I'm not arguing that Windows is as flexible as the average *nix. Nobody is. Comparing open source to closed source is a waste of time.
The previous posts were (as I understood it) implying the following:
1) He wrote a five line script that ran in user space and cached items in memory
2) SuperFetch must be part of the system to work in Windows
3) Therefore, Windows is inflexible
Still with me? Now, here's what your problem is: You don't know anything about SuperFetch. As another post suggested, you should watch the video and learn a thing or two before arguing a point. The flawed logic arises because the above proof is contingent upon the five line script and SuperFetch being comparable in functionality. They aren't at all.
You can make plenty of arguments for why Windows isn't as flexible as *nix, but comparing a five line script to SuperFetch isn't one of them. That's terrible logic, as my first post pointed out.
Now, if you want to make a claim that Windows is inflexible because you can't write a crappy user space script to load your files into memory, and you can on *nix, that's fine (and perhaps this is what you were trying to do). If that was the original intent, your argument has nothing to do with SuperFetch and is off topic. Comparing the script to SuperFetch leaves you looking like you can't discern between apples & oranges.
No. His logic is terrible, and so is yours. Windows doesn't let you muck around with most things in the system; that's a design decision to keep the average idiot out. Spin that however you must to make yourself feel like a big man.
From my last post:
... It's loading files *when* they're commonly used...
I didn't say it loads files when they're used. I said it loads them when they're commonly used. Understand it this time? There's a difference.
SuperFetch doesn't just load all of your commonly used files into memory, as your very impressive five line script does. It preloads them only for periods in which it has recognized you typically use the file.
Because this feature isn't difficult to implement, it's proof of how inflexible Windows? Your logic is terrible.
There are valid reasons to bash any software package, including Vista. If you use one, you have the added bonus of not sounding like a moron.
Also, this isn't just preloading commonly opened files at boot time. It's loading files *when* they're commonly used, and avoiding unnecessarily paging out files you may want again in the futre. Pat yourself on the back again for your 5 lines, dear script kiddie, and then take some time to research what you're talking about.
So they better have improved their techniques with this SuperFetch.
You don't really seem to know what you're talking about (although I suppose that doesn't prohibit anyone from being "5, insightful" on/.). They can't "improve their techniques," because there was no version of this feature in XP.
Those "quick start apps" you mention have nothing to do with XP, and everything to do with application writers who think you want their garbage running all the time. Those aren't just "pre-loaded" into memory, they're scheduled processes that are wasting your time and resources. SuperFetch is completely different.
SuperFetch just uses heuristics to manage memory in an effort to keep items you'll want from being paged out, and if memory is available it will load something it predicts (based on your usage) you'll want. It won't schedule anything new.
Obviously, a poor implementation would slow your system down. That's the case with any memory management techniques, and isn't worth noting. Unless you're on Slashdot, and it applies to Microsoft and not Apple or Google.
---------
If it has more features than an Apple product, it's like totally bloated.
For the FY ending 6/30/05, MSFT had a net income of US$ 12.2 billion. So, a fine of US $357 million IS significant -- it's roughly 3% of their net income.
Make sure we're all clear that net income is their profit. This isn't going to impact their budget for anything, since it is, by definition, the money that's left over after all the bills are paid.
This will come out of cash that piles up and gets distributed to shareholders in the future, and has zero potential to impact a current budget. It's also only ~10 days worth. Not such a big deal.
Really? MIT isn't one of the premier educational institutions in the world? Do you have evidence to support that claim?
Nobody said it was the best institution there is, but I'd like to see you form a (constructive) argument against calling it one of the premier institutions. In the future I suggest you avoid unfounded statements.
I'd also like to hear the reasoning behind your being awarded a score of "5, Insightful."
I know bashing the USA on Slashdot is as trendy as typing Microsoft with a dollar sign ("OMG M$ SUX!!!"), but please... lets be a bit more selective, moderators.
The only way MS wins is by manipulating the free and open markets.
Microsoft buying companies is part of the free and open market. The majority of people on slashdot need to grow up and stop using shitty arguments. I use *nix whenever possible, but it'll never take off as long as its fanboys make irrational comments/gripe & whine about how they're mistreated.
There are good arguments out there for/against MS, Apple, *nix, etc.. And if you use them, you have the added bonus of not embarrassing yourself.
Apple has a lot few programmers and a smaller budget than Microsoft....
Apple also based OSX on a research variation of Unix developed at CMU and in doing so kept their actual contributions down. Those "few programmers" didn't come close to taking the project from start to finish. Poor example, but I do appreciate your point.
THEIR software won a contest against a lot of competitors, and made it across the desert.
RedTeam may have won a public contest (er... second and third place?), but DARPA has been testing Spinner/Crusher privately on military bases for quite some time now - and often on more difficult terrain (who needs a path?). NREC has developed an incredibly reliable platform that exceeded DARPA's wildest expectations and it has been extensively tested. That being said, if there really are better algorithms now available, perhaps that's something that will appear in v3.0.
