- Intel and AMD are shipping dual-core CPUs *today*. IBM's Antares is still in the design phase.
- PPC970 has not scaled as IBM anticipated. It is approximately 10-20% slower, clock for clock, than AMD64. AMD64 has scaled well with process improvements, and it easily reaches 2.6GHz with moderate aircooling. PPC970 is only clocked at 2.7GHz, and it needs watercooling to do so.
- PCI Express chipsets are available for both Intel and AMD CPUs. No such chipsets exist for PPC970. PCI-X and AGP cards are being quickly replaced by cheaper and faster PCIe solutions. The latest GPUs are often not available in AGP variants at launch, and the lag time is increasing. AGP GPUs often now cost more than PCIe GPUs.
- IBM's semiconductor fabrication business isn't doing well. Ripe from failures such as NV30 (fabrication at IBM was such a disaster that NV35 was fabricated by TSMC), IBM is realizing that being able to deliver world-class server chips in small quantities doesn't necessarily translate into being able to deliver low-end chips in high quantities. Numberous supply problems delayed the launch of the XServe G5 and new Apple desktops. Intel and AMD, in contrast, have extensive experience shipping in large volume.
- The PPC970 is expensive. Apple's new lower-end products need inexpensive CPUs to remain cost-competitive. Right now, that niche is filled by Freescale's G4. The G4 is a dog, though, and Apple needs to replace it. Inte's Pentium-M CPUs are considerably faster, similar in power draw, and are constantly falling in price. Intel's Celeron-M is even more affordable, and it still offers superior performance to the G4.
Yeah, except that it doesn't have to worry about pesky things like *running on batteries for a decent amout of time*.
The PPC970FX is ~50W average. Intel's Pentium-M line is closer to 20W max.
Designing a PowerBook G5 would require:
- Severely reduced battery life (e.g. the 2 hours typical of P4-M notebooks instead of the 4-5 hours typical of P-M notebooks). This would be a disaster for Apple as their product would look stupid compared to P-M based notebooks that offer both performance and battery life.
- A lower power PPC. PPC970FX is already on the latest process that IBM has. IBM simply does not have the power-saving technology that Intel does. Intel has spent years optimizing their core and their cache to save power.
- A switch to Intel CPUs. That likely means Pentium-M or Celeron-M in their small-form-factor (Mini, iMac, eMac) and notebook (iBook, PowerBook) computers, and potentially Pentium-4 in their desktop line.
"With Longhorn still in the distant future, and Windows 2000 support starting to dry up, who wants to make a crappy pit stop at XP waiting for Longhorn?"
The hundreds of millions of users who run XP on a daily basis?
XP is like Windows 2000+. It's not a huge upgrade, but there are a number of nice and notable features (ClearType, RGBA icons, firewall, Internet Connection Sharing, WIA for scanners & cameras, faster bootup, wifi support). If you turn off the theme service and disable the search dog, it's really no more annoying than Windows 2000.
If you want to keep Windows 2000, security patch support won't end until 2010.
"Here we never had any other system than GSM, and EVERYONE texts. Every major cellphone company has several prime SMS packages. Some people don't even use their phones for voice!"
Every major cell company over here has SMS packages as well. I can get unlimited SMS for $10 a month, or 300 for $3 a month (though my $20 unlimited GPRS plan already includes unlimited SMS).
"I've been to the States... Blackberry devices and pagers are more prevalent than SMSers."
Have you been to a high-school? In general, 40-year-old adults aren't too hot on SMS. Teens, college students, and young adults are.
Blackberrys are common because they are useful for road-warriors. Being able to read and send email on the go, including full integration with Exchange, is far more useful than being able to send 160-character text messeges.
"Second, you add some new features. Like stealing compression code from Stacker, MS will just steal one of the "Tabbed browsing in IE" Plugins and muck the variable names up a bit."
Come on. That's a cheap shot. There is no indication that Microsoft stole any code whatsoever for IE7.
Ever wonder why Slashdot gets the reputation of being a bunch of crazy coots? Yep, it's people like you lobbing unfounded allegations against Microsoft.
