Somebody else said it first, but DNS was not supposed to be a keyword system.
But yet that's exactly how people have treated *.COM, and everytime we have a story about squatting or ICANN or CompanyReallySucks.com, slashdot gets up in arms and starts proposing reforms and reorgs and so on.
So some company comes along and tries to make a keyword system apart from DNS, and what's Slashdot's reaction. Well, they suck too, the leeching bastards. I don't get it.
Alternate NICs are not a solution the abuse of DNS -- they do nothing more than attempt to extend the misuse of the system.
Keywords and Search are the solution, and RealNames was not a bad idea. Lets put the stupid squatting and trademark junk in their database instead of our DNS system. Let them worry about the problem of how to find "Something" on the internet without resorting to "Something.com".
Having RealNames punted by MS is a good thing in this respect. Hopefully browser developers will come up with a system where you can plug in as many Keyword Systems as you'd like (just like you can with search engines). I wouldn't mind an automatic Google Are You Feeling Lucky? feature in my address bar.
I've been doing this for a while and in something like 90% of the cases the client proactively demands ownership of the code. And it's Work For Hire, so that makes sense.
This can get tricky because you might want to use some of it for in house code libraries and the like, and in some cases they have objected to using any pre-existing code and/or using any of "their" code for future projects. Yes, this affects the price, and yes you should get a contract signed that covers all of this.
Furthermore, there's the matter of good business sense. Even if you do own copyright, giving away what you just sold to your clients competitors doesn't sound like a good idea. It causes ill feelings when a developer is selling a app they were paid to write -- much worse if they just posted it on their website.
I'm curious why you think MS has a bad compatibility record. In my view, they value compatibilty far more than progress (or in their terms "innovation"), to the detriment of their userbase. Hense DOS/Windows lasting into the 21st century.
I've seen bizarre 10 year old PC business apps running without hickups on Win 2000, so I know that it works. And yeah, DEC & MS even did the hardware architecture thing with FX32 on Alpha, although it was optional so nobody bought it.
Now, Apple -- I've got a old Quadra sitting here, and I can tell you that most of the apps on the thing will not run on my PBG3/OSX. Now, that's not entirely Apple's fault, just that they refuse to maintain bug-compatible interfaces like Microsoft does. And that costs them users to some extent.
I guess what I'm getting at here is that Microsoft has enormous power, but they still can't tell people to rewrite their apps, buy new ones, or switch their CPU. But Apple has the balls to do those things, and that acts as a form of natural selection against thier marketshare.
Before System 6, Apple never had a real OS distribution, just little parts. If you wanted to know what you were running, you'd just Get Info on your "System" file, but your Finder and all the rest had completely different versions. They did increment the version number fairly quickly.
This was intentional - the "Macintosh Experience" was designed to obliviate the distinciton between software and the hardware. So it was against their philosphy to even admit there was an "OS". That only really changed when they started selling upgrads.
"But I don't hear anyone here on slashdot complaining about forcing gamers and other groups to a more NT OS foundation."
Bah. I was complaining they didn't do it 5 years earlier. But, they had markets to artifically segment.
That's the key difference between Apple as the vanguard leader of a small progressive wing and Microsoft as protector of the vast waddling middle of end-user computing -- MS took 9 years to complete the transition to it's Next Gen OS (longer if you count OS/2), while Apple has legacied the past in about 1 single year.
Of course, abrupt, compatiblity-breaking, "insanely great" progress is one big reason Apple's got 4% marketshare while MS has the rest.
TimeWarner was this overly huge conglomorate of smaller media companies that never seemed to make any (or very much) money.
I'm sure it had some other parts that were profitable, but everyone was going in different directions and the net result was nothing spectacular. Everytime TW sucked up another company, everyone predicted "synergy", but it never seemed to happen.
On the otherhand, AOL was very profitable and very good at coordinated marketing attacks.
