there was concern over the pressure of cell phone carriers who will be introducting competetive combo phone/DAP devices this year.
No they won't. They will be introducing cellphone/DAP devices that are in no way competitive with a standalone DAP device, because they will only be able to download songs from the carrier's own servers, just as they currently only download games at exorbitent rates from the carrier's servers. There is precicely ONE phone that will be introduced this year that works with normal servers, the Motorola iTunes-compatible phone. THAT phone is being stonewalled by the carriers.
Lastly, do you really, and take your time on this, really think that phone companies would charge $4.00 to download a song to a phone when iTMS sells songs for $0.99?
Absoluetly. If you buy an MP3 phone for $400 and it can only play songs you buy from Verizon or T-mobile for $4.00 each, and you can't download them from iTunes or install them from your computer, what else are you going to do.
This isn't hypothetical. This is precisely how the carriers operate for all the other software (games, ringtones, text messages) you access from your cellphones, and precisely why they're stonewalling the Motorola phone.
Except that there was no need for people with 2-3 year old machines to be left out of an upgrade cycle just because they didn't wait another year to get machines that were only partially supported with Mac OS X.
I've got neXTstep running right here on a NextStation Mono. This device is pretty much identical in specs to an LC475, with the same speed CPU, slightly less RAM, same expansion capabilities, comparable display.
ANY Powermac kicked its butt in every dimension.
And yet Nextstep on this 68030 was more responsive than OS 8 or OS 9 on my Powermac 7600/180. That's over 6 times the CPU speed, four times the VRAM, 8 times the RAM, and a bigger and faster disk drive.
Remember, we're not talking about Mac OS X's requirements, we're talking about whether Rhapsody was a misstep, and whether Apple's customers would be better off if Rhapsody had actually been shipped. Mac OS X is *much* heavier than Rhapsody, and Rhapsody would have been much faster than Mac OS 8 or 9 on any Powermac... all the way down to the 6100/60... when running Yellowbox applications.
Further, there was concern over the pressure of cell phone carriers who will be introducting competetive combo phone/DAP devices this year.
As far as I can see, the only competitive phone/DAP device out there is the one the carriers are trying to kill. The rest are "competitive" if paying $4.00 a song to rent it from your cellular carrier is "competitive". Hell, maybe it is, given what people are willing to pay for "ringtones".
Oh, and that one "competitive" phone pays royalties to Apple.
If MP3 players are rapidly approaching critical mass, how long will it take before car stereos will feature minijack inputs as standard?
Some time after cars start shipping with CD players as standard, and probably after they start shipping with data CD players, and by that time you'll be asking why cars aren't shipping with USB host for your iPod dock to plug into.
I mean, tanjit, cheap CD players have been on the market for what, 10 years, and they're still treating them as "upscale" options in cars.
Do tell o' wise one, where may I see this promised land of the MP3 market?
You make a 74 minute long playlist of DRMed AACs, WMAs, and RMs and save them to an audio CD. Then you click on "Import as MP3". Yeh, you lose a miniscule amount of audio quality that you MAY be able to hear if you listen to it, but when you're walking down the street or sitting on a train with earbuds connected to your MP3 player... you'll never notice it.
Panther: Expose, OK, though I prefer Peter Maurer's Witch. Massively broken Finder, Jaguar's was much more practical and less "in the way". Safari... uh, that came out in Jaguar. Fast User Switching... a big disappointment, free UNIX has had virtual consoles for years before OS X existed... and they work better. Preview... that came out in Jaguar as well. I can't comment on iChat AV, Filevailt, or Inkwell... I don't use them. Slight performance increase for some models. Several free third-party tools that worked under Jaguar required shareware upgrades for Panther.
If it wasn't for the fact that Panther came along when I upgraded to a Mac mini I'd still be using Jaguar.
Witness the trials of "program not responding."...
OS X has dumped me into similar situations, the main difference is that if I leave a Terminal window open I can do
% ps waux | grep program % kill process_id
and it goes away.
OSX will just unmount the drive when told to
Um, no, it won't. I've had to kill iTunes a couple of times to get my iPod unmounted. I also have quite a lot of hate about the way iTunes pretends it's not using my iPod Shuffle as a flash drive.
