.
While I understand you are aiming for humor, I'm getting mildly sick of seeing this kind of comparison: You can turn almost any work into almost any other work by abstracting it sufficiently.
For instance, lets take the early Star Wars universe (neglecting, for a moment, the movie Hidden Fortress) and Atlas Shrugged, by Ayn Rand.
Both deal with a small group of individuals (bussinessmen or the rebel alliance) who split off from society and form their own communities which are being hunted by the predominant governments in their areas and fighting back against them. The dark side in each story has developed a superweapon (Xylophone/Death Star) which can obliterate large areas (cities/planets) and have a host of other similarities (both employ torture on the heroes!)
Let's also take a look at your x-wing/broomstick comparison. In the radio drama, where Luke's skill is really demonstrated, it is with a/landspeeder/ at first (x-wings come later) and he is skilled at racing his friends.
Now lets compare with Harry, who is more like a very talented soccer player. It makes a lot of sense, considering Harry's world, that they would have at least one sport played on brooms (and brooms are an old throw-in here from other legends &c). Thus, if Harry is going to be a natural at that sport, then he is going to have to be a natural on the broom.
It should also be noted that Luke is portrayed as being more than slightly reckless with his landspeeder in order to get an edge in racing. Harry is *not*.
This is just a start, there are a *lot* of other problems with the Harry--Star Wars match-up. Saying that they have some base similarities is true, but that can be said of many things that/in no way used each other/ as sources.
"It's pretty sad. Rowling always said she would NEVER sell the rights to Harry Potter. Now if you look at he books, there's a small message on the copyright page saying that Time Warner own the rights to all the character names and likenesses."
She didn't sell the rights to the books (IIRC), only to the "names and likenesses," which was probably a mandatory step moving it into the movie industry, having action figures, &c. She probably didn't want to deal with all of the paperwork of subcontracting each individual entity for these things (a company to make toys, &c) and it may have been a mandatory, and not terribly offensive part of the deal for her.
"they're loosing money here all this time the iTunes store isn't running in Europe. "
Hardly. They aren't making as much as they could, but they are not "losing money" because they are not in Europe.
"I think they'll have a lot to explain to shareholders, eg. 'Why didn't you pursue this business model as soon as it showed it's potential?'."
Not Apple's problem--there are laws and licensing issues involved that they are attempting to negotiate their way through. These are a more major issue than was first anticipated. Period. End of discussion.
If they are turning a profit with it now and/or will turn a profit once the Windows version comes out, why on earth would they need to justify that they haven't managed to work out licensing with Europe just yet?
They aren't exactly trying to explain to the teacher why they haven't gotten their homework in on time.
Windows access is relatively critical, but only from the perspective of competing with whatever MS manages to cobbel together that actually has a chance of competing (I haven't seen anything yet) and will likely have very little to do with the long term life-and-death of the project as it stands now (the record companies seem happy with the deal, and so long as they have the big-5 and the big-5 remain profitable, Apple will probably not have to float the store).
International access, on the other hand, is in no-way critical to whether the project succeeds or whether Apple succeeds. The problems are the laws in other countries which are not conducive to this kind of pricing scheme, as well as the international contracts and licensing issues surrounding distribution of media.
If Apple and the big-5 are having trouble with it, so will anyone else who wants to expand across country boundaries.
Thus, the two *most* critical issues for the iTunesMusicStore, at this stage, are:
1) Getting more labels aboard (they seem to be doing an admirable job at getting this set up).
2) Getting iTunes ported to Windows (once again).
These would be the "logical next steps" to remain in front of the pack. International access is more of a nicety that is not overly critical for their success (though I'm sure they want to do it as quickly as possible--more profits are never a bad thing).
"imagine remotely accessing a machine in Finland, using that machine to download from Apple's store at a cheaper rate than your home country, and then downloading from the Finland machine to yours. "
This would probably be countered the same way that it is being handled now to prevent people from buying music this way from Finland by routing through a US American server--by checking the credit card address.
Thus, if your credit card has a home address in Estonia (and a Credit Card *is* required to use the store), then you won't be able to use the iTunes store through Finland after Finland is included in the plan or through the United States today.
It doesn't matter if it is a better product--someone will ask their friends "what will work for me?" their friends say "I use this, it works for me" and that prompts said person to go out and buy X.
Most people I talk to I can sway to buying a Mac--if I get to them first and let them get their hands on one.