There are plenty of valid reasons to attack Microsoft. You don't need to make shit up.
OOo can't hurt Office because, as of yet, OOo is a bloated crapfest. It takes nearly 120MB of memory on my desktop (and even more on Linux), takes *forever* to start (20 seconds on my P-M 1.5GHz notebook - Word takes less than 5), crashes frequently, doesn't follow the native UI (particularly on Linux, but also on Windows), and is all-around awkward to use.
OOo 2.0 is worlds better, but on my system the spellchecking engine doesn't work - and Impress crashes when I try to show a presentation full-screen.
Look, OOo 2.0 is the "Linux" of word processors - it's useful in specific applications, offers huge advantages in terms of cost saving and lack of vendor-lockin, but just isn't "polished" enough for everyday use by normal office workers.
That, combined with the lack of a good groupware suite (Evolution is a buggy and slow piece of garbage), has prevented OOo from taking the place of Office 2003 on the desktops of my company.
"If Microsoft was really opening up Office, why didn't they go for the OASIS Spec?"
Perhaps because it didn't meet their needs? Office is a complex product, with everything from macros to embedded COM objects. Trying to shove that into the OASIS spec may not be possible.
Frankly, so long as the new formats aren't hugely patent-encumbered, this is a great boon for open-source development. Not to mention the fact that the current binary formats are horribly easy to corrupt.
Note that the Office 12 schemas will be covered under a different license. Whether or not that license precludes open-source implementation is another matter entirely.
However, I don't believe that MS would be moving to this format if they didn't want to "open things up". They certainly could have kept their evil binary formats.
Yes, 1280x1024 is too low-resolution for a 17" monitor. It also has an odd aspect ratio (1.25:1 instead of 1.33:1). I'd much rather see 1600x1200.
However, this isn't a problem with LCD technology. My friend's 15" notebook LCD is 1600x1200, and my 12" notebook LCD is 1400x1050 (which actually has the standard 1.33:1 apsect ratio).
Single-link DVI can drive 1600x1200 without issues. You can go even further with dual-link DVI.
So, I'm waiting for the 17" 1600x1200 LCDs. I don't think that I'm going to see them anytime soon, though. That's why my desktop gets used for DVDs and gaming, not hacking code. For that, I use my notebook.
I specifically got a CRT for development work. I can switch resolutions for testing, and still get high picture quality.
Flat panel displays have a "sweet-spot" resulution. Anything outside that looks terrible."
A good high-resolution LCD looks fine when scaling down. Yes, it's "blurry", but hell, how much time do you spend actually testing at non-native resolutions?
I use LCDs for coding because the text is sharper, there's no geometry to screw with, you don't have to worry about color misalignment (older CRTs often drift so that white text has color fringing), there are no ghosting issues, and you can use subpixel antialiasing.
Add the fact that you have a less reflective screen (reduced glare) that's truly flat, requires less power, and doesn't rely on persistance of vision to display its image.
CRTs were a good deal in 1999 when LCDs were expensive garbage. Today, you can get an excellent 17" screen, with DVI, for under $300.
" Darned, I was hoping they were implementing the Opera back/forward action, where the page is simply redrawn, for lack of the proper term, as the page you saw, rather than re-executed or re-downloaded. In Opera, the page redraws were so fast as to be unnoticable, and there were no data-post limitations. It was a snapshot rather than a reload. Of course, if you wanted to, you could just reload the page manually to re-invoke the post (or whatever actually happened on the page)"
That's the way that Trident (the Internet Explorer rendering engine) does things, and it's one of the big new features in 1.1. Hopefully it will be debugged enough by the time 1.1 ships to make it into the release.
Right now, Firefox is ploddingly slow on my 1.5GHz P-M system. Interestingly, it's considerably faster on Windows.
That's how you lose customers and lose eyes. No one is going to switch to Firefox because you're being an ass and refusing to write pages compatible with their browser. If you're writing decent code, you shouldn't have any trouble getting your site to work on IE. Or Firefox. Or Opera. Or WebCore. Or KHTML.
"It is as fast as DSL if your DSL service sucks, which in the U.S. it basically does."
