Yup, we're reaching the size limits regarding standardized, interchangable interfaces.
Thankfully, we're already tossing out electrical connections for low-speed I/O with the introduction of Bluetooth, ad there's the portential for making high-speed I/O wireless with beefy 802.11n and later revisions.
The big problem is power. Power is going to kill innovation in the wireless devices field unless we can come up with some impressive storage capacity improvements. You can improve the efficiency of the transmitter and receiver, but you can't really reduce transmission power much futher than we've already done. That would require reducing the noise floor, which is impossible.
I don't expect any of these small concept PCs to pan-out precisely for this reason: many people want flexible interfaces in their devices so they can communicate.
Where things get smaller and faster every year! Amazing!
The same technology that allows us to fit hundreds of millions of transistors on a chip allows us to build a tiny botherboard the functionality in a few custom asics and processors. And so it goes.
The only reason this is "impressive" is because Via is the first company to show it off as if it were sexy. The industry has already been producing small, PCI bus motherboards for years.
The bounty hunters chasing the critters to Earth are shape-shifters, which allows them to fit-in with the local population. When they get to Earth, one of the hunters grabs a Playboy magazine out of a trash can, and transforms into the centerfold - complete with a staple in the belly.
At the bottom of the page are two graphs that compare Microsoft's various units against Google.
What you failed to notice that that the SECOND graph shows Microsoft's profits / losses in every division (including the huge losses in the home entertainment sector).
I'm continually surprised at how low-spec so many systems are in the Steam surveys.
Why are you surpsied? Many of these people only have Steam installed to play Counterstrike 1.6, whch is STILL more popular than CS:Source.
I originally played Half-Life on a Rendition 4MB v2200 graphics card, on a machine with 64MB ram. I originally played Counterstrike on a 16MB Matrox G400. Is it any wonder you can play CS 1.6 today on any old Intel GMA 900-series with geat framerates?
Also, you'd be suprised how smoothly CS:Source runs on a FX 5200 if you run it in DX7 or DX8 modes. Most serious CS:Source players play with all the details turned off anyway.
The one that's really stumped me over the years is how ATI lost their spot on-top in the Steam survey. The first Steam survey, fueled by the impressive price / performance ratio of the 9600 series and the Half-Life 2 voucher program, put ATI on-top. It took two years of ATI continually ignoring the midrange, but they finally lost that spot to Nvidia's 6600 series. ATI continued to ignore the midrange, allowing the 7600 to become entrenched, and only now are they producing enticing midrange products.
It's really amazing how accurate a picture this survey is of the PC 3D gaming industry.
One question I have: if MMC has the same 2GB limit as SD does (fixed by SDHC), what benefit is this backward-compatibility with older SD/MMC readers?
We are at a crossroads regarding capacity thanks to the popularity of SD and the 2GB spec limit. Why don't they just push the standard as USB-only? Most of the new cards sold under this standard will be too big for most MMC/SD readers to read anyway.
The launch time is a test the modern computer is guaranteed to lose. This is because the peak read rate of hard disks has only improved by 100x in the last 20 years.
Let's start with the admission that the modern OS is 15 thousand times larger (1MB versus 15GB). It's a fair assumption that most applications are at least larger by a small fraction of that - say one thousand times larger.
That old SCSI hard disk would have a peak read speed of around 1MB/s, while the best disks around today are approaching 100MB/s. That's a 100x increase in read delay speed, ignoring access time improvements.
Even if we're generous and say the application code being loaded is a slim one thousand times larger than the version on the Mac Plus, with the pitiful 100x increase in hard disk read speed, it's still going to load 10x slower.
So, why are they so close? Why is the modern system with the relatively slow hard disk so competitive?
* Fast caching architecture with prefetch (on the hard disk controller itself). * Shared libraries are already memory-resident. * OS attempts to preload your most-used applications into memory (new feature in Vista).
All these features are designed to get around the limitations imposed by the hard disk. Once we finally find a better solid-state solution, the modern system could potentialy blow the Mac Plus out of the water.
I guess you don't enjoy intentionally cheesy films like The Mask, Tank Girl, Ghost Rider, Howard The Duck, TMNT, etc.