"Nope. If it's not threaded properly, it will not take advantage of multiple processors. The application will run on one processor."
Yes, except for one thing: you generally have more than one application running, and each of them can have more time because you have two processors. Thus, one process (such as one application) can receive a huge speed improvement simply because it doesn't get switched off of the processor as often and doesn't have to make as much room for other things to run.
If you don't believe me do the simulation yourself. Draw out 10 processes on paper, don't forget the OS kicking in to decide who goes next, and assume a Shortest Job First, Preemptive (you can ignore all of the other fancy things, this is just a simulation).
>So why the fuck don't they include one then? Sell you an >overpriced machine, and then make you spend another >$30 on a real mouse? WTF?
Simple. The OS is not so fatally flawed as to require two (or three) buttons to use effectively and, for the vast majority of users, a single mouse button is all they ever need for anything that they do on the computer.
Professional/power users, who need 3 button mice (or 2 button, or 12 button, whatever), can buy the one that they want--since whatever version Apple choose it would not work for many users who even want a multi-button mouse would need (I want a 3 button mouse, with the third button under my thumb, built for a right-hand! I want the same, but a lefty design! I want a flat, 3 button optical mouse! etc.)
So instead they provide an elegant one-button optical mouse with a clickable surface (rather than buttons) that will work for virtually all of their non-power users and even their power users until they find the right replacement.
It also simplifies tech support if everyone has one, but that's another matter.
>I love my Linux desktop because *I*, not Steve J or Bill G, >am in charge. I can do what I want, I can combine, shells >are not hidden, and I have a choice of apps greater than >one (vs MS Office for Apple and not much more for the >Mac).
I can run GNome on a Mac via X11. I can also run OpenOffice and I've heard reports of people getting AbiWord up in running.
That being said, if you can't do something in Linux, what is your response? If you say that you can "write it yourself" that is certainly true, but no different than for the Mac. Yes, you can theoretically tweak anything in the OS, but knowing the calls and where to do that are not necessarily trivial things!
Just because it can be done doesn't mean that you can personally do it.
Either that or they got tired of waiting.
on
Safari 1.0 Released
·
· Score: 4, Insightful
It is also possible that they just got sick of waiting for a good, standards-complient web-browser or that they got sick of being so dependent on MS for a decent browser.
Camino and Mozilla weren't quite up to snuf and have serious flaws, as things are; IE was years behind everything else, was too slow, poorly threaded, and had a host of other issues... OmniWeb and Standard Compliance didn't belong in the same sentence--particularly when it came to CSS, and Opera just plain Sucked on the Mac.
The rational choice, particularly for such an important app as a web browser, is in-house development.
1) I can run a network into an Airport Extreme hub and connect to it from a desktop that is in an awkward place to lay cat5.
2) To communicate with my laptop and either share an internet connection with it (there is a preference panel for this) or just do things like establish a network with it or stream music to it if I don't have a central airport hub.
While I don't have any benchmarks, it seems faster and more responsive overall. This wouldn't surprise me, since they've probably removed a good deal of the debug code.
It still has a bad habit of trying to deeplink itself into CNN every time I go there and a few rendering fragments when a text box crosses the address/status bar, but other than that it seems very solid as a release.
There are no real improvements in the prefs panel since last time either, which is unfortunate.
Safari 1.0 is now available through Apple's software update.
The new version seems noticibly faster and has no bug button, but there is still a "Report Bugs To Apple" option under the Safari menu.
Try using a Mac.
on
Jaguar is Over
·
· Score: 2, Informative
10.1->10.2 was *not* a "minor" update by any stretch of the imagination, nor is 10.2->10.3 going to be a "minor" update (unless you call adding major features such as FileVault, an updated application suite, a font manager, iDisk syncing, a new appearance for the entire Finder, fast user switching, and a hundred other things "minor").
10.0->10.1, OTOH, *was* free.
Re:FreeBSD 5.0 as a lower-level enhancement?
on
Jaguar is Over
·
· Score: 1
Not emmulation anymore than it currently being built off of FreeBSD 4.4 is--it is built on top of it, there is no emmulation involved.
That being said, I suspect the same way they mean "low level" the same way assembly is "low level"--it does *not* mean less significant, it means something to the effect of "closer to the core."
