Seems to me there's a simple and major error in this document. Figure 2 shows storage latency in nanoseconds for a spinning disk... I don't know what kind of spinning disks they're using... maybe alien technology?
I don't think the marketing that MS has put out so far makes it clear enough that there are two different Windows 8s. The Windows 8 RT on the tablet abandons one of the primary historical Windows selling points, that is, legacy software compatibility.
In addition, from what I can tell so far, the current Surface with RT (as opposed to the upcoming Surface Pro) abandons another of MS traditional selling points - business integration: It doesn't integrate with active directory or group policy, and it doesn't have Outlook. In fact it implements a bunch of consumer-centric stock apps (Music, News, etc.) that I think are going to be the opposite of a selling point for the business market.
You only have to look at Vista and 7's Gadgets to see how well Microsoft has previously done pushing developers to a new app GUI. That said, I'm sure that if they stick with it this Windows 8 Applications concept, maybe they'll have a legitimate application ecosystem by the time Windows 9 rolls around. That said, a promise of perserverance and future legitimacy doesn't make me want to go out and buy a Surface right now.
I also understand that MS is planning to remove one of Surface's last remaining legitimate selling points by offering Office on iOS and Android.
The seeming incompetence of this new plan astounds me. Oh and how about that Windows Server 2012 with the new 'Metro' start interface? Why?? Did someone at MS think that system admins wanted to connect touch screens to their servers?
While I agree with your main statement: (Larger, cheaper SSDs would be nice even if we had to live with them being a bit slower), your estimates of hard disk and SSD speeds are pretty far off.
Yes, the SATA 6Gb/s standard supports up to about maybe 600MB/s in practical operation. No, mechanical hard disks do not approach 400MB/s (let alone >400MB/s). The fastest mechanical hard disks on the market right now might get you 170MB/s maximum in a sequential benchmark (read: not a real-world scenario), and only when reading from the fastest part of the disk. http://hothardware.com/Reviews/WD-VelociRaptor-600GB-Fastest-HD-Ever/?page=6
The fastest SATA SSDs are pushing maybe 300MB/s in these benchmarks, and usually only for read speeds. The reason people are seeing such disproportionate performance improvements with SSDs is access latency. The fastest hard disk might have 6.0ms average latency. SSDs are usually below 0.1ms.
Which OS is still able to do run 15-20 years unmodified on constantly-evolving hardware?
DOS (including versions from 1990 or even earlier) still works unmodified on the latest hardware. Might only format 2GB partitions or somesuch but it works and AFAIK it would be proportionately faster on new hardware.
I suspect an old Linux kernel would still run on new hardware too. As long as you are still satisfied with the functions those operating systems support they could still be a viable alternative.
Guess what. Linux developers actually WRITE most of the drivers for Linux THEMSELVES with very little help from hardware manufacturers, for FREE! In contrast, it seems like MS couldn't even be bothered to make an effort to work with hardware manufacturers in order to get stable drivers out in a reasonable time for Vista users.
I don't think this is a fair comparison. Switching from one version of a given distribution of Linux to the next is much different than going from XP to Vista. Your settings are preserved, your GUI is close to identical unless you've decided to move to a new major release or a different GUI altogether. Basically there is a whole lot more choice involved in exactly what parts of the OS you are upgrading and what kind of learning curve you want to deal with
Generally new changes aren't forced down your throat like with MS Windows. Things aren't moved around in the GUI for no reason, application compatibility usually only gets better, and worse comes to worse you can get out your wallet and bribe a package maintainer to repackage whatever application you need for the latest version of your distro. Personally I would much rather talk to a Linux developer than with Microsoft Support.
The average buyer hasn't paid for a Microsoft upgrade ever. He gets the new OS when he buys his new PC. How is that 'pushy upgrades'
The average Microsoft user BUYS a new OS when he buys his new PC. There, fixed that for you.
It is a pushy upgrade if the customer doesn't want the new OS and would be happy to buy the new PC at a discount with no OS, get rid of his old PC, and continue to use his current XP license on his new PC. It is a pushy upgrade if the customer is being forced to a new GUI and new program incompatibilities because of lack of bugfix support by the OS manufacturer.
I concur. Your post is both funny and insightful. One time I complained to faculty about the reading material in an operating system class (at a university which I will not mention here), and was basically told that the reason the material was factually incorrect was because things change so often in the CS world. I argued that there were plenty of books written decades ago that were still completely relevant. I was told that those books were few and far between - which is true, but irrelevant AFAICT. I dropped this line of questioning, got an A in the class, and remained frustrated about the ineffectiveness (and in some cases outright falseness) of our assigned reading material.
Mod parent up. Microsoft licensing terms are constantly becoming more confusing and more onerous.