Wall Street might had thought that new management at TimeWarner would allow it to better cross-market it's products (particularlly it's cable networks) and actually become profitable.
Why should MS recompile anything? They can just translate the binary from 32-bit to 64-bit opcodes. That is how they ported 8-bit 8080 BASIC to the 16-bit 8086 after all.
Did you install Sun Java 1.4? It has the "feature" of taking over Java support from the MS JVM. On my box, this almost always results in a hung browser.
You can disable this behavior in the Java control panel.
I have trouble trusting any random person's perception of reliability (unless say they come from a VAX background:)
People are just too beaten down with poor expectations bred by years of working with crappy hardware, crappy OSes (DOS/Win or MacOS), and beta-level video drivers. In the average person mind, a few reboots a week is probably perfectly acceptable as "rock solid". Plus they may feel the need to conform to the majority opinion that (Mobo) or (Chipset) is a good thing and avoid getting flamed.
It's also tough because the whole hardware hobbiest culture (the people who are hands-on and publishing the info) seem to treat their computers as disposable game consoles with 6 month upgrade cycles. For the most part they reject stability features like ECC RAM in favor of miniscule performance increases.
Anyway, I would kill for some reliable information about reliabity. My expectation has been set by my last couple boxes (Compaq and IBM 'workstation' stuff) that could run NT4/5 24x7 without trouble.
The problem is not unsolvable or necessarily overly tedious, just at some level we have to give up our faith in experts.
A) It was a compliant that Netscape liked to dictate what it thought it should be standards to the W3C and did not like to participate. Microsoft buddied up to the W3C committees for strategic reasons. Slashdot user Zeinfeld discusses this frequently.
B) If you go up the thread crappy market penetration is exactly what I'm complaining about -- only because it puts the Mozilla project's funding seriously at risk.
Within the last year or so, Netscape 4 lost 50% of it's userbase, almost entirely to IE. The remaining Netscape holdouts need a kick-in-the-ass to get up to modern standards. Otherwise, why shouldn't AOL take the Gecko engine in-house for their client and shut down Netscape? Why blow your wad for a 1% marketshare browser?
(Oh, and the last message was posted AC accidentally. I obviously created this account for a reason.)
I dunno, everything else I do seems "smooth" enough (well, except some new games). Hopefully two Pentium III CPUs and a half-gig of RAM isn't too wussy for Mozilla, because if it is, I'll just use something else.
One possibility is that it could be a SMP-related issue. Off to bugzilla.
Great defense of Netscape 4.0 pal. Although, you might want to check your watch though, because it's not 1996 anymore -- it's 2002, and most Netscape users are still clinging dearly to their old static and flawed product.
That's bad for Netscape, bad for the Mozilla project, and bad for the WWW. The only one winning in this situation is Microsoft, as their marketshare shoots above 90%.
Netscape 4 will crash/misrender on perfectly compliant webpages. Netscape also never released a "standards based web browser" until 6.x. The W3C CSS spec dates from 1996 and has nothing directly to do with Microsoft. Furthermore, Netscape 4 doesn't ignore it, it tries to convert it to it's own proprietary stuff and ends up crapping itself.
Are you that much in love with your browser that you would fabricate a bunch of complete lies?
Plug-ins -- something that I've been curious about.
Mozilla is supposed to be this 'base' browser that can be branded by anyone.
But is seems like there is no centralized plug-in directory. Which would mean that 3rd parties like Real have to write installers which handle each particular branded version of Mozilla which would lead to inevitable installation problems.
However, the idea is that "non-geeks" (who aren't doing testing of some sort) should use the Netscape releases, which do include Real/Flash/Java.
I'll agree that Mozilla renders fast -- my main complaint is that it "feels" jerky, unresponsive, or in layman's terms -- slow.