"Longhorn users may be waiting until 2005 for their next release, but I doubt theyâ(TM)ll have spent $460 or $690 by that point on keeping their OS up to date."...and then they will pay all of it at once, considering MS's history of pricing their OS.
Also, this is going to come out at the end of this year. Assuming that we get one $130 update every year (including this year), then staying up-to-date through 2005 will be $390--not $690.
You've obviously never seen a transistor tech-sheet, I would hate to have to read the V-I diagrams if they were compressed to be smaller than the transistor, and that's just one chart.
Would you and the other misc people cool it with this myth?
There have been *exactly* two pieces of software that I wished ran on the Mac at one point or another, but didn't.
One is OrCAD/PSpice, which now that I'm out of school I could die happy if I never have to use it again.
Another is an editor called TorilEdit, which is DOS only. It being dirt simple to replicate (just time consuming), I am writing my own version of it--in Python.
I've don't lack for a word processor or document renderer (MS Word, AppleWorks, OpenOffice, TeXShop/teTeX), a web browser (OmniWeb, Camino, Safari), a Presentation tool (Powerpoint, Keynote), Statistical Software (R, JMP), graphing software (R, ChartSmith, proFit), ABM modeling toolkits (Ascape, Swarm), Diagraming/UML tools (OmniGraffle), compilers (java, gcc), IDEs (BBEdit, Project Builder), or anything else that I use (Goban, Deimos Rising, QuickBooks, cvs, nmap, etc, etc, etc)....and you will note a distinct lack of DTP, Audio, or Graphical (sans graphing) work in all of that.
"SPEC is largely OS independent. You spend almost 0% of your time in the actual OS when running the apps."
This, by itself, makes it 100% irrelevant for real-world performance.
"You use a geometric mean for SPEC. There is no argument."
Yes, geometric means are always used whenver you have any form of normalized data or ratios that you are trying to make comparisons over. This is not my point--though I probably didn't state this very well:-(
For all of their strengths when averaging normalized scores, they can severely bias central tendencies--the same way arithmatic means are sensative to extremes, so are geometric means. Even weighted geometric means run into problems in this regard.
Thus, it does very well or very poorly one area, that offsets the score dramatically. Determining central tendancy is a more complex problem than it appears even when we all agree to the nature of the data set.
Also, IIRC, the fortran interpreter used on the mac is not up to snuff--not as fast as it could be by a long shot.
I know I'm on/., but is it too much to ask that people read their own posts? I mean okay, so you don't read the article or the little blurb at the top, but at least read what you are writing?
Your original statement:
"Macs still are best for people who want to do audio, graphics, DTP work and little else. "
Your followup:
"Are you trying to suggest that Windows is better for all of the above?"
No, I think that the author was implying that you had never used a Mac because of the "and little else" statement, not because of the "audio, graphs, [and] DTP work" part.
Largely because SPEC is irrelevant and doesn't represent actual performance?
Seriously, have you ever studied the problem of comparing two platforms that are even running the same operating system? This can be pretty nontrivial by itself, and comparing between different architecutres running different OSes to simply generate a "performance score" can be downright impossible.
SPEC also has its own issues.
For instance, with SPEC, do you use an arithmatic mean or a geometric mean? Each have their own disadvantages and can bias data unfairly.
Next, the configuration of the platform which these things are tested on matters... a lot. Yet many companies produce their own black boxes to do this sort of testing.
Slashdot Mirror
"The best benchmark is the app you want to use"
Wisest advice I've ever heard--it was in my machine org and assembly textbook.
*Any* cross-platform benchmark should be taken with a shaker full of salt--they simply do not represent real world performance.
SPEC, for all of its nice points, also falls into this same category. In the end, when all is said and done, people prefer to confuse the model with reality--they think that real world performance follows SPEC scores.
Part of the problem is that Rowling has to balance both the game in her world for the audiences in her world and for /her/ audience (i.e., us).
She indicated that games of Quidditch often take hours and can take days (if memory serves she said one Quidditch World Cup game lasted 4 *days*). This would give a little more opportunity for either or both sides to get a 150 point lead (perhaps it is still not enough, but as you said: Rowling is a storyteller, not a game theorist). The length of time before the snitch is caught can help balance out the game.
Then she has to balance out what we are willing to read through. She wants to describe the action in Quidditch and give us a fast-paced feel for it, but she can't really do that if the game lasts too long. She tries to describe Harry's thoughts and follow the character through the game, but how long would we be willing to put up with that if the game lasted 12 hours?