The comics WERE cheesy. Cheesy premise, cheesy background story, cheesy dialog, cheesy villians - these are all necessary components of a graphic novel that appeals to kids and teens.
The films are also cheesy, and I love it. A straight-faced comic-book hero film is a snorefest.
1. Form First Foundation. 2. Create new "technology" religion. 3. Watch the old Galactic Empire crumble from within. 4. Get taken over by The Mule. 5. Find the Second Foundation. 6. ?? 7. Write Foundation's Edge due to publisher pressure, and profit!!
DX10 or for the uninformed, Derendering eXtraction (10 megapixels/second) is a standard benchmark for measuring the performance of GPUs or Gradient Pixilization Units. Pretty much this is what the video card companies all base their prices on with price being directly related to how many pixels can be gradiated per unit (usually about 30 cents per pixel/ounce).
But do those nVidia and ATI cards have open source drivers? The Intel chips do!
Do Intel's X3000 open-source drivers have support for the T&L units and vertex shaders built-in to the X3000? The Windows drivers certainly don't, even after 9 months on the market!
Almost 3 months ago, beta drivers were promised, but they have yet to surface. The X3000 is still using the processor to perform vertex shading / T&L, just like the GMA 900 / 950, and that's why it still gets beat by the old Nvidia GeForce 6150.
I played the game for about an hour and a half and was utterly bored. Plus, I hadn't even left the village yet. Long adventure games should have something to grab you in the beginning, and this had nothing but corralling sheep and getting a bird to fly towards you.
I agree with you, the opening part of the game (up until the twilight appears) is pretty boring. There's no anticipation, and some of the puzzle elements are unclear.
For example, I wasted half an hour figuring out how to get the cat to go home because the solution was not obvious. SPOILER: you have to fish twice in a row next to the cat. I got fed-up after fishing by the cat because I couldn't use the fish for anything - it flopped around and disappeared. So, I wandered around before trying to fish again. If you do anything else in-between your fishing attempts, they don't count, and the puzzle remains unsolved. I had to look it up in a FAQ it was so annoing.
That said, I've learned not to judge games based on the first couple hours, especially Zelda games. The rest of the game is well-paced and worth the time.
I'm sure he's referring to the poor latency. Multiple seconds for a webserver to respond is much worse than ANY dialup service I've ever used* (typically 200-300 ms latency). Sure, if all you do is download large files, then bandwidth is king...however, if you need something interactive, latency is usually the biggest factor.
You'd be surprised what you can do on dialup speeds. One of my previous employers used to run a Citrix Winframe server connected to a pool of 28.8k modems, and the interface was surpringly fast so long as you were just using Word/Excel or mail apps. Latency, again, was the most important attribute.
*Note that I've used some pretty shitty dialup, like the service at my Aunt's place out in Bumblefuck, Iowa. On a GOOD day she got a 28.8k connection to the outside world with her 56k modem (thanks to shitty old telco lines)! The latency was still well under a second.
I think this number might be a bit overestimated. There are still lots of notebooks sold with core duo (or pre-merom celeron M), though yes on the desktop you'd have trouble finding a pc without x86-64.
Think so, eh? Yeah, I pulled those numbers out of my ass, based on the general trend to replace Core Duo models with Core 2 Duo models. I figured the only places still selling Core Duos and Pentium Ms were selling closeout models (with the exception of embedded shops). Unfortunately, there's no easy way to get sales numbers for a single procesor model.
Intel essentially threw a wrench in the 64-bit machine when they announced that the Core Duo was 32-bit. I think Microsoft would have a lot more solid ground to stand on regarding a 64-bit-only Vista if Intel hadn't dragged their feet so long on the Pentium M, and then released that stopgap.
In 1995 the Pentium Pro introduced 36-bit addressing and combined with windows PAE at the time, could address 64 gibibytes (yes, "gibibytes") of memory.
Do you make a habit out of reading the first sentence of a post and then commenting on it with a smart response?
The second sentence of my post, for your reference:
Most motherboards don't support PAE (either due to lack of re-mappable PCI address space, or even lack of 36-bit address lines!), so we have a hard-limit of 2-3GB in the most popular version of Vista (32-bit).