For example, if you are (say) loading a large slashdot page in the background, the UI and the scrolling of your foreground window becomes very unresponsive. This gets kind of annoying if you click the wrong link and find that your Stop button doesn't want to register and the page loads anyway. (2x PIII-600, 512MB, Win2K)
This is all probably threading issues rather than actual performance -- it's just that perceptually looks like a performance problem.
Also, IMO, the incremental renderer adds to this perception. On IE you might wait just as long, but when the page appears it looks right. Mozilla shows you various half-done bizarro-versions of the page along the way, which can look klunky on some sites.
(The graphs are interesting because they show the OS X version to be much slower than the Windows version. Yet because the competition is worse on Mac, Mozilla feels much better there for some reason, on much slower hardware than my Winbox.)
Sorry, pages that use Nuscrapisms like FONT tags instead of CSS are not "designed with a clue". Not to mention that completely proprietary LAYER stuff in the place of W3C standards.
But I'm glad you came out of the woodwork as an example of the embittered Netscape 4 user. You'd rather fight than switch, even as the noose of the modern www tightens around your neck. 6% marketshare and declining -- don't expect to see many clueful NS4 compatible sites coming on line in the future.
Mhz ratings are not only confusing, they are misleading.
Actually, what I would argue is misleading is the entire way that computers are sold to the Home/SOHO "Joe Sixpack" market. Wrapping up a "2000+" or "2.0 Ghz" CPU with cheap components, lowspec memory, and some doodads and marketing it as a high end system is the real point of confusion. I think we all know that a much lower-clocked system could beat most of the stuff you find at retail. Both AMD and Intel are playing that game rather than trying to improve the situation.
For the corporate markets however, Mhz isn't as imporant as everyone's making out, and that's where AMD really needs to grow. Yet their PR scheme is causing some level of confusion and distrust among semi-informed customers. As you point out, it's risky marketing.
If only more Netscape 4 users would act like your coworker and give Netscape 6 another shot. Most of them took a look at 6.0 and went right back to their "beloved" crap NS4 broswer.
Stats I've seen show like a 5:1 advantage of NS4 over NS6/Mozilla. With many more NS4 users bleeding over to IE than upgrading to NS6.
Not that it really matters, but attracting users would help ensure that the Mozilla project continues to get it's funding from AOL/Netscape.
Based on this data, what really happened, what is really happening, and what disinterested parties seem to believe, I would conclude that the AMD PR Rating system provides a very nice comparison of Athlon performance relative to P4 performance at the clockspeed of the PR rating.
Intel funded FUD or not, the thrust of the report is because the PR ratings are based on pseudo-objective measurements, they are ultimately confusing to buyers.
Mhz might be a very poor buying consideration, but at least it's an objective number, unlike AMD's rating system. Thus AMD has put the informed buyer (such as yourself) in the position where you need to independently "conclude" whether it's fair or not. No matter what we all think about the P4 2.2A, we all conclude unanimously that it actually runs at ~2.2Ghz.
Even someone like you who is informed and thinks highly of AMD CPUs have confused the PR Rating as a "Pentium Rating" -- it's actually (supposedly) a benchmark comparison with a 1Ghz K7 chip, not a Pentium 4.
In practice though, AMD bumps up 66Mhz and adds 100 PR points every time. As you point out, everytime they do this, their lead at a particular "rating" becomes narrower over Intel. If I was them, I'd be very worried about the perception that they over-speced the PR numbers -- if the "2500+" chip benchmarks slower than a P4 2.5Ghz, they are going to be blasted to high heaven -- even it's supposedly not a directly comparsion.
The other issue is that AMD is trying to make inroads into the server and workstation markets, and it's really questionable if "PR" ratings are needed there.
I'm a member of the target market (PC user who picked up a used PowerBook G3 to run OS X), and the mouse button issue (and pressing ctrl) is annoying as hell.
They might plonk me, but I'm going to suggest that they put some of that world famous HCI talent to use and design a single mouse button that could be software-switched into 2 button mode.
