I'll make you a bet. If SCO gives up it its suit within the period 2 months before Vista's release through its release date, I'll quit posting on Slashdot. If they don't, you quit posting. Deal or No Deal?
As I said in another post, the mere fact that NT doesn't contain legacy code doesn't mean it doesn't have legacy issues to contend with. Did you throw away all your legacy applications when upgrading from Win9x to NT?
My orginal point was that if MS had to give up the ability to be compatible with legacy applications they could produce a much better OS.
"Microsoft has (tens of) thousands of software developers. Yes, they have a lot of extremely poor programmers."
Well, if what you originally intended to say is that MS has a lot of extremely poor programmers simply because they have a lot of programmers, that's at least plausible. It would be equally plausible to argue that MS has a lot of extremely good programmers for exactly the same reason. What I don't find plausible is the idea that MS has a far greater percentage of extremely poor programmers than everybody else.
"You claim that Microsoft is hobbled by the "8088 legacy" (ignoring the fact that the NT line of software carries virtually no legacy from the DOS days), which would only make sense if they were a) still writing software for the 8088 (which of course they aren't), b) modern x86 shares the similar problems. Of course the latter is only laughably true, and modern x86 is extremely rich and full featured"
The fact that MS doesn't program for the 8088 anymore or that the NT line of software doesn't have a single line of real mode code doesn't mean that the legacy of the 8088 era (and the eras in between then and now) isn't a factor. Legacy code written by third parties that customers use must still be able to run (in most cases).
Keep in mind the UNIX was created at a time that teletypes or dumb terminals were state of the art and a CLI command shell was the best interface you could use at that time. Today we don't use use teletypes or dumb terminals (for the most part) yet Linux would not have been successful if the command shell had been eliminated because all the legacy utilities that users had been accostumed to wouldn't work. This despite the fact that no legacy code was used to create Linux.
So legacy issues are a bit more complicated than you imagine.
"..but they also have a lot of extraordinarily poor programmers."
That's exactly the myth that my post was intended to provoke. Thanks for falling for it.
"Do you have any idea how rich, and simple, modern x86 is?"
How did we get from the 8088 to the "modern x86"? The 8088 and 8086 have far more in common with the 8 bit 8080 than they do with Intel's current processors. Here's a homework assignment: try porting the Linux kernel to an 8088 PC platform. It should be easy since you believe the 8088 to be so "rich and simple". I'll make it easier: You can use the full 1MB address space of the 8088.
"And, as touched upon, the idea that starting from scratch is a desirable position is usually the foolish mistake of beginners. Most people starting from scratch just make the same, or similar, mistakes, without the advantage of launching off of prior work."
Actually an OS like DOS doesn't represent a mistake. It was an appropriate OS for the limitations of early PC's. Maintaning compatiblity over the years was a great business decision even if it made code ugly.
But mistake or not, It's rather silly to suggest, that MS (or anyone else) would design the same thing today as they did in 1981, or make the same mistakes.
"They understand that someday they will have to ditch the entire spaghetti code base that is Windows(TM)."
I'm sure that MS's competitors (including OSS) look forward to that day. On the other hand, the idea of MS's formidable programmers writing brand-new code without the legacy of the 8088 PC platform and years of legacy applications to deal with should scare the crap out of their competitors. They've been fighting with their left hand for years.
This is a common scenario I encounter when discussing Linux. When the discussion is about features and capabilities, people point to the biggest distros, when you talk about system requirements or security issues, suddenly Linux is just a kernel with a few utilities. Should we use Windows CE as the benchmark for the minimum system requirements for Windows?
I think its fair to conclude that a claim that the change in system requirements from version to version of Linux is negligable is false unless it is qualified by naming the specific distro it applies to. It certainly doesn't apply to RH Linux.
For example, the minimum system requirements for Red Hat Linux 5.2 was a 386 PC with 8MB of RAM and 100MB of hard disk space. Compare that to the current version.
There are plenty of successful applications that don't come "in the box" with Windows. Remember AOL made the bulk of its money as a proprietary portal using software that didn't come with the PC. If a real "Web 2.0" were compelling enough to be favored over what we have today, installing a new application to enable it would not be a big deal.