PAE is not some magic bullet. Support is required in three places: OS, chipset and motherboard itself.
Vista supports PAE just fine, so that's not a problem. The problem lies in the hardware:
While the limitations of 32-bit Vista may spark a renewed interest in motherboards having proper PAE support, it's too little, too late. I also consider PAE a pretty crappy band-aid for a problem that already has a solution (Vista 64). This was the focus of my original post: we could have avoided this whole memory mess if Vista were purely flat 64-bit.
That's why they have WOW64 mode on their 64-bit operating systems. Tghe x86-64 platform was designed with full 32-bit compatibility in mind. This is why I was amazed MS didn't make Vista a core 64-bit OS - like the 386 (running 16-bit applications in 32-bit mode), it has support for running 32-bit applications in 64-bit mode.
This provides excellent backward-compatibility for 32-bit apps, with all the benefits of a 64-bit OS and drivers. This allows you to unify development of applications and drivers to one platform, instead of splittiing it between Vista 32 and 64.
Right now Vista 64 is the black sheep becausde all focus is on the 32-bit version.
The 4GB memory barrier is fast-aproaching for high-end users, and dealing with it is a MESS. Most motherboards don't support PAE (either due to lack of re-mappable PCI address space, or even lack of 36-bit address lines!), so we have a hard-limit of 2-3GB in the most popular version of Vista (32-bit). This is going to be a rough few years for game developers.
I really don't see why Microsoft went 32-bit on this version anyway...I'd say over %80 of the potential upgrade platforms and over %95 of all shipping PCs today support x86-64 mode. But when you look back, history paves the way:
Windows 386 = Windows 2.0 with 32-bit enhancements bolted-on. Equivilant of Windows XP 64 Windows 3 = crossover version with support for 16-bit and 32-bit processors. Equivilant of Vista. Windows 95 = supports only 32-bit processors. Equivilant to the next revision of Windows.
Too bad Microsoft didn't have the balls to jump the gun and make Vista 64-bit only.
The article states that %78 precent of sales are from the "premium edition," but we all know with such a segmented product lineup (5 versions sold in the US) there's no way JUST Home Premium made up %78 of total sales. This is an extremely poorly worded article.
I'm more inclined to interpret this as: %22 of all sales were Vista Home Basic, and the rest were Vista Home Premium or higher. This would also include Vista Business, Vista Ultimate, and yes, Vista Enterprise, which adds features on-top of the $299 Vista Business Edition.
Yeah, it wouldn't surprise me at all. The Software Assurance is issued in 3-year intervals, and since it was introduced in 2001, many companies have re-signed Software Assurance plans in 2004 and recently in 2006/2007. In fact, this was one of the driving factors behind The November '06 release of Vista: Microsoft quelled unrest in their biggest customers by releasing Vista before their Software Assurance plans ran out.
The fact is, very few companies are going to drop their Software Assurance contracts, and since every single SA contact comes with a Vista Enterpeise license (it is the only way to get one, according to Microsoft), I wouldn't be surprised at all if Microsoft counted every re-signed SA contract in their numbers.
I've been wondering this whole time where MS was getting their numbers, but now it all makes sense. What also makes sense is how Microsoft posted huge profits for this quarter: lots of re-signed SA contracts, plus SOME genuine sales of Vista (with new accounting practices, the entire profit from the sale is applied to only this quarter, instead of the software lifetime).
Exactly. When I was first introduced to BASIC in my 8th grade programming class, the teacher never even covered GOTO. She mentioned that the keyword existed, but that we should use GOSUB. Unfortunately, we never got an explanation for why this was so.
I didn't actually discover why GOTO was such a bad thing until I started programming games on my TI-82 in high school. In TI-BASIC, when using GOTO to escape a loop, the loop never completed, and the loop variable remained allocated. Over time, this would allocate too many loop variables, and cause a program to slow and crash (used to be called "program fatgue" until the real cause was found). I ran into this problem when I built my most ambitious game, a graphical Scorched Earth clone.