Not true. CDMA2000 1xRTT or EDGE both deliver around 80-150kbps in the real world, with gateway latencies in the 600ms range.
Even Qwest's crappy cheap DSL is 256kbps (~200kbps actually), and even with the crappy interleaving it's only around 35ms to the default gateway (~100ms to Google). That's considerably better in both bandwidth and latency then the cellular technologies.
Not to mention that even Qwest offers 1.5mbps and 7mbps tiers, SBC has 6mbps, and Verizon offers 3mbps.
So, no, DSL is considerably better. Not quite up to cable standards (~6mbps with uncompressable data, 10ms gateway latency) in my area, but still quite good.
Get a Seasonic S12-430 instead. Quiet (20dbA/1m below 150W), cool (less than 6C difference between intake and output), efficent (over 80%), friendly to the grid (active PFC), and, perhaps most importantly, UL certified.
I have the older Super Tornado 300W, and it's excellent. All the power I need to keep my system running without a lot of noise or heat.
"virtual terminal, to the point where most Windows users hate the terminal"
Windows Server 2003 significantly improves the terminal. It's not quite BASH, but it's actually pretty usable.
"Microsoft's decision to not support "real" links, just shortcuts, in their user-mode software."
This is a UI choice. Symlinks can behave poorly if they are abused. If you want them in Windows, there's a command-line tool to do it (in the XP resource kit) or you can get a program like Junction Link Magic. It's part of the kernel, just not part of the UI.
" Never precisely specifying API behavior -- MSDN is more of a tutorial or basic user guide to the API than a true specification."
I've never had any trouble with the MSDN documentation. Linux man pages don't generally specify API at all - they specify configuration file formats and potential arguments. Perhaps that's because Linux doesn't really have "an API", it instead has a number of different APIs (GTK+/Qt, ALSA/OSS, V4L, GStreamer, X, SDL, CUPS, etc.). Manpages can't really be compared with MSDN - you have to compare the individual API guides. When you do that, you'll see that frequently they are out-of-date or simply plain wrong (as is the case with much of the GTK+ documentation).
"Bad VM algorithms. I don't know what they use, but try running low on memory on a Windows box and the system becomes bloody unusable."
As it does on Linux. Apache grinds to a halt on my older 256M webserver as soon as PHP/MySQL decide to suck up the available memory.
What *really* kills Windows systems is disk I/O. The system becomes useless once you start hacking at the disk. It's better on SCSI systems, and Windows isn't alone in this regard, but it somehow seems worse on Windows - perhaps because the shell is always hacking away at the disk, or perhaps because of the registry.
"Allowing application software to "block" a shutdown."
You can always force a shutdown with shutdown.exe. It's good that Word can block a shutdown and ask you to save - it prevents stupid mistakes in a lot of cases.
"So, for the past two weeks he's sat around with no internet and no sound and shitty video while I tore the house apart looking for the discs."
You're an idiot, then. We have this thing called the internet now. You can go on the manufacturer's website and download the drivers. You don't need to look for CDs.
Hell, I only need two drivers for my setup - the NForce Unified drivers & NVIDIA ForceWare drivers. That's it.
"Too big to be run on today's hardware, too incompatible with many critical applications, and too few business reasons to make the switch."
Longhorn is not too big to run on today's hardware (in fact, each successive alpha build has gotten faster and leaner), it's not incompatible with most applications, and there are plenty of business reasons to make the switch.
Oh, and the RTM date has already been set in May, 2006. And programmers inside Microsoft have indicated that the date isn't going to change.
"No. Coffee shops provide coffee so you can drink it. Everything else is ancillary."
No. Coffee shops are not selling coffee. Nor are restaurants selling food. They are selling *service*.
Anyone can make coffee with inexpensive equipment. Coffee shops are selling the service of preparing that coffee (or other beverage) and serving it in an inviting atmosphere.
I put down $2.50 for Java, they let me hang out with friends or other associates for a few hours and use the internet. Sounds like a fair trade to me.
How so? Even my year-old Sidekick II runs 24+ hours, regardless of how much I use it. Unless you're talking 5+ hours a day on the phone, what's the big deal? Charge it when you go to sleep and it's ready to go in the morning.