So its somewhat likely that there are multiple issues involved in trying to balance the entire mess.
'compare The Two Towers where there are three simultaneous stories- and they're relentlessly literal where they could be surreal. "
Actually, this can be said to be part of litterary style--neither method is necessarily better than another anymore than you can make the unilateral statement "acrylic is better than oil in painting" or "oil is better than watercolor".
I took a class about a year ago in Writing Fiction with the teacher Joanne Greenberg (The Running of the Dear, I never promised you a rose garden, and many others). She explained that writing a good story with multiple points of view is more difficult than a singular point of view simply because every time you switched your point of view you had a tremendous potential for losing power with your audience. You can also amplify it, but it is signficiantly more difficult to do so.
Which one you choose is a matter of preference--it doesn't make your writing automatically "better" if you have 16 points of view which you follow throughout.
While I understand you are aiming for humor, I'm getting mildly sick of seeing this kind of comparison: You can turn almost any work into almost any other work by abstracting it sufficiently.
/landspeeder/ at first (x-wings come later) and he is skilled at racing his friends.
/in no way used each other/ as sources.
For instance, lets take the early Star Wars universe (neglecting, for a moment, the movie Hidden Fortress) and Atlas Shrugged, by Ayn Rand.
Both deal with a small group of individuals (bussinessmen or the rebel alliance) who split off from society and form their own communities which are being hunted by the predominant governments in their areas and fighting back against them. The dark side in each story has developed a superweapon (Xylophone/Death Star) which can obliterate large areas (cities/planets) and have a host of other similarities (both employ torture on the heroes!)
Let's also take a look at your x-wing/broomstick comparison. In the radio drama, where Luke's skill is really demonstrated, it is with a
Now lets compare with Harry, who is more like a very talented soccer player. It makes a lot of sense, considering Harry's world, that they would have at least one sport played on brooms (and brooms are an old throw-in here from other legends &c). Thus, if Harry is going to be a natural at that sport, then he is going to have to be a natural on the broom.
It should also be noted that Luke is portrayed as being more than slightly reckless with his landspeeder in order to get an edge in racing. Harry is *not*.
This is just a start, there are a *lot* of other problems with the Harry--Star Wars match-up. Saying that they have some base similarities is true, but that can be said of many things that
"It's pretty sad. Rowling always said she would NEVER sell the rights to Harry Potter. Now if you look at he books, there's a small message on the copyright page saying that Time Warner own the rights to all the character names and likenesses."
She didn't sell the rights to the books (IIRC), only to the "names and likenesses," which was probably a mandatory step moving it into the movie industry, having action figures, &c. She probably didn't want to deal with all of the paperwork of subcontracting each individual entity for these things (a company to make toys, &c) and it may have been a mandatory, and not terribly offensive part of the deal for her.
Looking at my copy of Book 4 (I don't have book 5 on hand) it says, in the jacket:
"Text copyright © 2000 by J.K. Rowling"
Thus, I would say calling her an "unmitigated sellout" is probably a bit harsh.
You, sir, are ignorant.
"they're loosing money here all this time the iTunes store isn't running in Europe. "
Hardly. They aren't making as much as they could, but they are not "losing money" because they are not in Europe.
"I think they'll have a lot to explain to shareholders, eg. 'Why didn't you pursue this business model as soon as it showed it's potential?'."
Not Apple's problem--there are laws and licensing issues involved that they are attempting to negotiate their way through. These are a more major issue than was first anticipated. Period. End of discussion.
If they are turning a profit with it now and/or will turn a profit once the Windows version comes out, why on earth would they need to justify that they haven't managed to work out licensing with Europe just yet?
They aren't exactly trying to explain to the teacher why they haven't gotten their homework in on time.
This is blatant hyperbole.
Windows access is relatively critical, but only from the perspective of competing with whatever MS manages to cobbel together that actually has a chance of competing (I haven't seen anything yet) and will likely have very little to do with the long term life-and-death of the project as it stands now (the record companies seem happy with the deal, and so long as they have the big-5 and the big-5 remain profitable, Apple will probably not have to float the store).
International access, on the other hand, is in no-way critical to whether the project succeeds or whether Apple succeeds. The problems are the laws in other countries which are not conducive to this kind of pricing scheme, as well as the international contracts and licensing issues surrounding distribution of media.