The game "worked," but crashed after a few minutes of play...I was clueless as to the cause. I hadn't learned proper programming concepts, and with the TI-82's lack of real function calls, it certainly wasn't helping my programming style. After a suggestion from someone with more programming experience, I dropped the GOTOs and redesigned the entire program to run with if trees and loops. No more memory leak.
I don't think it's the language that's to blame, I think it's a lack of quality documentation, and a severe lack of clued-in indivuduals willing to offer advice. I remember reading the TI-BASIC section of my TI-82 manual thoroughly, and nowhere did they mention a simple programming tip to always let your loops complete and exit naturally...even though the temptation of using all 37 possible labels was so great.
If I hadn't been able to poke the brains of a more talented programmer, I might never have learrned that lesson. One thing is for certain: there's no way ANY programming language is going to just magically teach a lesson like that, at least in a way you will REMEMEBER and LIVE by. You have to make mistakes to really LEARN. That's why I like BASIC: it encourages people to try and make many mistakes...but like any language, it does require some oversight and feedback to bring people up to the next level.
I was pretty happy when I picked up my 90nm $300 7900 GT last year - same power consumption as my old 6600 GT, but three times the performance! If I HAD to buy today, I'd get the 8800 GTS, but because Nvidia didn't design it with different 2D and 3D clocks, the 2D idle consumption is higher than it should be. The x2900 XT has a "low" idle power consumption in respect to the 8800-series because it supports a lower 2D clock.
Hopefully, this will be corrected in the 65nm refresh. Even if it isn't, the overall power consumption should drop significantly.
Slashdot Mirror
Yup, we're reaching the size limits regarding standardized, interchangable interfaces.
Thankfully, we're already tossing out electrical connections for low-speed I/O with the introduction of Bluetooth, ad there's the portential for making high-speed I/O wireless with beefy 802.11n and later revisions.
The big problem is power. Power is going to kill innovation in the wireless devices field unless we can come up with some impressive storage capacity improvements. You can improve the efficiency of the transmitter and receiver, but you can't really reduce transmission power much futher than we've already done. That would require reducing the noise floor, which is impossible.
I don't expect any of these small concept PCs to pan-out precisely for this reason: many people want flexible interfaces in their devices so they can communicate.
Where things get smaller and faster every year! Amazing!
The same technology that allows us to fit hundreds of millions of transistors on a chip allows us to build a tiny botherboard the functionality in a few custom asics and processors. And so it goes.
The only reason this is "impressive" is because Via is the first company to show it off as if it were sexy. The industry has already been producing small, PCI bus motherboards for years.
*Flash*
8=================D
There' you're scarred for life! Now I'm facing forty years!
This reminds me of Critters.
The bounty hunters chasing the critters to Earth are shape-shifters, which allows them to fit-in with the local population. When they get to Earth, one of the hunters grabs a Playboy magazine out of a trash can, and transforms into the centerfold - complete with a staple in the belly.
At the bottom of the page are two graphs that compare Microsoft's various units against Google.
What you failed to notice that that the SECOND graph shows Microsoft's profits / losses in every division (including the huge losses in the home entertainment sector).
I'm continually surprised at how low-spec so many systems are in the Steam surveys.
Why are you surpsied? Many of these people only have Steam installed to play Counterstrike 1.6, whch is STILL more popular than CS:Source.
I originally played Half-Life on a Rendition 4MB v2200 graphics card, on a machine with 64MB ram. I originally played Counterstrike on a 16MB Matrox G400. Is it any wonder you can play CS 1.6 today on any old Intel GMA 900-series with geat framerates?
Also, you'd be suprised how smoothly CS:Source runs on a FX 5200 if you run it in DX7 or DX8 modes. Most serious CS:Source players play with all the details turned off anyway.
The one that's really stumped me over the years is how ATI lost their spot on-top in the Steam survey. The first Steam survey, fueled by the impressive price / performance ratio of the 9600 series and the Half-Life 2 voucher program, put ATI on-top. It took two years of ATI continually ignoring the midrange, but they finally lost that spot to Nvidia's 6600 series. ATI continued to ignore the midrange, allowing the 7600 to become entrenched, and only now are they producing enticing midrange products.