We agree on your last sentence, but not on the rest.
A truly new approach isn't going to be based on any current browser, which means that IE's 85% of the browser market would be irrelevent. Many popular applications that run on the PC were not developed by MS. Historically MS's support and cooperation are not necessary for a successful product (obviously, it doesn't hurt).
"If you're working at a huge company like HP, you're not that sort of person. You're mediocre, that's why you're there. You're a person who wants security. You'll sacrifice creative control for it, you'll do a shoddy job because thats what the manager wants, you just want to get paid. You went to school for a career, and this is it. You like safe and lazy. Not the best attitudes for telecommuters."
What a bunch of crap. I've worked at both very large and very small companies and the fact is that people in small companies can be just as mediocre, want security just as much, do shoddy work etc as anyone in a large company.
On the other hand, If you work as an indepedent consultant your nose is going to be a lot browner than most company men.
There are very few engineers in the real world who have "creative control" (not exactly an engineering term anyway) unless they have money from daddy and are pretending to be entrepreneur.
"1. Microsoft doesn't want the web to enable something that will threaten its monopoly in OS and Office software"
What does MS have to do with it? Nobody got the OK from MS to create HTTP, HTML or web browsers.
Of course, if what you mean is that you want a common standard based on the above legacy technologies than it's not going to improve things much anyway. The solution isn't to make a universal JavaScript, it's to eliminate the need for JavaScript in most scenarios.
"If you want to call something Web 2.0, pick a new port and design a new open protocol."
This is a great idea that probably won't happen. There's so much boiler-plate Javascript just to add capabilites that should have been part of the protocol from the beginning.
If neither RMS nor the FSF own any trademarks, than you're right, it doesn't make sense. On the other hand, if they do, than they've already reserved the right to take such actions and it's not unreasonable to imagine circumstances where they would. Nobody establishes a trademark unless they want legal protection.
The whole browser-war scenario was a great ploy for Netscape to attack MS and get their big payday by selling AOL the legal windfall, but it really hasn't made MS the dominant portal or stopped anyone else from being successful on the Internet. In other words, none of the scary scenarios that were supposed to happen if IE became popular have come to pass.
So just as having the dominant browser hasn't made MS dominant on the net, losing share to FireFox isn't going to hurt much more than their pride.
"Remember, there are different rules for monopolies."
And the typical Slashdotter view of these rules says that MS is not allowed to do much of anything. It's obvious that suing MS lately has been a great way of providing income that would otherwise have to come from honest labor, but the fact that MS has been willing to settle these suits says little about those "different rules" that warm the hearts of many Slashdotters.
I don't know if MS's concern is valid or not, but keep in mind that nobody ever made of a ton of money suing Apple. A number of companies, however, have made a few dollars suing MS:
Do you recommend that applications be written in assembly? If you're willing to accept the performance hit of using C instead of assembly, than perhaps you do believe there is an excuse for inefficient code after all.
The difference between choosing C over Java or C# is the degree of balance between performance and productivity. The compromise is different for different languages, but make no mistake, it is a compromise that all languages make.
I have been reading about thin client business failures for about a decade now. Despite the TCO hype, these devices need to be a lot cheaper than low-end PC's before they're going to be more than a niche product.
"It can come with up to 64MB of flash memory and 128MB RAM."
That means that the entry level system includes less than 64MB of flash or less than 128MB of RAM or less of both. Given the price, this looks like a PC for suckers.
Slashdot Mirror
I'll make you a bet. If SCO gives up it its suit within the period 2 months before Vista's release through its release date, I'll quit posting on Slashdot. If they don't, you quit posting. Deal or No Deal?
I don't mind if you mod me down but why don't you use "Flamebait" or "Troll". You just look stupid when you use "overrated" on an unmoderated post.
As I said in another post, the mere fact that NT doesn't contain legacy code doesn't mean it doesn't have legacy issues to contend with. Did you throw away all your legacy applications when upgrading from Win9x to NT?
My orginal point was that if MS had to give up the ability to be compatible with legacy applications they could produce a much better OS.