If Apple and the big-5 are having trouble with it, so will anyone else who wants to expand across country boundaries.
Thus, the two *most* critical issues for the iTunesMusicStore, at this stage, are:
1) Getting more labels aboard (they seem to be doing an admirable job at getting this set up).
2) Getting iTunes ported to Windows (once again).
These would be the "logical next steps" to remain in front of the pack. International access is more of a nicety that is not overly critical for their success (though I'm sure they want to do it as quickly as possible--more profits are never a bad thing).
"imagine remotely accessing a machine in Finland, using that machine to download from Apple's store at a cheaper rate than your home country, and then downloading from the Finland machine to yours. "
This would probably be countered the same way that it is being handled now to prevent people from buying music this way from Finland by routing through a US American server--by checking the credit card address.
Thus, if your credit card has a home address in Estonia (and a Credit Card *is* required to use the store), then you won't be able to use the iTunes store through Finland after Finland is included in the plan or through the United States today.
It is called "market lock-in"
It doesn't matter if it is a better product--someone will ask their friends "what will work for me?" their friends say "I use this, it works for me" and that prompts said person to go out and buy X.
Most people I talk to I can sway to buying a Mac--if I get to them first and let them get their hands on one.
"Nope. If it's not threaded properly, it will not take advantage of multiple processors. The application will run on one processor."
Yes, except for one thing: you generally have more than one application running, and each of them can have more time because you have two processors. Thus, one process (such as one application) can receive a huge speed improvement simply because it doesn't get switched off of the processor as often and doesn't have to make as much room for other things to run.
If you don't believe me do the simulation yourself. Draw out 10 processes on paper, don't forget the OS kicking in to decide who goes next, and assume a Shortest Job First, Preemptive (you can ignore all of the other fancy things, this is just a simulation).
Another poster was kind enough to point out to me that you can enable the bug button from inside of the "View" menu.
>So why the fuck don't they include one then? Sell you an
>overpriced machine, and then make you spend another
>$30 on a real mouse? WTF?
Simple. The OS is not so fatally flawed as to require two (or three) buttons to use effectively and, for the vast majority of users, a single mouse button is all they ever need for anything that they do on the computer.
Professional/power users, who need 3 button mice (or 2 button, or 12 button, whatever), can buy the one that they want--since whatever version Apple choose it would not work for many users who even want a multi-button mouse would need (I want a 3 button mouse, with the third button under my thumb, built for a right-hand! I want the same, but a lefty design! I want a flat, 3 button optical mouse! etc.)
So instead they provide an elegant one-button optical mouse with a clickable surface (rather than buttons) that will work for virtually all of their non-power users and even their power users until they find the right replacement.
It also simplifies tech support if everyone has one, but that's another matter.
>I love my Linux desktop because *I*, not Steve J or Bill G,
>am in charge. I can do what I want, I can combine, shells
>are not hidden, and I have a choice of apps greater than
>one (vs MS Office for Apple and not much more for the
>Mac).
I can run GNome on a Mac via X11. I can also run OpenOffice and I've heard reports of people getting AbiWord up in running.
That being said, if you can't do something in Linux, what is your response? If you say that you can "write it yourself" that is certainly true, but no different than for the Mac. Yes, you can theoretically tweak anything in the OS, but knowing the calls and where to do that are not necessarily trivial things!
Just because it can be done doesn't mean that you can personally do it.
It is also possible that they just got sick of waiting for a good, standards-complient web-browser or that they got sick of being so dependent on MS for a decent browser.
Camino and Mozilla weren't quite up to snuf and have serious flaws, as things are; IE was years behind everything else, was too slow, poorly threaded, and had a host of other issues... OmniWeb and Standard Compliance didn't belong in the same sentence--particularly when it came to CSS, and Opera just plain Sucked on the Mac.
The rational choice, particularly for such an important app as a web browser, is in-house development.
>Why WI-FI in a desktop box?
:-p
A couple of reasons that I can think of.
1) I can run a network into an Airport Extreme hub and connect to it from a desktop that is in an awkward place to lay cat5.
2) To communicate with my laptop and either share an internet connection with it (there is a preference panel for this) or just do things like establish a network with it or stream music to it if I don't have a central airport hub.
3) Because it is 1337
While I don't have any benchmarks, it seems faster and more responsive overall. This wouldn't surprise me, since they've probably removed a good deal of the debug code.