It's really amazing how accurate a picture this survey is of the PC 3D gaming industry.
One question I have: if MMC has the same 2GB limit as SD does (fixed by SDHC), what benefit is this backward-compatibility with older SD/MMC readers?
We are at a crossroads regarding capacity thanks to the popularity of SD and the 2GB spec limit. Why don't they just push the standard as USB-only? Most of the new cards sold under this standard will be too big for most MMC/SD readers to read anyway.
Those cores aren't real.
Seriously, that "pic" is a render.
The launch time is a test the modern computer is guaranteed to lose. This is because the peak read rate of hard disks has only improved by 100x in the last 20 years.
Let's start with the admission that the modern OS is 15 thousand times larger (1MB versus 15GB). It's a fair assumption that most applications are at least larger by a small fraction of that - say one thousand times larger.
That old SCSI hard disk would have a peak read speed of around 1MB/s, while the best disks around today are approaching 100MB/s. That's a 100x increase in read delay speed, ignoring access time improvements.
Even if we're generous and say the application code being loaded is a slim one thousand times larger than the version on the Mac Plus, with the pitiful 100x increase in hard disk read speed, it's still going to load 10x slower.
So, why are they so close? Why is the modern system with the relatively slow hard disk so competitive?
* Fast caching architecture with prefetch (on the hard disk controller itself).
* Shared libraries are already memory-resident.
* OS attempts to preload your most-used applications into memory (new feature in Vista).
All these features are designed to get around the limitations imposed by the hard disk. Once we finally find a better solid-state solution, the modern system could potentialy blow the Mac Plus out of the water.
Huh?
I guess you don't enjoy intentionally cheesy films like The Mask, Tank Girl, Ghost Rider, Howard The Duck, TMNT, etc.
The comics WERE cheesy. Cheesy premise, cheesy background story, cheesy dialog, cheesy villians - these are all necessary components of a graphic novel that appeals to kids and teens.
The films are also cheesy, and I love it. A straight-faced comic-book hero film is a snorefest.
1. Form First Foundation.
2. Create new "technology" religion.
3. Watch the old Galactic Empire crumble from within.
4. Get taken over by The Mule.
5. Find the Second Foundation.
6. ??
7. Write Foundation's Edge due to publisher pressure, and profit!!
DX10 or for the uninformed, Derendering eXtraction (10 megapixels/second) is a standard benchmark for measuring the performance of GPUs or Gradient Pixilization Units. Pretty much this is what the video card companies all base their prices on with price being directly related to how many pixels can be gradiated per unit (usually about 30 cents per pixel/ounce).
Hey, how much is that in furlongs/fortnight?
But do those nVidia and ATI cards have open source drivers? The Intel chips do!
Do Intel's X3000 open-source drivers have support for the T&L units and vertex shaders built-in to the X3000? The Windows drivers certainly don't, even after 9 months on the market!
Almost 3 months ago, beta drivers were promised, but they have yet to surface. The X3000 is still using the processor to perform vertex shading / T&L, just like the GMA 900 / 950, and that's why it still gets beat by the old Nvidia GeForce 6150.
Yup, pretty worthless if you ask me. Might be more cost-efficient over the lifetime of the line to lay 10 regular conductors than one superconductor.
And since when did "high temperature" refer to -230C?
Oh, I guess I was thinking of Room Temperature Superconductors, which are still a pipe-dream.
I played the game for about an hour and a half and was utterly bored. Plus, I hadn't even left the village yet. Long adventure games should have something to grab you in the beginning, and this had nothing but corralling sheep and getting a bird to fly towards you.
I agree with you, the opening part of the game (up until the twilight appears) is pretty boring. There's no anticipation, and some of the puzzle elements are unclear.
For example, I wasted half an hour figuring out how to get the cat to go home because the solution was not obvious. SPOILER: you have to fish twice in a row next to the cat. I got fed-up after fishing by the cat because I couldn't use the fish for anything - it flopped around and disappeared. So, I wandered around before trying to fish again. If you do anything else in-between your fishing attempts, they don't count, and the puzzle remains unsolved. I had to look it up in a FAQ it was so annoing.