"Microsoft has (tens of) thousands of software developers. Yes, they have a lot of extremely poor programmers."
Well, if what you originally intended to say is that MS has a lot of extremely poor programmers simply because they have a lot of programmers, that's at least plausible. It would be equally plausible to argue that MS has a lot of extremely good programmers for exactly the same reason. What I don't find plausible is the idea that MS has a far greater percentage of extremely poor programmers than everybody else.
"You claim that Microsoft is hobbled by the "8088 legacy" (ignoring the fact that the NT line of software carries virtually no legacy from the DOS days), which would only make sense if they were a) still writing software for the 8088 (which of course they aren't), b) modern x86 shares the similar problems. Of course the latter is only laughably true, and modern x86 is extremely rich and full featured"
The fact that MS doesn't program for the 8088 anymore or that the NT line of software doesn't have a single line of real mode code doesn't mean that the legacy of the 8088 era (and the eras in between then and now) isn't a factor. Legacy code written by third parties that customers use must still be able to run (in most cases).
Keep in mind the UNIX was created at a time that teletypes or dumb terminals were state of the art and a CLI command shell was the best interface you could use at that time. Today we don't use use teletypes or dumb terminals (for the most part) yet Linux would not have been successful if the command shell had been eliminated because all the legacy utilities that users had been accostumed to wouldn't work. This despite the fact that no legacy code was used to create Linux.
So legacy issues are a bit more complicated than you imagine.
"What sort of idiotic comment is that? Are you disagreeing with even the thought that Microsoft has some poor programmers? Grow up."
How did we get from "a lot of extraordinarily poor programmers" to "some poor programmers".
"Your claim was that Microsoft was constrained and limited by the "legacy of the 8088 platform". Carry your own argument forward, moron."
I have no idea what you mean.
"Go troll MySpace personas or something."
Sorry but I'm too out of touch to understand the insult. What is MySpace?
You're referring to the crap code you've never read? Which MS applications are you referring to that don't have to deal with legacy issues?
"..but they also have a lot of extraordinarily poor programmers."
That's exactly the myth that my post was intended to provoke. Thanks for falling for it.
"Do you have any idea how rich, and simple, modern x86 is?"
How did we get from the 8088 to the "modern x86"? The 8088 and 8086 have far more in common with the 8 bit 8080 than they do with Intel's current processors. Here's a homework assignment: try porting the Linux kernel to an 8088 PC platform. It should be easy since you believe the 8088 to be so "rich and simple". I'll make it easier: You can use the full 1MB address space of the 8088.
"And, as touched upon, the idea that starting from scratch is a desirable position is usually the foolish mistake of beginners. Most people starting from scratch just make the same, or similar, mistakes, without the advantage of launching off of prior work."
Actually an OS like DOS doesn't represent a mistake. It was an appropriate OS for the limitations of early PC's. Maintaning compatiblity over the years was a great business decision even if it made code ugly.
But mistake or not, It's rather silly to suggest, that MS (or anyone else) would design the same thing today as they did in 1981, or make the same mistakes.
"They understand that someday they will have to ditch the entire spaghetti code base that is Windows(TM)."
I'm sure that MS's competitors (including OSS) look forward to that day. On the other hand, the idea of MS's formidable programmers writing brand-new code without the legacy of the 8088 PC platform and years of legacy applications to deal with should scare the crap out of their competitors. They've been fighting with their left hand for years.
This is a common scenario I encounter when discussing Linux. When the discussion is about features and capabilities, people point to the biggest distros, when you talk about system requirements or security issues, suddenly Linux is just a kernel with a few utilities. Should we use Windows CE as the benchmark for the minimum system requirements for Windows?
I think its fair to conclude that a claim that the change in system requirements from version to version of Linux is negligable is false unless it is qualified by naming the specific distro it applies to. It certainly doesn't apply to RH Linux.
For example, the minimum system requirements for Red Hat Linux 5.2 was a 386 PC with 8MB of RAM and 100MB of hard disk space. Compare that to the current version.