A Microsoft Windows user is besieged. And I mean not just with spyware and worms, but also with Windows Updates. They're doing the same thing as Apple's updates (make no mistake--both companies are giving you bug fixes), but there are so many updates for this mysterious vulnerability or that compromise that a typical home user is overwhelmed by not only by the OS prompting them to the point of annoyance that you have new Windows Updates as well as the number of patches and attacks.
Oh yes, that same Game box I have at home. I hadn't used it since some time before Christmas, but I brought it up recently to try out NASA World Wind and before I could run it I needed to install 39 "critical security updates" and two updates to.NET and Direct X.
Microsoft, having given away Windows XP Service Pack 2 for free eight months ago, still can't seem to convince enough of their users to adopt it to even hit the one-quarter mark.
Blow that for a joke, I'm sticking with Windows 2000 on my GameOS box. Why should I trust an OS like XP with a suicide switch in the kernel when the older version works fine?
Well, criticizing stuff that you're a fan of on/. gets comments like "MOD PARENT DOWN!" and "TROLL! HYPOCRITE!". I personally think Paul's analysis is really shallow and misses a lot of points, but he is at least being straight as he sees it.
You mean Microsoft's page to download SP4 for NT4.0?
No, I mean a web application that doesn't work with IE 6, or IE 5.4, or if IE 5.6 if that exists, but only IE 5.5 and only with certain obscure options set.
But I have to say your example is mighty hateful too.
I don't consider Rhapsody and "Yellow Box" missteps. What we have now in OS X, Cocoa, that's what it comes from. The problem wasn't anything to do with Rhapsody, it was with Adobe. Adobe refused to port their apps to what became Cocoa, forcing Apple to make this huge unweildy sidestep through Carbon that kept the abysmal OS 9 alive for years longer than it deserved, and where we ended up now is EXACTLY where we'd be if Apple had been able to convince Adobe and a few other key developers to stick with the program... except years later *and* most of the key apps that Apple bent over for ended up getting dropped anyway.
Spotlight is a HUGE step forward. It's killer functionality, and needs OS support to work right, but just a touch. OS X provides that support... without creating a new "search-based" file system.
Because you don't (despite what Be and Microsoft say) need a new file system to manage metadata, you just need a mechanism for applications to talk about it... WHEREVER it's stored. That's why Find on palm OS already exists, and has existed since 1996, despite Palm OS having a file system that's hardly worthy of the name. And that's why Spotlight works and Microsoft's dithering on WinFS.
And if I hadn't bought a Mac Mini I would have skipped Panther altogether because Panther *is* a relatively minor release. Tiger, with Spotlight, is a different kind of cat altogether.
I wish I had the power to ban applications like that.
I wish I still did. Unfortunately they brought us into line with the "corporate standards" and merged our Windows domains just in time for Blaster to hit. Lovely timing, that.
It improves capabilities too. There's some jobs you just can't reasonably tackle (read: they're a huge PITA) on 32-bit machines no matter how fast.
Multiple precision arithmetic libraries that make bignum (both extended and arbitrary precision) calculations trivial have existed since the '60s, and have been integrated into high performance programming languages (eg, Maclisp) since the '70s.
There is absolutely nothing I can think of for which the difference between 32- and 64- bit processing comes down to anything but a performance issue, except for large address spaces. And for GUI applications, Tiger doesn't provide a way to use large address spaces directly... and we're talking about laptops here.
I'm not into graphics stuff, but it's not difficult to imagine something like Photoshop benefitting
I'm only peripherally into graphics stuff, but I am into encryption, and it's hard for me to see anything that 64-bit integer code can do that Altivec can't do better in either area.
When Microsoft do the "right thing" (such as XP SP2),
Microsoft has yet to do the right thing. The security community has been beggng them to back out of the tight browser/desktop integration and "security zones" since 1997, and split the rendering and access functionality of the HTML control into separate components so you CAN run a locked-down sandboxed version of Internet Explorer if you want to... but instead Microsoft refuses to admit they made a mistake and patches symptom after symptom instead of attacking the disease.
That's why I, wearing my "security hat", banned all internet-capable applications that used the MS HTML control for rendering... back in 1997. As long as that ban was in effect we had zero virus and security panics, and we were the only division of our company for which that was the case.