That said, I've learned not to judge games based on the first couple hours, especially Zelda games. The rest of the game is well-paced and worth the time.
I'm sure he's referring to the poor latency. Multiple seconds for a webserver to respond is much worse than ANY dialup service I've ever used* (typically 200-300 ms latency). Sure, if all you do is download large files, then bandwidth is king...however, if you need something interactive, latency is usually the biggest factor.
You'd be surprised what you can do on dialup speeds. One of my previous employers used to run a Citrix Winframe server connected to a pool of 28.8k modems, and the interface was surpringly fast so long as you were just using Word/Excel or mail apps. Latency, again, was the most important attribute.
*Note that I've used some pretty shitty dialup, like the service at my Aunt's place out in Bumblefuck, Iowa. On a GOOD day she got a 28.8k connection to the outside world with her 56k modem (thanks to shitty old telco lines)! The latency was still well under a second.
I think this number might be a bit overestimated. There are still lots of notebooks sold with core duo (or pre-merom celeron M), though yes on the desktop you'd have trouble finding a pc without x86-64.
Think so, eh? Yeah, I pulled those numbers out of my ass, based on the general trend to replace Core Duo models with Core 2 Duo models. I figured the only places still selling Core Duos and Pentium Ms were selling closeout models (with the exception of embedded shops). Unfortunately, there's no easy way to get sales numbers for a single procesor model.
Intel essentially threw a wrench in the 64-bit machine when they announced that the Core Duo was 32-bit. I think Microsoft would have a lot more solid ground to stand on regarding a 64-bit-only Vista if Intel hadn't dragged their feet so long on the Pentium M, and then released that stopgap.
In 1995 the Pentium Pro introduced 36-bit addressing and combined with windows PAE at the time, could address 64 gibibytes (yes, "gibibytes") of memory.
Do you make a habit out of reading the first sentence of a post and then commenting on it with a smart response?
The second sentence of my post, for your reference:
Most motherboards don't support PAE (either due to lack of re-mappable PCI address space, or even lack of 36-bit address lines!), so we have a hard-limit of 2-3GB in the most popular version of Vista (32-bit).
PAE is not some magic bullet. Support is required in three places: OS, chipset and motherboard itself.
Vista supports PAE just fine, so that's not a problem. The problem lies in the hardware:
Most consumer motherboard chipsets do not support the remapping of memory-mapped I/O space required to support PAE. Even then, some motherboards omit the extra 4 bits of address lines as a cost savings feature. You usually have to trade-up to server boards to get these types of features, and even some of those fall short.
While the limitations of 32-bit Vista may spark a renewed interest in motherboards having proper PAE support, it's too little, too late. I also consider PAE a pretty crappy band-aid for a problem that already has a solution (Vista 64). This was the focus of my original post: we could have avoided this whole memory mess if Vista were purely flat 64-bit.
That's why they have WOW64 mode on their 64-bit operating systems. Tghe x86-64 platform was designed with full 32-bit compatibility in mind. This is why I was amazed MS didn't make Vista a core 64-bit OS - like the 386 (running 16-bit applications in 32-bit mode), it has support for running 32-bit applications in 64-bit mode.
This provides excellent backward-compatibility for 32-bit apps, with all the benefits of a 64-bit OS and drivers. This allows you to unify development of applications and drivers to one platform, instead of splittiing it between Vista 32 and 64.
Right now Vista 64 is the black sheep becausde all focus is on the 32-bit version.
Windows 95 contained 16-bit code, but the OS required a 32-bit processor.
This is as-opposed to Windows 3.x, which can run in real mode on a 16-bit processor, or in protected mode on a 32-bit processor.
The 4GB memory barrier is fast-aproaching for high-end users, and dealing with it is a MESS. Most motherboards don't support PAE (either due to lack of re-mappable PCI address space, or even lack of 36-bit address lines!), so we have a hard-limit of 2-3GB in the most popular version of Vista (32-bit). This is going to be a rough few years for game developers.