"Well, than and there was no desktop sync software"
Intellisync is now available. And you have always been able to import / edit contacts using the web interface.
A few years ago, a number of companies produced Windows CE powered "handheld PCs".
Many of the second-generation models had full-sized keyboards. They are cheap (often sub-$150), and have excellent battery life (10+ hours).
Not to mention that they boot instantly, have no moving parts, often have USB host ports, and have PCMCIA for adding Wifi or ethernet. Not to mention the color touchscreen and capability to add software like a full-featured web browser (NetFront) or word processor (TextMaker).
SPDIF outputs are usually pretty consistent at passing the PCM data or the DD/DTS sountrack if you have them configured right.
Some cards, however, such as Creative's Audigy series, are notorious for resampling inputs/outputs, so you might want to check.
Even a cheap card, like the $15 cards on Newegg, should provide a clean output. Don't buy the garbage about "jitter" that I'm sure someone will bring up - so long as your card and cabling are operating within the specification, you won't have any problems.
Do consider TOSLINK instead, however. TOSLINK uses fiber-optics, so your audio equipment and PC are electrically isolated. This reduces the chance of creating a ground loop or introducing RF noise into your reciever/amp. Moreover, it protects your equipment in the event of an electrical mishap.
The "Java is faster than C/C++" argument argues this"
No, it argues that a JIT compiler can make optimizations at runtime that could not be made at compile time. Which, indeed, is true. Certain specific algorithms are faster in Java than in C++.
Of course, we're talking about technicalities here. I'll agree that C is genarally faster than Java. But Java isn't necessarily slow - take the T-Mobile Sidekick, which has a quasi-Java OS and Java applications, yet maintains excellent performance on a 50MHz ARM CPU.
Slashdot Mirror
Let's look at the reality:
- Intel and AMD are shipping dual-core CPUs *today*. IBM's Antares is still in the design phase.
- PPC970 has not scaled as IBM anticipated. It is approximately 10-20% slower, clock for clock, than AMD64. AMD64 has scaled well with process improvements, and it easily reaches 2.6GHz with moderate aircooling. PPC970 is only clocked at 2.7GHz, and it needs watercooling to do so.
- PCI Express chipsets are available for both Intel and AMD CPUs. No such chipsets exist for PPC970. PCI-X and AGP cards are being quickly replaced by cheaper and faster PCIe solutions. The latest GPUs are often not available in AGP variants at launch, and the lag time is increasing. AGP GPUs often now cost more than PCIe GPUs.
- IBM's semiconductor fabrication business isn't doing well. Ripe from failures such as NV30 (fabrication at IBM was such a disaster that NV35 was fabricated by TSMC), IBM is realizing that being able to deliver world-class server chips in small quantities doesn't necessarily translate into being able to deliver low-end chips in high quantities. Numberous supply problems delayed the launch of the XServe G5 and new Apple desktops. Intel and AMD, in contrast, have extensive experience shipping in large volume.
- The PPC970 is expensive. Apple's new lower-end products need inexpensive CPUs to remain cost-competitive. Right now, that niche is filled by Freescale's G4. The G4 is a dog, though, and Apple needs to replace it. Inte's Pentium-M CPUs are considerably faster, similar in power draw, and are constantly falling in price. Intel's Celeron-M is even more affordable, and it still offers superior performance to the G4.
"I mean, that things almost a laptop already!"
Yeah, except that it doesn't have to worry about pesky things like *running on batteries for a decent amout of time*.
The PPC970FX is ~50W average. Intel's Pentium-M line is closer to 20W max.
Designing a PowerBook G5 would require:
- Severely reduced battery life (e.g. the 2 hours typical of P4-M notebooks instead of the 4-5 hours typical of P-M notebooks). This would be a disaster for Apple as their product would look stupid compared to P-M based notebooks that offer both performance and battery life.
- A lower power PPC. PPC970FX is already on the latest process that IBM has. IBM simply does not have the power-saving technology that Intel does. Intel has spent years optimizing their core and their cache to save power.