In short, maybe RedTeam's software is a breakthrough. I'm certainly not going to discredit/give praise to their work without reviewing it personally (unlike the Wired author). I fully agree with you that brilliance often comes in small packages - that's part of what keeps the frontiers of technology so exciting. And I'm VERY proud of what my peers who were a part of RedTeam accomplished. Being surrounded by these people on a daily basis is really something I'm going to miss.
Not sharing technology seems kind of limiting to me.
Your comment is based on the assumption that the research/techniques/algorithms used on these vehicles doesn't already match/surpass what was implemented by students as an extra-curricular project. Here's a bit of background:
Before Spinner was deployed, there were two separate projects at NREC:
1) PerceptOR - a project devoted to development of advanced processing/sensing techniques on autonomous vehicles
2) UGCV - (Unmanned Ground Combat Vehicle) a project devoted to exploring physical constraints & developing the mechanics of the system.
These two projects were combined to form UPI (UGCV + PerceptOR Integrated), and gave birth to Spinner/Crusher. As such, these vehicles already achieved very high levels of sophistication prior to the Grand Challenge. That said, it wouldn't exactly be permissible for NREC to donate years of semi-private research to RedTeam. Plus, they did just fine on their own. (Many RedTeam participants performed their research for course credit. Why eliminate the learning experience?)
In terms of mounting weapons, that's the military's call. CMU builds vehicles that meet certain payload specifications. What is done with the platform we develop is entirely up to the military.
DARPA probably wouldn't be too thrilled if I ranted about technical design decisions or specific sensors, so to be safe I think I'll leave it up to them to disclose what they deem fit. Sorry.
Preface: I'm a CMU student, and have worked on this specific project.
--
This project has always been kept separate from the grand challenge. Spinner/Crusher are brought to us by CMU's NREC (National Robotics Engineering Center), while the RedTeam is responsible for CMU's involvement in the Grand Challenge. We never had communication or shared technology with the RedTeam.
It would be rather unfair to the Grand Challenge if years of DARPA funded research/development (like Spinner/Crusher) were used to win a DARPA sponsored competition. And I can say with confidence that Spinner/Crusher would have dominated the Grand Challenge. The resources at NREC's disposal can't be matched by a (mostly) volunteer group of students.
In short, CMU's contract with DARPA was established well before the Grand Challenge - the Spinner/Crusher research/development began several years ago. I'm flattered you think we're good enough to throw this together as a result of the Grand Challenge, though;)
Slashdot Mirror
Let me try and help you get back on track.
I'm not arguing that Windows is as flexible as the average *nix. Nobody is. Comparing open source to closed source is a waste of time.
The previous posts were (as I understood it) implying the following:
1) He wrote a five line script that ran in user space and cached items in memory
2) SuperFetch must be part of the system to work in Windows
3) Therefore, Windows is inflexible
Still with me? Now, here's what your problem is: You don't know anything about SuperFetch. As another post suggested, you should watch the video and learn a thing or two before arguing a point. The flawed logic arises because the above proof is contingent upon the five line script and SuperFetch being comparable in functionality. They aren't at all.
You can make plenty of arguments for why Windows isn't as flexible as *nix, but comparing a five line script to SuperFetch isn't one of them. That's terrible logic, as my first post pointed out.
Now, if you want to make a claim that Windows is inflexible because you can't write a crappy user space script to load your files into memory, and you can on *nix, that's fine (and perhaps this is what you were trying to do). If that was the original intent, your argument has nothing to do with SuperFetch and is off topic. Comparing the script to SuperFetch leaves you looking like you can't discern between apples & oranges.
No. His logic is terrible, and so is yours. Windows doesn't let you muck around with most things in the system; that's a design decision to keep the average idiot out. Spin that however you must to make yourself feel like a big man.
... It's loading files *when* they're commonly used...
From my last post:
I didn't say it loads files when they're used. I said it loads them when they're commonly used. Understand it this time? There's a difference.
SuperFetch doesn't just load all of your commonly used files into memory, as your very impressive five line script does. It preloads them only for periods in which it has recognized you typically use the file.
Because this feature isn't difficult to implement, it's proof of how inflexible Windows? Your logic is terrible.
There are valid reasons to bash any software package, including Vista. If you use one, you have the added bonus of not sounding like a moron.
Also, this isn't just preloading commonly opened files at boot time. It's loading files *when* they're commonly used, and avoiding unnecessarily paging out files you may want again in the futre. Pat yourself on the back again for your 5 lines, dear script kiddie, and then take some time to research what you're talking about.
So they better have improved their techniques with this SuperFetch.
/.). They can't "improve their techniques," because there was no version of this feature in XP.
You don't really seem to know what you're talking about (although I suppose that doesn't prohibit anyone from being "5, insightful" on
Those "quick start apps" you mention have nothing to do with XP, and everything to do with application writers who think you want their garbage running all the time. Those aren't just "pre-loaded" into memory, they're scheduled processes that are wasting your time and resources. SuperFetch is completely different.
SuperFetch just uses heuristics to manage memory in an effort to keep items you'll want from being paged out, and if memory is available it will load something it predicts (based on your usage) you'll want. It won't schedule anything new.