The other big bitch I have is the braindead keyboard which is missing several standard keys (PgUp/PgDn/Home/End, and most importantly FORWARD DELETE). Not having these keys basically means I have to fight muscle memory while editing text which is very grating. Plus, I'm going to point out that my olde Mac SE had a full extended keyboard, so removing keys is something new over at Apple.
It's gotten so that when I have to 'real work' on the powerbook, I get to frustrated and walk across the house to fire up the PC. All due to the lack of a couple extra keys/buttons.
Don't get defensive because Apple's asking for input. I love the OS and the environment, but these minor HCI issues really are the difference between me buying a G4 PB and not buying one.
IBM's own LAN Manager networking was rock solid once you had it set up properly, but as others have pointed out, it was a bitch to set up.
Just for the outsiders, IBM's networking was the faimilar NetBEUI or SMB-over-TCP/IP. Except they had some Dynamic DNS thing in addition to NBNS (WINS). As pointed out by the Samba guys, "SMB sucks, it really really sucks". Still, I'd take the NT implementation of this over the OS/2 version any day, where it's pretty much plug-n-play.
I lived through the pain of the NetWare drivers -- they didn't even use the standard NDIS hardware drivers, but instead required a seperate ODI card driver and some sort of bitchy NDIS-ODI shim. They also liked to forget where the NDS tree was and crash the system.
Not only was IBM's networking a bitch to setup, it was even a bitch to order. You had to understand a mess of TLAs and ETLAs and FRU numbers to even get right software. Little things like TCP/IP support cost more than the entire OS. The administrative tools were some of the fuglist things I ever pushed a mouse pointer at.
It was so bad that I'm convinced that they actively were discouraging you from putting OS/2 on a LAN -- Just run your 3270 client and be happy.
Too little, too late, IBM fixed these issues with the Warp Connect Edition. By then, we'd converted our OS/2 boxes to NT.
I always saw this as the core flaw in IBM's marketing that killed OS/2 and almost killed the entire company. While the world was going ga-ga over Client-Server computing, IBM was slogging forward with this this mainframe-centric (SNA) view of computing that nobody outside of a few Big Blue shops was buying into.
Consequentally, OS/2 would come with various terminal emulators and such, but a LAN client or a TCP/IP stack cost extra $$$ and wasn't easy to order. Meanwhile MS was pushing out products based on TCP/IP, RPC, ODBC, and agressively marketing into the small server space, and UNIX ("the network is the computer") use was exploding.
IBM acted like the PC was nothing more than a glorified dumb terminal and seemed to want to 'keep it it's place' in order to protect their fat midrange profit margins. That probably explains their completely half-hearted software development for the platform, and their unwillingness to offer anything that looked like 'server' hardware in PC space.
When IBM finally figured out they were getting slaughtered by UNIX on the high-end and Windows on the low-end, they moved quickly. But by then it was too late for OS/2.
C) IBM Released a version of OS/2 called "Extended Edition" which supposedly only ran on PS/2 machines. This lead to the perception that OS/2 only ran on PS/2 period (when in fact even OS/2 EE ran on some clones).
What was true is that IBM was very slow in promoting 3rd party compatibility for OS/2. After Microsoft ditched them, you couldn't get a Hardware Compatibility List or any such information that didn't only list IBM model numbers. The most infamous thing about OS/2 2.0 was that out-of-box it would only print to IBM-branded printers, for example.
Also, the marketing materials continually linked OS/2 and PS/2 -- Look at the names! IBM certainly wanted you think that they were peas-in-a-pod. (Which was a PR disaster for OS/2 as people rejected MCA.) So it should be understandable that people think that OS/2 only ran on PS/2 -- that's exactly the message IBM was spreading.
Slashdot Mirror
Somebody else said it first, but DNS was not supposed to be a keyword system.