I really don't see why Microsoft went 32-bit on this version anyway...I'd say over %80 of the potential upgrade platforms and over %95 of all shipping PCs today support x86-64 mode. But when you look back, history paves the way:
Windows 386 = Windows 2.0 with 32-bit enhancements bolted-on. Equivilant of Windows XP 64
Windows 3 = crossover version with support for 16-bit and 32-bit processors. Equivilant of Vista.
Windows 95 = supports only 32-bit processors. Equivilant to the next revision of Windows.
Too bad Microsoft didn't have the balls to jump the gun and make Vista 64-bit only.
The article states that %78 precent of sales are from the "premium edition," but we all know with such a segmented product lineup (5 versions sold in the US) there's no way JUST Home Premium made up %78 of total sales. This is an extremely poorly worded article.
I'm more inclined to interpret this as: %22 of all sales were Vista Home Basic, and the rest were Vista Home Premium or higher. This would also include Vista Business, Vista Ultimate, and yes, Vista Enterprise, which adds features on-top of the $299 Vista Business Edition.
Yeah, it wouldn't surprise me at all. The Software Assurance is issued in 3-year intervals, and since it was introduced in 2001, many companies have re-signed Software Assurance plans in 2004 and recently in 2006/2007. In fact, this was one of the driving factors behind The November '06 release of Vista: Microsoft quelled unrest in their biggest customers by releasing Vista before their Software Assurance plans ran out.
The fact is, very few companies are going to drop their Software Assurance contracts, and since every single SA contact comes with a Vista Enterpeise license (it is the only way to get one, according to Microsoft), I wouldn't be surprised at all if Microsoft counted every re-signed SA contract in their numbers.
I've been wondering this whole time where MS was getting their numbers, but now it all makes sense. What also makes sense is how Microsoft posted huge profits for this quarter: lots of re-signed SA contracts, plus SOME genuine sales of Vista (with new accounting practices, the entire profit from the sale is applied to only this quarter, instead of the software lifetime).
Exactly. When I was first introduced to BASIC in my 8th grade programming class, the teacher never even covered GOTO. She mentioned that the keyword existed, but that we should use GOSUB. Unfortunately, we never got an explanation for why this was so.
I didn't actually discover why GOTO was such a bad thing until I started programming games on my TI-82 in high school. In TI-BASIC, when using GOTO to escape a loop, the loop never completed, and the loop variable remained allocated. Over time, this would allocate too many loop variables, and cause a program to slow and crash (used to be called "program fatgue" until the real cause was found). I ran into this problem when I built my most ambitious game, a graphical Scorched Earth clone.
The game "worked," but crashed after a few minutes of play...I was clueless as to the cause. I hadn't learned proper programming concepts, and with the TI-82's lack of real function calls, it certainly wasn't helping my programming style. After a suggestion from someone with more programming experience, I dropped the GOTOs and redesigned the entire program to run with if trees and loops. No more memory leak.
I don't think it's the language that's to blame, I think it's a lack of quality documentation, and a severe lack of clued-in indivuduals willing to offer advice. I remember reading the TI-BASIC section of my TI-82 manual thoroughly, and nowhere did they mention a simple programming tip to always let your loops complete and exit naturally...even though the temptation of using all 37 possible labels was so great.
If I hadn't been able to poke the brains of a more talented programmer, I might never have learrned that lesson. One thing is for certain: there's no way ANY programming language is going to just magically teach a lesson like that, at least in a way you will REMEMEBER and LIVE by. You have to make mistakes to really LEARN. That's why I like BASIC: it encourages people to try and make many mistakes...but like any language, it does require some oversight and feedback to bring people up to the next level.
I'm waiting for the 65nm refresh, myself.
I was pretty happy when I picked up my 90nm $300 7900 GT last year - same power consumption as my old 6600 GT, but three times the performance! If I HAD to buy today, I'd get the 8800 GTS, but because Nvidia didn't design it with different 2D and 3D clocks, the 2D idle consumption is higher than it should be. The x2900 XT has a "low" idle power consumption in respect to the 8800-series because it supports a lower 2D clock.
Hopefully, this will be corrected in the 65nm refresh. Even if it isn't, the overall power consumption should drop significantly.