- A switch to Intel CPUs. That likely means Pentium-M or Celeron-M in their small-form-factor (Mini, iMac, eMac) and notebook (iBook, PowerBook) computers, and potentially Pentium-4 in their desktop line.
"With Longhorn still in the distant future, and Windows 2000 support starting to dry up, who wants to make a crappy pit stop at XP waiting for Longhorn?"
The hundreds of millions of users who run XP on a daily basis?
XP is like Windows 2000+. It's not a huge upgrade, but there are a number of nice and notable features (ClearType, RGBA icons, firewall, Internet Connection Sharing, WIA for scanners & cameras, faster bootup, wifi support). If you turn off the theme service and disable the search dog, it's really no more annoying than Windows 2000.
If you want to keep Windows 2000, security patch support won't end until 2010.
"Here we never had any other system than GSM, and EVERYONE texts. Every major cellphone company has several prime SMS packages. Some people don't even use their phones for voice!"
Every major cell company over here has SMS packages as well. I can get unlimited SMS for $10 a month, or 300 for $3 a month (though my $20 unlimited GPRS plan already includes unlimited SMS).
"I've been to the States... Blackberry devices and pagers are more prevalent than SMSers."
Have you been to a high-school? In general, 40-year-old adults aren't too hot on SMS. Teens, college students, and young adults are.
Blackberrys are common because they are useful for road-warriors. Being able to read and send email on the go, including full integration with Exchange, is far more useful than being able to send 160-character text messeges.
"Second, you add some new features. Like stealing compression code from Stacker, MS will just steal one of the "Tabbed browsing in IE" Plugins and muck the variable names up a bit."
Come on. That's a cheap shot. There is no indication that Microsoft stole any code whatsoever for IE7.
Ever wonder why Slashdot gets the reputation of being a bunch of crazy coots? Yep, it's people like you lobbing unfounded allegations against Microsoft.
There are plenty of valid reasons to attack Microsoft. You don't need to make shit up.
"It hardly has GSM!"
I'd hardly call 60 million GSM users "hardly having GSM". Not to mention that CDMA2000 also has SMS support.
Remember, Debian's servers were hacked a while back. People who live in glass houses shouldn't throw stones.
OOo can't hurt Office because, as of yet, OOo is a bloated crapfest. It takes nearly 120MB of memory on my desktop (and even more on Linux), takes *forever* to start (20 seconds on my P-M 1.5GHz notebook - Word takes less than 5), crashes frequently, doesn't follow the native UI (particularly on Linux, but also on Windows), and is all-around awkward to use.
OOo 2.0 is worlds better, but on my system the spellchecking engine doesn't work - and Impress crashes when I try to show a presentation full-screen.
Look, OOo 2.0 is the "Linux" of word processors - it's useful in specific applications, offers huge advantages in terms of cost saving and lack of vendor-lockin, but just isn't "polished" enough for everyday use by normal office workers.
That, combined with the lack of a good groupware suite (Evolution is a buggy and slow piece of garbage), has prevented OOo from taking the place of Office 2003 on the desktops of my company.
"If Microsoft was really opening up Office, why didn't they go for the OASIS Spec?"
Perhaps because it didn't meet their needs? Office is a complex product, with everything from macros to embedded COM objects. Trying to shove that into the OASIS spec may not be possible.
Frankly, so long as the new formats aren't hugely patent-encumbered, this is a great boon for open-source development. Not to mention the fact that the current binary formats are horribly easy to corrupt.
Note that the Office 12 schemas will be covered under a different license. Whether or not that license precludes open-source implementation is another matter entirely.
However, I don't believe that MS would be moving to this format if they didn't want to "open things up". They certainly could have kept their evil binary formats.
Yes, 1280x1024 is too low-resolution for a 17" monitor. It also has an odd aspect ratio (1.25:1 instead of 1.33:1). I'd much rather see 1600x1200.
However, this isn't a problem with LCD technology. My friend's 15" notebook LCD is 1600x1200, and my 12" notebook LCD is 1400x1050 (which actually has the standard 1.33:1 apsect ratio).
Single-link DVI can drive 1600x1200 without issues. You can go even further with dual-link DVI.