Sometimes I have a hard time understanding the license terms myself and Microsoft doesn't make it easy to get in touch with someone who knows.
I work for a system builder and, as the parent said, Microsoft is no longer letting us distribute installation CDs with Microsoft Office 2007 which is just a huge headache all around.
Here is a link to a site describing the scientific method. Can you test the creation of a universe?? Even if you could, can you say for sure that just because you created it one way in your tests, that it wasn't created a different way the first time around? Evolution or creation is an event, an event cannot be scientifically proven to have happened. Maybe you could prove that it's possible, but beyond that is not the place for theories, only educated guessing.
I don't know how this got a 4 Insightful... Evolution cannot be proved or disproven by the scientific method either. We have no way of knowing what happened in order for the universe as we know it to come into existance, we weren't there. We can make educated guesses based on scientific evidence, but in my opinion, evolution is still a hyphothesis, and should never have gained theory status. Also, Creationism in itself does not contain any kind of anti-evolution agenda, its just different way of explaining how the universe came into existence, and in my opinion, is actually a more complete theory, as it takes into account how not just how the universe was formed, but how the underlying laws which govern those systems came to be.
Evolution is too broad and is compatible with any state of the world (through scientific guesses based on circumstancial evidence, which is what evolution amounts to at this point). Evolution is not a scientific theory, because it cannot be verified through use of the scientific method.
I haven't read anything very interesting on any of the IT review sites lately. This ExtremeTech nonsense is just a continuation of that trend. I started to skim the article, saw the part that said they had implemented keyboard shortcuts for the gmail interface. Thats way cool! Now how did they do it? Theres no mention of what type of coding gMail is built with, is it just fancy CSS? is it Javascript? Is the mail client actually a Java app?! WTH is the MEAT?! I want to know how it runs! The comparison chart at the end is totally lame, nothing important or non-obvious was compared there. Also I was unimpressed with the blatant dissing of whoever the government official was who talked about gMail bringing billboards to your desktop. Obviously the advertising area in gMail is not a big deal, but you can say that, and leave it alone. This bashing of the politician did NOT enhance the article at all.
Slashdot Mirror
Seems to me there's a simple and major error in this document. Figure 2 shows storage latency in nanoseconds for a spinning disk... I don't know what kind of spinning disks they're using... maybe alien technology?
The site you linked requires registration to actually see the posted benchmark score. Can you give a summary?
I don't think the marketing that MS has put out so far makes it clear enough that there are two different Windows 8s. The Windows 8 RT on the tablet abandons one of the primary historical Windows selling points, that is, legacy software compatibility.
In addition, from what I can tell so far, the current Surface with RT (as opposed to the upcoming Surface Pro) abandons another of MS traditional selling points - business integration: It doesn't integrate with active directory or group policy, and it doesn't have Outlook. In fact it implements a bunch of consumer-centric stock apps (Music, News, etc.) that I think are going to be the opposite of a selling point for the business market.
You only have to look at Vista and 7's Gadgets to see how well Microsoft has previously done pushing developers to a new app GUI. That said, I'm sure that if they stick with it this Windows 8 Applications concept, maybe they'll have a legitimate application ecosystem by the time Windows 9 rolls around. That said, a promise of perserverance and future legitimacy doesn't make me want to go out and buy a Surface right now.
I also understand that MS is planning to remove one of Surface's last remaining legitimate selling points by offering Office on iOS and Android.
The seeming incompetence of this new plan astounds me. Oh and how about that Windows Server 2012 with the new 'Metro' start interface? Why?? Did someone at MS think that system admins wanted to connect touch screens to their servers?
While I agree with your main statement: (Larger, cheaper SSDs would be nice even if we had to live with them being a bit slower), your estimates of hard disk and SSD speeds are pretty far off.
Yes, the SATA 6Gb/s standard supports up to about maybe 600MB/s in practical operation. No, mechanical hard disks do not approach 400MB/s (let alone >400MB/s). The fastest mechanical hard disks on the market right now might get you 170MB/s maximum in a sequential benchmark (read: not a real-world scenario), and only when reading from the fastest part of the disk.
http://hothardware.com/Reviews/WD-VelociRaptor-600GB-Fastest-HD-Ever/?page=6
The fastest SATA SSDs are pushing maybe 300MB/s in these benchmarks, and usually only for read speeds. The reason people are seeing such disproportionate performance improvements with SSDs is access latency. The fastest hard disk might have 6.0ms average latency. SSDs are usually below 0.1ms.
I never said I would use it. I'm just saying that whatever you were using it in 1990, it would still do those things and faster on a modern machine.
Which OS is still able to do run 15-20 years unmodified on constantly-evolving hardware?