But yet that's exactly how people have treated *.COM, and everytime we have a story about squatting or ICANN or CompanyReallySucks.com, slashdot gets up in arms and starts proposing reforms and reorgs and so on.
So some company comes along and tries to make a keyword system apart from DNS, and what's Slashdot's reaction. Well, they suck too, the leeching bastards. I don't get it.
Alternate NICs are not a solution the abuse of DNS -- they do nothing more than attempt to extend the misuse of the system.
Keywords and Search are the solution, and RealNames was not a bad idea. Lets put the stupid squatting and trademark junk in their database instead of our DNS system. Let them worry about the problem of how to find "Something" on the internet without resorting to "Something.com".
Having RealNames punted by MS is a good thing in this respect. Hopefully browser developers will come up with a system where you can plug in as many Keyword Systems as you'd like (just like you can with search engines). I wouldn't mind an automatic Google Are You Feeling Lucky? feature in my address bar.
I've been doing this for a while and in something like 90% of the cases the client proactively demands ownership of the code. And it's Work For Hire, so that makes sense.
This can get tricky because you might want to use some of it for in house code libraries and the like, and in some cases they have objected to using any pre-existing code and/or using any of "their" code for future projects. Yes, this affects the price, and yes you should get a contract signed that covers all of this.
Furthermore, there's the matter of good business sense. Even if you do own copyright, giving away what you just sold to your clients competitors doesn't sound like a good idea. It causes ill feelings when a developer is selling a app they were paid to write -- much worse if they just posted it on their website.
I'm curious why you think MS has a bad compatibility record. In my view, they value compatibilty far more than progress (or in their terms "innovation"), to the detriment of their userbase. Hense DOS/Windows lasting into the 21st century.
I've seen bizarre 10 year old PC business apps running without hickups on Win 2000, so I know that it works. And yeah, DEC & MS even did the hardware architecture thing with FX32 on Alpha, although it was optional so nobody bought it.
Now, Apple -- I've got a old Quadra sitting here, and I can tell you that most of the apps on the thing will not run on my PBG3/OSX. Now, that's not entirely Apple's fault, just that they refuse to maintain bug-compatible interfaces like Microsoft does. And that costs them users to some extent.
I guess what I'm getting at here is that Microsoft has enormous power, but they still can't tell people to rewrite their apps, buy new ones, or switch their CPU. But Apple has the balls to do those things, and that acts as a form of natural selection against thier marketshare.
Before System 6, Apple never had a real OS distribution, just little parts. If you wanted to know what you were running, you'd just Get Info on your "System" file, but your Finder and all the rest had completely different versions. They did increment the version number fairly quickly.
This was intentional - the "Macintosh Experience" was designed to obliviate the distinciton between software and the hardware. So it was against their philosphy to even admit there was an "OS". That only really changed when they started selling upgrads.
"But I don't hear anyone here on slashdot complaining about forcing gamers and other groups to a more NT OS foundation."
Bah. I was complaining they didn't do it 5 years earlier. But, they had markets to artifically segment.
That's the key difference between Apple as the vanguard leader of a small progressive wing and Microsoft as protector of the vast waddling middle of end-user computing -- MS took 9 years to complete the transition to it's Next Gen OS (longer if you count OS/2), while Apple has legacied the past in about 1 single year.
Of course, abrupt, compatiblity-breaking, "insanely great" progress is one big reason Apple's got 4% marketshare while MS has the rest.
TimeWarner was this overly huge conglomorate of smaller media companies that never seemed to make any (or very much) money.
I'm sure it had some other parts that were profitable, but everyone was going in different directions and the net result was nothing spectacular. Everytime TW sucked up another company, everyone predicted "synergy", but it never seemed to happen.
On the otherhand, AOL was very profitable and very good at coordinated marketing attacks.
Wall Street might had thought that new management at TimeWarner would allow it to better cross-market it's products (particularlly it's cable networks) and actually become profitable.