The fundamental design of the HTML control is broken and unfixable. THe only solution is to back out of that design at a very low level, and rewrite all the applications that use it to handle access themselves. In 1997 I expected that Microsoft would do that... by now, it's obvious that they won't. They're afraid of losing face.
The right thing, from a security point of view, is to stop using Internet Explorer, Outlook, Outlook Express, Windows Media Player, Realplayer, and all other applications that use the MS HTML control to display potentially untrusted data whether they're shipped by Microsoft or some third party. Microsoft has proven over and over again for the last seven years that there is no other rational course of action.
SP2 and every other "security" patch that Microsoft provides are just smoke and mirrors.
Avoiding it means your systems are running on a legacy OS.
You say that like that's a bad thing.
How long before the legal or finance departments need to use a business-critical Web site that requires IE7 for access?
I don't know, you tell me: how long before some criminally stupid web developer creates a business-critical website that requires a specific version of a browser to even work? Not just "doesn't work on Firefox" (which is already in the "criminally stupid" department) but "doesn't work on recent versions of Internet Explorer"? Yes, I know, that's already happened... but in my case it was a website that didn't work on anything later than IE 5.5. Or older, either. Basically, Doctor Evil, this is a sword that cuts both ways.
I think 64-bitness is more important, and you think performance is more imporant.
I think 64-bitness is only important when it improves performance. And I've been working with real live gotta-be-fast 64-bit software since Apple was still using Nubus, and, well... the performance advantage of 64-bit is much overrated. If you really need 64-bit you know you really need it because you're running into 32-bit limitations.
The 64-bit address registers are the big win, especially for database apps. The 64-bit integer registers are nice, but Altivec already operates on multi-word objects... I'm not sure there's really a lot to gain there. Do you really know about algorithms that are faster in 64-bit mode, or are you guessing?
If by "Pentium-style", you mean "half as long and much wider than a Pentium",
No, I don't mean that. Unless by "Pentium-style" you mean "only Pentium 4" rather than "any Pentium-style processor", you don't mean that either.
The G3 had 4 stages, the G4 started with 4 and ended up with 7 or 9 depending on who you talk to, the G5 up to 25 for MIMD instructions... again depending on how you define "stages". Regardless, the G5 is a "brainiac" long-pipeline multiple-functional unit processor. This means it's more sensitive to pipeline stalls and cache misses, and requires much more power than a simpler design.
And, yes, I think "2 MHz sometime next year" is reasonably close to "3 MHz sometime next year we promise, honest, I know we said 2004, we meant 2005, uh, 2006" given that the G4 gets more work done per stage and thus, for typical code, will probably get more work done per clock IF it can get its bus problems resolved.
See, the G4 vs G5 question has a lot in common with the similar PIII vs P4 debate, except the G4 has had both arms tied behind its back by the crummy memory bus. If Freescale solves that problem, dual-core will just be icing on the cake.
That's why 64-bit OSes (including Tiger) can run 32-bit code.
You're being disengenuous. Tiger is primarily 32-bit, all the graphic code is still 32-bit, and any Cocoa software is 32-bit. It just has the ability to run 64-bit command-line programs, which is more than enough for most 64-bit software: for example, tthe 64-bit software I've been working on on Tru64 (which IS a native 64-bit OS) for most of the past 10 years... most of that is server code.
Getting back to the original point, for a laptop (which is what we're talking about) the ability to run 64-bit code is pointless to all but a tiny minority of the population... mostly software developers who want to test their server code when they're working on their laptop.
Slashdot Mirror
there was concern over the pressure of cell phone carriers who will be introducting competetive combo phone/DAP devices this year.
No they won't. They will be introducing cellphone/DAP devices that are in no way competitive with a standalone DAP device, because they will only be able to download songs from the carrier's own servers, just as they currently only download games at exorbitent rates from the carrier's servers. There is precicely ONE phone that will be introduced this year that works with normal servers, the Motorola iTunes-compatible phone. THAT phone is being stonewalled by the carriers.
Lastly, do you really, and take your time on this, really think that phone companies would charge $4.00 to download a song to a phone when iTMS sells songs for $0.99?