The only reasonable way that you can compare an oven and a computer is to overclock the CPU and disable the fan.
There are plenty of successful applications that don't come "in the box" with Windows. Remember AOL made the bulk of its money as a proprietary portal using software that didn't come with the PC. If a real "Web 2.0" were compelling enough to be favored over what we have today, installing a new application to enable it would not be a big deal.
""Agile methods" are largely an acknowledgement that mistakes are inevitable and should be planned for."
This is hardly something that orginated with "Agile methods".
We agree on your last sentence, but not on the rest.
A truly new approach isn't going to be based on any current browser, which means that IE's 85% of the browser market would be irrelevent. Many popular applications that run on the PC were not developed by MS. Historically MS's support and cooperation are not necessary for a successful product (obviously, it doesn't hurt).
"If you're working at a huge company like HP, you're not that sort of person. You're mediocre, that's why you're there. You're a person who wants security. You'll sacrifice creative control for it, you'll do a shoddy job because thats what the manager wants, you just want to get paid. You went to school for a career, and this is it. You like safe and lazy. Not the best attitudes for telecommuters."
What a bunch of crap. I've worked at both very large and very small companies and the fact is that people in small companies can be just as mediocre, want security just as much, do shoddy work etc as anyone in a large company.
On the other hand, If you work as an indepedent consultant your nose is going to be a lot browner than most company men.
There are very few engineers in the real world who have "creative control" (not exactly an engineering term anyway) unless they have money from daddy and are pretending to be entrepreneur.
"1. Microsoft doesn't want the web to enable something that will threaten its monopoly in OS and Office software"
What does MS have to do with it? Nobody got the OK from MS to create HTTP, HTML or web browsers.
Of course, if what you mean is that you want a common standard based on the above legacy technologies than it's not going to improve things much anyway. The solution isn't to make a universal JavaScript, it's to eliminate the need for JavaScript in most scenarios.
"If you want to call something Web 2.0, pick a new port and design a new open protocol."
This is a great idea that probably won't happen. There's so much boiler-plate Javascript just to add capabilites that should have been part of the protocol from the beginning.
If neither RMS nor the FSF own any trademarks, than you're right, it doesn't make sense. On the other hand, if they do, than they've already reserved the right to take such actions and it's not unreasonable to imagine circumstances where they would. Nobody establishes a trademark unless they want legal protection.
The whole browser-war scenario was a great ploy for Netscape to attack MS and get their big payday by selling AOL the legal windfall, but it really hasn't made MS the dominant portal or stopped anyone else from being successful on the Internet. In other words, none of the scary scenarios that were supposed to happen if IE became popular have come to pass.
So just as having the dominant browser hasn't made MS dominant on the net, losing share to FireFox isn't going to hurt much more than their pride.
"Remember, there are different rules for monopolies."
And the typical Slashdotter view of these rules says that MS is not allowed to do much of anything. It's obvious that suing MS lately has been a great way of providing income that would otherwise have to come from honest labor, but the fact that MS has been willing to settle these suits says little about those "different rules" that warm the hearts of many Slashdotters.
I don't know if MS's concern is valid or not, but keep in mind that nobody ever made of a ton of money suing Apple. A number of companies, however, have made a few dollars suing MS:
Sun: $20 million (2001), $2 billion (2004)
AOL: $750 millon (2003)
Real: $460 millon(2005)
Do you recommend that applications be written in assembly? If you're willing to accept the performance hit of using C instead of assembly, than perhaps you do believe there is an excuse for inefficient code after all.
The difference between choosing C over Java or C# is the degree of balance between performance and productivity. The compromise is different for different languages, but make no mistake, it is a compromise that all languages make.
I have been reading about thin client business failures for about a decade now. Despite the TCO hype, these devices need to be a lot cheaper than low-end PC's before they're going to be more than a niche product.
"It can come with up to 64MB of flash memory and 128MB RAM."
That means that the entry level system includes less than 64MB of flash or less than 128MB of RAM or less of both. Given the price, this looks like a PC for suckers.
"If they did pull something like this then I can't imagine a better incentive to switch to a different OS."
Which is exactly why it will never happen.