Why should MS recompile anything? They can just translate the binary from 32-bit to 64-bit opcodes. That is how they ported 8-bit 8080 BASIC to the 16-bit 8086 after all.
Having "punch the monkey" crash your browser radically limits what you are able to view.
Java 1.4 works fine for me other places. It's no good in IE, at least for me.
Did you install Sun Java 1.4? It has the "feature" of taking over Java support from the MS JVM. On my box, this almost always results in a hung browser.
You can disable this behavior in the Java control panel.
I have trouble trusting any random person's perception of reliability (unless say they come from a VAX background :)
People are just too beaten down with poor expectations bred by years of working with crappy hardware, crappy OSes (DOS/Win or MacOS), and beta-level video drivers. In the average person mind, a few reboots a week is probably perfectly acceptable as "rock solid". Plus they may feel the need to conform to the majority opinion that (Mobo) or (Chipset) is a good thing and avoid getting flamed.
It's also tough because the whole hardware hobbiest culture (the people who are hands-on and publishing the info) seem to treat their computers as disposable game consoles with 6 month upgrade cycles. For the most part they reject stability features like ECC RAM in favor of miniscule performance increases.
Anyway, I would kill for some reliable information about reliabity. My expectation has been set by my last couple boxes (Compaq and IBM 'workstation' stuff) that could run NT4/5 24x7 without trouble.
The problem is not unsolvable or necessarily overly tedious, just at some level we have to give up our faith in experts.
A) It was a compliant that Netscape liked to dictate what it thought it should be standards to the W3C and did not like to participate. Microsoft buddied up to the W3C committees for strategic reasons. Slashdot user Zeinfeld discusses this frequently.
B) If you go up the thread crappy market penetration is exactly what I'm complaining about -- only because it puts the Mozilla project's funding seriously at risk.
Within the last year or so, Netscape 4 lost 50% of it's userbase, almost entirely to IE. The remaining Netscape holdouts need a kick-in-the-ass to get up to modern standards. Otherwise, why shouldn't AOL take the Gecko engine in-house for their client and shut down Netscape? Why blow your wad for a 1% marketshare browser?
(Oh, and the last message was posted AC accidentally. I obviously created this account for a reason.)
I dunno, everything else I do seems "smooth" enough (well, except some new games). Hopefully two Pentium III CPUs and a half-gig of RAM isn't too wussy for Mozilla, because if it is, I'll just use something else.
One possibility is that it could be a SMP-related issue. Off to bugzilla.
Great defense of Netscape 4.0 pal. Although, you might want to check your watch though, because it's not 1996 anymore -- it's 2002, and most Netscape users are still clinging dearly to their old static and flawed product.
That's bad for Netscape, bad for the Mozilla project, and bad for the WWW. The only one winning in this situation is Microsoft, as their marketshare shoots above 90%.
My god you are delusional crackmonkey.
Netscape 4 will crash/misrender on perfectly compliant webpages. Netscape also never released a "standards based web browser" until 6.x. The W3C CSS spec dates from 1996 and has nothing directly to do with Microsoft. Furthermore, Netscape 4 doesn't ignore it, it tries to convert it to it's own proprietary stuff and ends up crapping itself.
Are you that much in love with your browser that you would fabricate a bunch of complete lies?
My fav Model M variant is the TrackPoint II. (Just get a USB mouse if you want more than 2 buttons.)
Plug-ins -- something that I've been curious about.
Mozilla is supposed to be this 'base' browser that can be branded by anyone.
But is seems like there is no centralized plug-in directory. Which would mean that 3rd parties like Real have to write installers which handle each particular branded version of Mozilla which would lead to inevitable installation problems.
However, the idea is that "non-geeks" (who aren't doing testing of some sort) should use the Netscape releases, which do include Real/Flash/Java.
I'll agree that Mozilla renders fast -- my main complaint is that it "feels" jerky, unresponsive, or in layman's terms -- slow.