Absoluetly. If you buy an MP3 phone for $400 and it can only play songs you buy from Verizon or T-mobile for $4.00 each, and you can't download them from iTunes or install them from your computer, what else are you going to do.
This isn't hypothetical. This is precisely how the carriers operate for all the other software (games, ringtones, text messages) you access from your cellphones, and precisely why they're stonewalling the Motorola phone.
Except that there was no need for people with 2-3 year old machines to be left out of an upgrade cycle just because they didn't wait another year to get machines that were only partially supported with Mac OS X.
I've got neXTstep running right here on a NextStation Mono. This device is pretty much identical in specs to an LC475, with the same speed CPU, slightly less RAM, same expansion capabilities, comparable display.
ANY Powermac kicked its butt in every dimension.
And yet Nextstep on this 68030 was more responsive than OS 8 or OS 9 on my Powermac 7600/180. That's over 6 times the CPU speed, four times the VRAM, 8 times the RAM, and a bigger and faster disk drive.
Remember, we're not talking about Mac OS X's requirements, we're talking about whether Rhapsody was a misstep, and whether Apple's customers would be better off if Rhapsody had actually been shipped. Mac OS X is *much* heavier than Rhapsody, and Rhapsody would have been much faster than Mac OS 8 or 9 on any Powermac... all the way down to the 6100/60... when running Yellowbox applications.
Problem for apple is that everyone already has an ipod
200 people at work here. 3 of them, including me, have iPods. I guess your definition of "everyone" is different from mine.
A market containing only myself isn't really much of a market now, is it?
You're selling DRMed AACs, WMAs, or RMs to yourself?
YKINMK, but YKIOK.
Further, there was concern over the pressure of cell phone carriers who will be introducting competetive combo phone/DAP devices this year.
As far as I can see, the only competitive phone/DAP device out there is the one the carriers are trying to kill. The rest are "competitive" if paying $4.00 a song to rent it from your cellular carrier is "competitive". Hell, maybe it is, given what people are willing to pay for "ringtones".
Oh, and that one "competitive" phone pays royalties to Apple.
If MP3 players are rapidly approaching critical mass, how long will it take before car stereos will feature minijack inputs as standard?
Some time after cars start shipping with CD players as standard, and probably after they start shipping with data CD players, and by that time you'll be asking why cars aren't shipping with USB host for your iPod dock to plug into.
I mean, tanjit, cheap CD players have been on the market for what, 10 years, and they're still treating them as "upscale" options in cars.
Do tell o' wise one, where may I see this promised land of the MP3 market?
You make a 74 minute long playlist of DRMed AACs, WMAs, and RMs and save them to an audio CD. Then you click on "Import as MP3". Yeh, you lose a miniscule amount of audio quality that you MAY be able to hear if you listen to it, but when you're walking down the street or sitting on a train with earbuds connected to your MP3 player... you'll never notice it.
Oh, I *love* that phrase. That's so cute, and so justly descriptive. Hello, Gartner? You just got a new nickname.
Panther:
Expose, OK, though I prefer Peter Maurer's Witch.
Massively broken Finder, Jaguar's was much more practical and less "in the way".
Safari... uh, that came out in Jaguar.
Fast User Switching... a big disappointment, free UNIX has had virtual consoles for years before OS X existed... and they work better.
Preview... that came out in Jaguar as well.
I can't comment on iChat AV, Filevailt, or Inkwell... I don't use them.
Slight performance increase for some models.
Several free third-party tools that worked under Jaguar required shareware upgrades for Panther.
If it wasn't for the fact that Panther came along when I upgraded to a Mac mini I'd still be using Jaguar.
Tiger:
Spotlight... a killer new functionality that I'm truly hungry for.
20-50% improvement in GUI performance.
Just those two points alone make it worthwhile.
Witness the trials of "program not responding."...
OS X has dumped me into similar situations, the main difference is that if I leave a Terminal window open I can do
% ps waux | grep program
% kill process_id
and it goes away.
OSX will just unmount the drive when told to
Um, no, it won't. I've had to kill iTunes a couple of times to get my iPod unmounted. I also have quite a lot of hate about the way iTunes pretends it's not using my iPod Shuffle as a flash drive.
I'm hard-pressed to describe how Win 95/98/ME differ at all.