So, I'm waiting for the 17" 1600x1200 LCDs. I don't think that I'm going to see them anytime soon, though. That's why my desktop gets used for DVDs and gaming, not hacking code. For that, I use my notebook.
"Amen to that.
I specifically got a CRT for development work. I can switch resolutions for testing, and still get high picture quality.
Flat panel displays have a "sweet-spot" resulution. Anything outside that looks terrible."
A good high-resolution LCD looks fine when scaling down. Yes, it's "blurry", but hell, how much time do you spend actually testing at non-native resolutions?
I use LCDs for coding because the text is sharper, there's no geometry to screw with, you don't have to worry about color misalignment (older CRTs often drift so that white text has color fringing), there are no ghosting issues, and you can use subpixel antialiasing.
Add the fact that you have a less reflective screen (reduced glare) that's truly flat, requires less power, and doesn't rely on persistance of vision to display its image.
CRTs were a good deal in 1999 when LCDs were expensive garbage. Today, you can get an excellent 17" screen, with DVI, for under $300.
We're moving away from CRTs because CRTs suck.
" Darned, I was hoping they were implementing the Opera back/forward action, where the page is simply redrawn, for lack of the proper term, as the page you saw, rather than re-executed or re-downloaded. In Opera, the page redraws were so fast as to be unnoticable, and there were no data-post limitations. It was a snapshot rather than a reload. Of course, if you wanted to, you could just reload the page manually to re-invoke the post (or whatever actually happened on the page)"
That's the way that Trident (the Internet Explorer rendering engine) does things, and it's one of the big new features in 1.1. Hopefully it will be debugged enough by the time 1.1 ships to make it into the release.
Right now, Firefox is ploddingly slow on my 1.5GHz P-M system. Interestingly, it's considerably faster on Windows.
That's how you lose customers and lose eyes. No one is going to switch to Firefox because you're being an ass and refusing to write pages compatible with their browser. If you're writing decent code, you shouldn't have any trouble getting your site to work on IE. Or Firefox. Or Opera. Or WebCore. Or KHTML.
"It is as fast as DSL if your DSL service sucks, which in the U.S. it basically does."
Not true. CDMA2000 1xRTT or EDGE both deliver around 80-150kbps in the real world, with gateway latencies in the 600ms range.
Even Qwest's crappy cheap DSL is 256kbps (~200kbps actually), and even with the crappy interleaving it's only around 35ms to the default gateway (~100ms to Google). That's considerably better in both bandwidth and latency then the cellular technologies.
Not to mention that even Qwest offers 1.5mbps and 7mbps tiers, SBC has 6mbps, and Verizon offers 3mbps.
So, no, DSL is considerably better. Not quite up to cable standards (~6mbps with uncompressable data, 10ms gateway latency) in my area, but still quite good.
Get a Seasonic S12-430 instead. Quiet (20dbA/1m below 150W), cool (less than 6C difference between intake and output), efficent (over 80%), friendly to the grid (active PFC), and, perhaps most importantly, UL certified.
I have the older Super Tornado 300W, and it's excellent. All the power I need to keep my system running without a lot of noise or heat.
Seasonic knows what they're doing.
"virtual terminal, to the point where most Windows users hate the terminal"
Windows Server 2003 significantly improves the terminal. It's not quite BASH, but it's actually pretty usable.
"Microsoft's decision to not support "real" links, just shortcuts, in their user-mode software."
This is a UI choice. Symlinks can behave poorly if they are abused. If you want them in Windows, there's a command-line tool to do it (in the XP resource kit) or you can get a program like Junction Link Magic. It's part of the kernel, just not part of the UI.
" Never precisely specifying API behavior -- MSDN is more of a tutorial or basic user guide to the API than a true specification."
I've never had any trouble with the MSDN documentation. Linux man pages don't generally specify API at all - they specify configuration file formats and potential arguments. Perhaps that's because Linux doesn't really have "an API", it instead has a number of different APIs (GTK+/Qt, ALSA/OSS, V4L, GStreamer, X, SDL, CUPS, etc.). Manpages can't really be compared with MSDN - you have to compare the individual API guides. When you do that, you'll see that frequently they are out-of-date or simply plain wrong (as is the case with much of the GTK+ documentation).