DOS (including versions from 1990 or even earlier) still works unmodified on the latest hardware. Might only format 2GB partitions or somesuch but it works and AFAIK it would be proportionately faster on new hardware.
I suspect an old Linux kernel would still run on new hardware too. As long as you are still satisfied with the functions those operating systems support they could still be a viable alternative.
Guess what. Linux developers actually WRITE most of the drivers for Linux THEMSELVES with very little help from hardware manufacturers, for FREE! In contrast, it seems like MS couldn't even be bothered to make an effort to work with hardware manufacturers in order to get stable drivers out in a reasonable time for Vista users.
I don't think this is a fair comparison. Switching from one version of a given distribution of Linux to the next is much different than going from XP to Vista. Your settings are preserved, your GUI is close to identical unless you've decided to move to a new major release or a different GUI altogether. Basically there is a whole lot more choice involved in exactly what parts of the OS you are upgrading and what kind of learning curve you want to deal with
Generally new changes aren't forced down your throat like with MS Windows. Things aren't moved around in the GUI for no reason, application compatibility usually only gets better, and worse comes to worse you can get out your wallet and bribe a package maintainer to repackage whatever application you need for the latest version of your distro. Personally I would much rather talk to a Linux developer than with Microsoft Support.
The average buyer hasn't paid for a Microsoft upgrade ever. He gets the new OS when he buys his new PC. How is that 'pushy upgrades'
The average Microsoft user BUYS a new OS when he buys his new PC. There, fixed that for you.
It is a pushy upgrade if the customer doesn't want the new OS and would be happy to buy the new PC at a discount with no OS, get rid of his old PC, and continue to use his current XP license on his new PC. It is a pushy upgrade if the customer is being forced to a new GUI and new program incompatibilities because of lack of bugfix support by the OS manufacturer.
I concur. Your post is both funny and insightful. One time I complained to faculty about the reading material in an operating system class (at a university which I will not mention here), and was basically told that the reason the material was factually incorrect was because things change so often in the CS world. I argued that there were plenty of books written decades ago that were still completely relevant. I was told that those books were few and far between - which is true, but irrelevant AFAICT. I dropped this line of questioning, got an A in the class, and remained frustrated about the ineffectiveness (and in some cases outright falseness) of our assigned reading material.
Mod parent up. Microsoft licensing terms are constantly becoming more confusing and more onerous.
Sometimes I have a hard time understanding the license terms myself and Microsoft doesn't make it easy to get in touch with someone who knows.
I work for a system builder and, as the parent said, Microsoft is no longer letting us distribute installation CDs with Microsoft Office 2007 which is just a huge headache all around.
This is a hugely insightful post IMO.
http://phyun5.ucr.edu/~wudka/Physics7/Notes_www/no de6.html#SECTION02121000000000000000
Here is a link to a site describing the scientific method. Can you test the creation of a universe?? Even if you could, can you say for sure that just because you created it one way in your tests, that it wasn't created a different way the first time around? Evolution or creation is an event, an event cannot be scientifically proven to have happened. Maybe you could prove that it's possible, but beyond that is not the place for theories, only educated guessing.
I don't know how this got a 4 Insightful... Evolution cannot be proved or disproven by the scientific method either. We have no way of knowing what happened in order for the universe as we know it to come into existance, we weren't there. We can make educated guesses based on scientific evidence, but in my opinion, evolution is still a hyphothesis, and should never have gained theory status. Also, Creationism in itself does not contain any kind of anti-evolution agenda, its just different way of explaining how the universe came into existence, and in my opinion, is actually a more complete theory, as it takes into account how not just how the universe was formed, but how the underlying laws which govern those systems came to be.
Evolution is too broad and is compatible with any state of the world (through scientific guesses based on circumstancial evidence, which is what evolution amounts to at this point). Evolution is not a scientific theory, because it cannot be verified through use of the scientific method.
Mod Parent Up
I haven't read anything very interesting on any of the IT review sites lately. This ExtremeTech nonsense is just a continuation of that trend. I started to skim the article, saw the part that said they had implemented keyboard shortcuts for the gmail interface. Thats way cool! Now how did they do it? Theres no mention of what type of coding gMail is built with, is it just fancy CSS? is it Javascript? Is the mail client actually a Java app?! WTH is the MEAT?! I want to know how it runs! The comparison chart at the end is totally lame, nothing important or non-obvious was compared there. Also I was unimpressed with the blatant dissing of whoever the government official was who talked about gMail bringing billboards to your desktop. Obviously the advertising area in gMail is not a big deal, but you can say that, and leave it alone. This bashing of the politician did NOT enhance the article at all.
ermm IPv6 is 128 bits rather than the 64 the article incorrectly states.