Obviously, a poor implementation would slow your system down. That's the case with any memory management techniques, and isn't worth noting. Unless you're on Slashdot, and it applies to Microsoft and not Apple or Google.
---------
If it has more features than an Apple product, it's like totally bloated.
If I decide that I just love 1990's Ska, at some point I am going to download all of the good 90's ska that there is.
/.
I stopped reading your post after this, and I really don't care what else you had to say.
Why? I'm too busy downloading 1990's Ska with my Zune Pass.
That's right. You heard it. I love my Zune, and Linux, and I read
For the FY ending 6/30/05, MSFT had a net income of US$ 12.2 billion. So, a fine of US $357 million IS significant -- it's roughly 3% of their net income.
Make sure we're all clear that net income is their profit. This isn't going to impact their budget for anything, since it is, by definition, the money that's left over after all the bills are paid.
This will come out of cash that piles up and gets distributed to shareholders in the future, and has zero potential to impact a current budget. It's also only ~10 days worth. Not such a big deal.
Really? MIT isn't one of the premier educational institutions in the world? Do you have evidence to support that claim?
Nobody said it was the best institution there is, but I'd like to see you form a (constructive) argument against calling it one of the premier institutions. In the future I suggest you avoid unfounded statements.
I'd also like to hear the reasoning behind your being awarded a score of "5, Insightful." I know bashing the USA on Slashdot is as trendy as typing Microsoft with a dollar sign ("OMG M$ SUX!!!"), but please... lets be a bit more selective, moderators.
The only way MS wins is by manipulating the free and open markets.
Microsoft buying companies is part of the free and open market. The majority of people on slashdot need to grow up and stop using shitty arguments. I use *nix whenever possible, but it'll never take off as long as its fanboys make irrational comments/gripe & whine about how they're mistreated.
There are good arguments out there for/against MS, Apple, *nix, etc.. And if you use them, you have the added bonus of not embarrassing yourself.
Apple has a lot few programmers and a smaller budget than Microsoft....
Apple also based OSX on a research variation of Unix developed at CMU and in doing so kept their actual contributions down. Those "few programmers" didn't come close to taking the project from start to finish. Poor example, but I do appreciate your point.
THEIR software won a contest against a lot of competitors, and made it across the desert.
RedTeam may have won a public contest (er... second and third place?), but DARPA has been testing Spinner/Crusher privately on military bases for quite some time now - and often on more difficult terrain (who needs a path?). NREC has developed an incredibly reliable platform that exceeded DARPA's wildest expectations and it has been extensively tested. That being said, if there really are better algorithms now available, perhaps that's something that will appear in v3.0.
In short, maybe RedTeam's software is a breakthrough. I'm certainly not going to discredit/give praise to their work without reviewing it personally (unlike the Wired author). I fully agree with you that brilliance often comes in small packages - that's part of what keeps the frontiers of technology so exciting. And I'm VERY proud of what my peers who were a part of RedTeam accomplished. Being surrounded by these people on a daily basis is really something I'm going to miss.
Not sharing technology seems kind of limiting to me.
Your comment is based on the assumption that the research/techniques/algorithms used on these vehicles doesn't already match/surpass what was implemented by students as an extra-curricular project. Here's a bit of background:
Before Spinner was deployed, there were two separate projects at NREC:
1) PerceptOR - a project devoted to development of advanced processing/sensing techniques on autonomous vehicles
2) UGCV - (Unmanned Ground Combat Vehicle) a project devoted to exploring physical constraints & developing the mechanics of the system.
These two projects were combined to form UPI (UGCV + PerceptOR Integrated), and gave birth to Spinner/Crusher. As such, these vehicles already achieved very high levels of sophistication prior to the Grand Challenge. That said, it wouldn't exactly be permissible for NREC to donate years of semi-private research to RedTeam. Plus, they did just fine on their own. (Many RedTeam participants performed their research for course credit. Why eliminate the learning experience?)
In terms of mounting weapons, that's the military's call. CMU builds vehicles that meet certain payload specifications. What is done with the platform we develop is entirely up to the military.
DARPA probably wouldn't be too thrilled if I ranted about technical design decisions or specific sensors, so to be safe I think I'll leave it up to them to disclose what they deem fit. Sorry.
Preface: I'm a CMU student, and have worked on this specific project. --
;)
This project has always been kept separate from the grand challenge. Spinner/Crusher are brought to us by CMU's NREC (National Robotics Engineering Center), while the RedTeam is responsible for CMU's involvement in the Grand Challenge. We never had communication or shared technology with the RedTeam.
It would be rather unfair to the Grand Challenge if years of DARPA funded research/development (like Spinner/Crusher) were used to win a DARPA sponsored competition. And I can say with confidence that Spinner/Crusher would have dominated the Grand Challenge. The resources at NREC's disposal can't be matched by a (mostly) volunteer group of students.
In short, CMU's contract with DARPA was established well before the Grand Challenge - the Spinner/Crusher research/development began several years ago. I'm flattered you think we're good enough to throw this together as a result of the Grand Challenge, though