For example, if you are (say) loading a large slashdot page in the background, the UI and the scrolling of your foreground window becomes very unresponsive. This gets kind of annoying if you click the wrong link and find that your Stop button doesn't want to register and the page loads anyway. (2x PIII-600, 512MB, Win2K)
This is all probably threading issues rather than actual performance -- it's just that perceptually looks like a performance problem.
Also, IMO, the incremental renderer adds to this perception. On IE you might wait just as long, but when the page appears it looks right. Mozilla shows you various half-done bizarro-versions of the page along the way, which can look klunky on some sites.
(The graphs are interesting because they show the OS X version to be much slower than the Windows version. Yet because the competition is worse on Mac, Mozilla feels much better there for some reason, on much slower hardware than my Winbox.)
Sorry, pages that use Nuscrapisms like FONT tags instead of CSS are not "designed with a clue". Not to mention that completely proprietary LAYER stuff in the place of W3C standards.
But I'm glad you came out of the woodwork as an example of the embittered Netscape 4 user. You'd rather fight than switch, even as the noose of the modern www tightens around your neck. 6% marketshare and declining -- don't expect to see many clueful NS4 compatible sites coming on line in the future.
Mhz ratings are not only confusing, they are misleading.
Actually, what I would argue is misleading is the entire way that computers are sold to the Home/SOHO "Joe Sixpack" market. Wrapping up a "2000+" or "2.0 Ghz" CPU with cheap components, lowspec memory, and some doodads and marketing it as a high end system is the real point of confusion. I think we all know that a much lower-clocked system could beat most of the stuff you find at retail. Both AMD and Intel are playing that game rather than trying to improve the situation.
For the corporate markets however, Mhz isn't as imporant as everyone's making out, and that's where AMD really needs to grow. Yet their PR scheme is causing some level of confusion and distrust among semi-informed customers. As you point out, it's risky marketing.
If only more Netscape 4 users would act like your coworker and give Netscape 6 another shot. Most of them took a look at 6.0 and went right back to their "beloved" crap NS4 broswer.
Stats I've seen show like a 5:1 advantage of NS4 over NS6/Mozilla. With many more NS4 users bleeding over to IE than upgrading to NS6.
Not that it really matters, but attracting users would help ensure that the Mozilla project continues to get it's funding from AOL/Netscape.
Based on this data, what really happened, what is really happening, and what disinterested parties seem to believe, I would conclude that the AMD PR Rating system provides a very nice comparison of Athlon performance relative to P4 performance at the clockspeed of the PR rating.
Intel funded FUD or not, the thrust of the report is because the PR ratings are based on pseudo-objective measurements, they are ultimately confusing to buyers.
Mhz might be a very poor buying consideration, but at least it's an objective number, unlike AMD's rating system. Thus AMD has put the informed buyer (such as yourself) in the position where you need to independently "conclude" whether it's fair or not. No matter what we all think about the P4 2.2A, we all conclude unanimously that it actually runs at ~2.2Ghz.
Even someone like you who is informed and thinks highly of AMD CPUs have confused the PR Rating as a "Pentium Rating" -- it's actually (supposedly) a benchmark comparison with a 1Ghz K7 chip, not a Pentium 4.
In practice though, AMD bumps up 66Mhz and adds 100 PR points every time. As you point out, everytime they do this, their lead at a particular "rating" becomes narrower over Intel. If I was them, I'd be very worried about the perception that they over-speced the PR numbers -- if the "2500+" chip benchmarks slower than a P4 2.5Ghz, they are going to be blasted to high heaven -- even it's supposedly not a directly comparsion.
The other issue is that AMD is trying to make inroads into the server and workstation markets, and it's really questionable if "PR" ratings are needed there.
I'm a member of the target market (PC user who picked up a used PowerBook G3 to run OS X), and the mouse button issue (and pressing ctrl) is annoying as hell.