Windows 98 was a lot more solid, given its limitations. More importantly USB support in Windows 98 actually worked.
Windows ME was basically a crippled version of Windows 98 with the DOS stuff hidden so you couldn't easily go in and fix it when it fell over.
Microsoft is working on similar, if further-reaching, technology for Longhorn.
What's further-reaching about what they're putting in Longhorn? WinFS? *cough*
Is that actually going to be in there? They seem to be having problems with it.
A Microsoft Windows user is besieged. And I mean not just with spyware and worms, but also with Windows Updates. They're doing the same thing as Apple's updates (make no mistake--both companies are giving you bug fixes), but there are so many updates for this mysterious vulnerability or that compromise that a typical home user is overwhelmed by not only by the OS prompting them to the point of annoyance that you have new Windows Updates as well as the number of patches and attacks.
.NET and Direct X.
Oh yes, that same Game box I have at home. I hadn't used it since some time before Christmas, but I brought it up recently to try out NASA World Wind and before I could run it I needed to install 39 "critical security updates" and two updates to
Microsoft, having given away Windows XP Service Pack 2 for free eight months ago, still can't seem to convince enough of their users to adopt it to even hit the one-quarter mark.
Blow that for a joke, I'm sticking with Windows 2000 on my GameOS box. Why should I trust an OS like XP with a suicide switch in the kernel when the older version works fine?
Well, criticizing stuff that you're a fan of on /. gets comments like "MOD PARENT DOWN!" and "TROLL! HYPOCRITE!". I personally think Paul's analysis is really shallow and misses a lot of points, but he is at least being straight as he sees it.
You mean Microsoft's page to download SP4 for NT4.0?
No, I mean a web application that doesn't work with IE 6, or IE 5.4, or if IE 5.6 if that exists, but only IE 5.5 and only with certain obscure options set.
But I have to say your example is mighty hateful too.
I don't consider Rhapsody and "Yellow Box" missteps. What we have now in OS X, Cocoa, that's what it comes from. The problem wasn't anything to do with Rhapsody, it was with Adobe. Adobe refused to port their apps to what became Cocoa, forcing Apple to make this huge unweildy sidestep through Carbon that kept the abysmal OS 9 alive for years longer than it deserved, and where we ended up now is EXACTLY where we'd be if Apple had been able to convince Adobe and a few other key developers to stick with the program... except years later *and* most of the key apps that Apple bent over for ended up getting dropped anyway.
Spotlight is a HUGE step forward. It's killer functionality, and needs OS support to work right, but just a touch. OS X provides that support... without creating a new "search-based" file system.
Because you don't (despite what Be and Microsoft say) need a new file system to manage metadata, you just need a mechanism for applications to talk about it... WHEREVER it's stored. That's why Find on palm OS already exists, and has existed since 1996, despite Palm OS having a file system that's hardly worthy of the name. And that's why Spotlight works and Microsoft's dithering on WinFS.
And if I hadn't bought a Mac Mini I would have skipped Panther altogether because Panther *is* a relatively minor release. Tiger, with Spotlight, is a different kind of cat altogether.
I wish I had the power to ban applications like that.
I wish I still did. Unfortunately they brought us into line with the "corporate standards" and merged our Windows domains just in time for Blaster to hit. Lovely timing, that.
It improves capabilities too. There's some jobs you just can't reasonably tackle (read: they're a huge PITA) on 32-bit machines no matter how fast.
Multiple precision arithmetic libraries that make bignum (both extended and arbitrary precision) calculations trivial have existed since the '60s, and have been integrated into high performance programming languages (eg, Maclisp) since the '70s.
There is absolutely nothing I can think of for which the difference between 32- and 64- bit processing comes down to anything but a performance issue, except for large address spaces. And for GUI applications, Tiger doesn't provide a way to use large address spaces directly... and we're talking about laptops here.
I'm not into graphics stuff, but it's not difficult to imagine something like Photoshop benefitting
I'm only peripherally into graphics stuff, but I am into encryption, and it's hard for me to see anything that 64-bit integer code can do that Altivec can't do better in either area.