"Bad VM algorithms. I don't know what they use, but try running low on memory on a Windows box and the system becomes bloody unusable."
As it does on Linux. Apache grinds to a halt on my older 256M webserver as soon as PHP/MySQL decide to suck up the available memory.
What *really* kills Windows systems is disk I/O. The system becomes useless once you start hacking at the disk. It's better on SCSI systems, and Windows isn't alone in this regard, but it somehow seems worse on Windows - perhaps because the shell is always hacking away at the disk, or perhaps because of the registry.
"Allowing application software to "block" a shutdown."
You can always force a shutdown with shutdown.exe. It's good that Word can block a shutdown and ask you to save - it prevents stupid mistakes in a lot of cases.
"So, for the past two weeks he's sat around with no internet and no sound and shitty video while I tore the house apart looking for the discs."
You're an idiot, then. We have this thing called the internet now. You can go on the manufacturer's website and download the drivers. You don't need to look for CDs.
Hell, I only need two drivers for my setup - the NForce Unified drivers & NVIDIA ForceWare drivers. That's it.
"Too big to be run on today's hardware, too incompatible with many critical applications, and too few business reasons to make the switch."
Longhorn is not too big to run on today's hardware (in fact, each successive alpha build has gotten faster and leaner), it's not incompatible with most applications, and there are plenty of business reasons to make the switch.
Oh, and the RTM date has already been set in May, 2006. And programmers inside Microsoft have indicated that the date isn't going to change.
"No. Coffee shops provide coffee so you can drink it. Everything else is ancillary."
No. Coffee shops are not selling coffee. Nor are restaurants selling food. They are selling *service*.
Anyone can make coffee with inexpensive equipment. Coffee shops are selling the service of preparing that coffee (or other beverage) and serving it in an inviting atmosphere.
I put down $2.50 for Java, they let me hang out with friends or other associates for a few hours and use the internet. Sounds like a fair trade to me.
"I never really did understand why mice won out over trackballs."
Preciscion of tracking. Try playing UT2004 with a trackball.
" Both were deal-killers for me."
How so? Even my year-old Sidekick II runs 24+ hours, regardless of how much I use it. Unless you're talking 5+ hours a day on the phone, what's the big deal? Charge it when you go to sleep and it's ready to go in the morning.
"Well, than and there was no desktop sync software"
Intellisync is now available. And you have always been able to import / edit contacts using the web interface.
A few years ago, a number of companies produced Windows CE powered "handheld PCs".
Many of the second-generation models had full-sized keyboards. They are cheap (often sub-$150), and have excellent battery life (10+ hours).
Not to mention that they boot instantly, have no moving parts, often have USB host ports, and have PCMCIA for adding Wifi or ethernet. Not to mention the color touchscreen and capability to add software like a full-featured web browser (NetFront) or word processor (TextMaker).
SPDIF outputs are usually pretty consistent at passing the PCM data or the DD/DTS sountrack if you have them configured right.
Some cards, however, such as Creative's Audigy series, are notorious for resampling inputs/outputs, so you might want to check.
Even a cheap card, like the $15 cards on Newegg, should provide a clean output. Don't buy the garbage about "jitter" that I'm sure someone will bring up - so long as your card and cabling are operating within the specification, you won't have any problems.
Do consider TOSLINK instead, however. TOSLINK uses fiber-optics, so your audio equipment and PC are electrically isolated. This reduces the chance of creating a ground loop or introducing RF noise into your reciever/amp. Moreover, it protects your equipment in the event of an electrical mishap.
The "Java is faster than C/C++" argument argues this"
No, it argues that a JIT compiler can make optimizations at runtime that could not be made at compile time. Which, indeed, is true. Certain specific algorithms are faster in Java than in C++.
Of course, we're talking about technicalities here. I'll agree that C is genarally faster than Java. But Java isn't necessarily slow - take the T-Mobile Sidekick, which has a quasi-Java OS and Java applications, yet maintains excellent performance on a 50MHz ARM CPU.