They might plonk me, but I'm going to suggest that they put some of that world famous HCI talent to use and design a single mouse button that could be software-switched into 2 button mode.
The other big bitch I have is the braindead keyboard which is missing several standard keys (PgUp/PgDn/Home/End, and most importantly FORWARD DELETE). Not having these keys basically means I have to fight muscle memory while editing text which is very grating. Plus, I'm going to point out that my olde Mac SE had a full extended keyboard, so removing keys is something new over at Apple.
It's gotten so that when I have to 'real work' on the powerbook, I get to frustrated and walk across the house to fire up the PC. All due to the lack of a couple extra keys/buttons.
Don't get defensive because Apple's asking for input. I love the OS and the environment, but these minor HCI issues really are the difference between me buying a G4 PB and not buying one.
IBM's own LAN Manager networking was rock solid once you had it set up properly, but as others have pointed out, it was a bitch to set up.
Just for the outsiders, IBM's networking was the faimilar NetBEUI or SMB-over-TCP/IP. Except they had some Dynamic DNS thing in addition to NBNS (WINS). As pointed out by the Samba guys, "SMB sucks, it really really sucks". Still, I'd take the NT implementation of this over the OS/2 version any day, where it's pretty much plug-n-play.
I lived through the pain of the NetWare drivers -- they didn't even use the standard NDIS hardware drivers, but instead required a seperate ODI card driver and some sort of bitchy NDIS-ODI shim. They also liked to forget where the NDS tree was and crash the system.
Not only was IBM's networking a bitch to setup, it was even a bitch to order. You had to understand a mess of TLAs and ETLAs and FRU numbers to even get right software. Little things like TCP/IP support cost more than the entire OS. The administrative tools were some of the fuglist things I ever pushed a mouse pointer at.
It was so bad that I'm convinced that they actively were discouraging you from putting OS/2 on a LAN -- Just run your 3270 client and be happy.
Too little, too late, IBM fixed these issues with the Warp Connect Edition. By then, we'd converted our OS/2 boxes to NT.
OS/2 always did talk to the mainframes very well.
I always saw this as the core flaw in IBM's marketing that killed OS/2 and almost killed the entire company. While the world was going ga-ga over Client-Server computing, IBM was slogging forward with this this mainframe-centric (SNA) view of computing that nobody outside of a few Big Blue shops was buying into.
Consequentally, OS/2 would come with various terminal emulators and such, but a LAN client or a TCP/IP stack cost extra $$$ and wasn't easy to order. Meanwhile MS was pushing out products based on TCP/IP, RPC, ODBC, and agressively marketing into the small server space, and UNIX ("the network is the computer") use was exploding.
IBM acted like the PC was nothing more than a glorified dumb terminal and seemed to want to 'keep it it's place' in order to protect their fat midrange profit margins. That probably explains their completely half-hearted software development for the platform, and their unwillingness to offer anything that looked like 'server' hardware in PC space.
When IBM finally figured out they were getting slaughtered by UNIX on the high-end and Windows on the low-end, they moved quickly. But by then it was too late for OS/2.
C) IBM Released a version of OS/2 called "Extended Edition" which supposedly only ran on PS/2 machines. This lead to the perception that OS/2 only ran on PS/2 period (when in fact even OS/2 EE ran on some clones).
What was true is that IBM was very slow in promoting 3rd party compatibility for OS/2. After Microsoft ditched them, you couldn't get a Hardware Compatibility List or any such information that didn't only list IBM model numbers. The most infamous thing about OS/2 2.0 was that out-of-box it would only print to IBM-branded printers, for example.
Also, the marketing materials continually linked OS/2 and PS/2 -- Look at the names! IBM certainly wanted you think that they were peas-in-a-pod. (Which was a PR disaster for OS/2 as people rejected MCA.) So it should be understandable that people think that OS/2 only ran on PS/2 -- that's exactly the message IBM was spreading.