When Microsoft do the "right thing" (such as XP SP2),
Microsoft has yet to do the right thing. The security community has been beggng them to back out of the tight browser/desktop integration and "security zones" since 1997, and split the rendering and access functionality of the HTML control into separate components so you CAN run a locked-down sandboxed version of Internet Explorer if you want to... but instead Microsoft refuses to admit they made a mistake and patches symptom after symptom instead of attacking the disease.
That's why I, wearing my "security hat", banned all internet-capable applications that used the MS HTML control for rendering... back in 1997. As long as that ban was in effect we had zero virus and security panics, and we were the only division of our company for which that was the case.
The fundamental design of the HTML control is broken and unfixable. THe only solution is to back out of that design at a very low level, and rewrite all the applications that use it to handle access themselves. In 1997 I expected that Microsoft would do that... by now, it's obvious that they won't. They're afraid of losing face.
The right thing, from a security point of view, is to stop using Internet Explorer, Outlook, Outlook Express, Windows Media Player, Realplayer, and all other applications that use the MS HTML control to display potentially untrusted data whether they're shipped by Microsoft or some third party. Microsoft has proven over and over again for the last seven years that there is no other rational course of action.
SP2 and every other "security" patch that Microsoft provides are just smoke and mirrors.
What if I'm running Windows 2000?
Avoiding it means your systems are running on a legacy OS.
You say that like that's a bad thing.
How long before the legal or finance departments need to use a business-critical Web site that requires IE7 for access?
I don't know, you tell me: how long before some criminally stupid web developer creates a business-critical website that requires a specific version of a browser to even work? Not just "doesn't work on Firefox" (which is already in the "criminally stupid" department) but "doesn't work on recent versions of Internet Explorer"? Yes, I know, that's already happened... but in my case it was a website that didn't work on anything later than IE 5.5. Or older, either. Basically, Doctor Evil, this is a sword that cuts both ways.
I think 64-bitness is more important, and you think performance is more imporant.
I think 64-bitness is only important when it improves performance. And I've been working with real live gotta-be-fast 64-bit software since Apple was still using Nubus, and, well... the performance advantage of 64-bit is much overrated. If you really need 64-bit you know you really need it because you're running into 32-bit limitations.
The 64-bit address registers are the big win, especially for database apps. The 64-bit integer registers are nice, but Altivec already operates on multi-word objects... I'm not sure there's really a lot to gain there. Do you really know about algorithms that are faster in 64-bit mode, or are you guessing?
Just to add to the confusion, here's what APPLE says about the pipelines:
A much longer execution pipeline (up to 23 stages vs. 7 stages for the G4).
That doesn't sound like "twice as long" as the G4 or "half as long" as the Pentium 4 to *me*.
If by "Pentium-style", you mean "half as long and much wider than a Pentium",
No, I don't mean that. Unless by "Pentium-style" you mean "only Pentium 4" rather than "any Pentium-style processor", you don't mean that either.
The G3 had 4 stages, the G4 started with 4 and ended up with 7 or 9 depending on who you talk to, the G5 up to 25 for MIMD instructions... again depending on how you define "stages". Regardless, the G5 is a "brainiac" long-pipeline multiple-functional unit processor. This means it's more sensitive to pipeline stalls and cache misses, and requires much more power than a simpler design.
And, yes, I think "2 MHz sometime next year" is reasonably close to "3 MHz sometime next year we promise, honest, I know we said 2004, we meant 2005, uh, 2006" given that the G4 gets more work done per stage and thus, for typical code, will probably get more work done per clock IF it can get its bus problems resolved.
See, the G4 vs G5 question has a lot in common with the similar PIII vs P4 debate, except the G4 has had both arms tied behind its back by the crummy memory bus. If Freescale solves that problem, dual-core will just be icing on the cake.
That's why 64-bit OSes (including Tiger) can run 32-bit code.
You're being disengenuous. Tiger is primarily 32-bit, all the graphic code is still 32-bit, and any Cocoa software is 32-bit. It just has the ability to run 64-bit command-line programs, which is more than enough for most 64-bit software: for example, tthe 64-bit software I've been working on on Tru64 (which IS a native 64-bit OS) for most of the past 10 years... most of that is server code.
Getting back to the original point, for a laptop (which is what we're talking about) the ability to run 64-bit code is pointless to all but a tiny minority of the population... mostly software developers who want to test their server code when they're working on their laptop.