It still has a bad habit of trying to deeplink itself into CNN every time I go there and a few rendering fragments when a text box crosses the address/status bar, but other than that it seems very solid as a release.
There are no real improvements in the prefs panel since last time either, which is unfortunate.
Safari 1.0 is now available through Apple's software update.
The new version seems noticibly faster and has no bug button, but there is still a "Report Bugs To Apple" option under the Safari menu.
10.1->10.2 was *not* a "minor" update by any stretch of the imagination, nor is 10.2->10.3 going to be a "minor" update (unless you call adding major features such as FileVault, an updated application suite, a font manager, iDisk syncing, a new appearance for the entire Finder, fast user switching, and a hundred other things "minor").
10.0->10.1, OTOH, *was* free.
Not emmulation anymore than it currently being built off of FreeBSD 4.4 is--it is built on top of it, there is no emmulation involved.
That being said, I suspect the same way they mean "low level" the same way assembly is "low level"--it does *not* mean less significant, it means something to the effect of "closer to the core."
"Longhorn users may be waiting until 2005 for their next release, but I doubt theyâ(TM)ll have spent $460 or $690 by that point on keeping their OS up to date." ...and then they will pay all of it at once, considering MS's history of pricing their OS.
Also, this is going to come out at the end of this year. Assuming that we get one $130 update every year (including this year), then staying up-to-date through 2005 will be $390--not $690.
You've obviously never seen a transistor tech-sheet, I would hate to have to read the V-I diagrams if they were compressed to be smaller than the transistor, and that's just one chart.
Would you and the other misc people cool it with this myth?
...and you will note a distinct lack of DTP, Audio, or Graphical (sans graphing) work in all of that.
There have been *exactly* two pieces of software that I wished ran on the Mac at one point or another, but didn't.
One is OrCAD/PSpice, which now that I'm out of school I could die happy if I never have to use it again.
Another is an editor called TorilEdit, which is DOS only. It being dirt simple to replicate (just time consuming), I am writing my own version of it--in Python.
I've don't lack for a word processor or document renderer (MS Word, AppleWorks, OpenOffice, TeXShop/teTeX), a web browser (OmniWeb, Camino, Safari), a Presentation tool (Powerpoint, Keynote), Statistical Software (R, JMP), graphing software (R, ChartSmith, proFit), ABM modeling toolkits (Ascape, Swarm), Diagraming/UML tools (OmniGraffle), compilers (java, gcc), IDEs (BBEdit, Project Builder), or anything else that I use (Goban, Deimos Rising, QuickBooks, cvs, nmap, etc, etc, etc).
"SPEC is largely OS independent. You spend almost 0% of your time in the actual OS when running the apps."
:-(
This, by itself, makes it 100% irrelevant for real-world performance.
"You use a geometric mean for SPEC. There is no argument."
Yes, geometric means are always used whenver you have any form of normalized data or ratios that you are trying to make comparisons over. This is not my point--though I probably didn't state this very well
For all of their strengths when averaging normalized scores, they can severely bias central tendencies--the same way arithmatic means are sensative to extremes, so are geometric means. Even weighted geometric means run into problems in this regard.
Thus, it does very well or very poorly one area, that offsets the score dramatically. Determining central tendancy is a more complex problem than it appears even when we all agree to the nature of the data set.
Also, IIRC, the fortran interpreter used on the mac is not up to snuff--not as fast as it could be by a long shot.
I know I'm on /., but is it too much to ask that people read their own posts? I mean okay, so you don't read the article or the little blurb at the top, but at least read what you are writing?
Your original statement:
"Macs still are best for people who want to do audio, graphics, DTP work and little else. "
Your followup:
"Are you trying to suggest that Windows is better for all of the above?"
No, I think that the author was implying that you had never used a Mac because of the "and little else" statement, not because of the "audio, graphs, [and] DTP work" part.
Largely because SPEC is irrelevant and doesn't represent actual performance?
Seriously, have you ever studied the problem of comparing two platforms that are even running the same operating system? This can be pretty nontrivial by itself, and comparing between different architecutres running different OSes to simply generate a "performance score" can be downright impossible.
SPEC also has its own issues.
For instance, with SPEC, do you use an arithmatic mean or a geometric mean? Each have their own disadvantages and can bias data unfairly.
Next, the configuration of the platform which these things are tested on matters... a lot. Yet many companies produce their own black boxes to do this sort of testing.