Re:Various conspiracy theories...
on
CherryOS On Hold
·
· Score: 1
Here's a conspiracy theory for you:
Maui software actually was "inspired" by PearPC and hired a Pakistani company to create a similar product. The Pakistani company secretly stole PearPC to deliver the product on time (partly because they had no idea what they were doing) and then Maui compounded the issue by covering the fact that they'd used an overseas company. (They didn't want to drive customers away by admitting that.) So now Maui has a hot (as in stolen) product on their hands, and is trying like hell to get rid of it.
How's that for a conspiracy theory? Considering everything we know, quite plausable too.
Ah yes. Only on Slashdot does the first post get modded "redundant". If I'd instead reposted my text of the first post from the original article, that some mod would have given it a +1 Informative. Perhaps the mod in question should rethink his/her strategy?
Does anyone have a clue as to the (very approximate) size of the/. crowd?
See that number next to your name? That's your userid. Those userids *are* sequential, and started at (wait for it...) 1. With that in mind, run around and find the highest userid you can (I've see over 800,000 myself) and then you'll have a "very approximate" idea of how large the Slashdot crowd is.
I can't say that I'd ever consider leasing unless it turned out to be MUCH cheaper. Old hardware can always be reused and/or sold off for temporary budget increases. Not to mention that high-wear equipment like laptops tend to break, thus requiring you to pay for the damage.
That being said, it does have certain political advantages. Having your equipment on lease ensures that the company *must* allow you to upgrade the equipment or go without.
It makes sense to register *both* bob-wilson.com and bobwilson.com, but if you can't get bobwilson.com, it does NOT make sense to register bob-wilson.com. It might also get you in trouble with the other domain owner.
On a sidenote, your karma is now good enough to get an automatic +2. Use the "No Karma Bonus" checkbox when you feel like shouting might get you in trouble. Congratulations, and welcome to the good posters club.
I mean, they are both distributed filesystems with the same name. What are the odds?;)
Considering that it's in vogue to name file systems with one letter in front of "FS"? About 1 in 26. The odds are even better if you discount commonly used file systems such as XFS, UFS, FFS, NFS, and JFS.
At the time OpenMosix had their own distributed filesystem called MFS. But it's proved inadequate, which is why they are switching to GFS
I'm sorry, did I miss the point at which Google made an open source implementation of GFS? Last I knew, the only docs for GFS were the papers that Google published on the concept. And those papers (unfortunately) seemed to lack a few of the finer details of implementation.
Funny you should mention that. It just so happens that COM was a technology originally developed for Unix, but Microsoft purchased the company that developed it and made it into their basis for Windows components.
I haven't seen a commercial product that paid the slightest attention to CompSci foundations in years, leaving them with the sort of saleable "usefulness" that pleases the marketing department, but a bit lacking in the sort of usefulness that "gets shit done."
While I agree with your general complaint, allow me to point something out:All modern computer products are based upon the CompSci foundations laid out by researchers years ago. They don't have to pay much attention to CompSci theory, because the APIs, hardware, OSes, and Virtual Machines do all the work for them.
That being said, there are a lot of idiots in the field who cheated or slept their way through CompSci. (Or perhaps they were taught the "marketable" brand of "Comp[Not]Sci") That, however, is a separate problem from scientific research.
It could move CompSci research back to an academic field conducted in the universities, if the universities themselves hadn't already forgotten what CompSci was and devolved into Java trade schools, because Java is "useful."
The sad truth, however, is that it's happening in ALL fields. For example, most of the crack aerospace and nuclear engineers I've talked to have iterated the same complaint as you. The only difference is that they're speaking about their own field instead of CompSci. Feel safer about flying yet?;-)
Experimented with and designed in the 70's and 80's. Commercially available in the 90's.
- the Web
Experimented with and designed in the 80's. Commercially available in the 90's.
- scalable clusters and Internet services
Experimented with and designed in the 80's. Commercially available in the 90's.
- mobile computing
Commercially available since the 80's. Lowering costs of commercial hardware made mobile devices more popular in the 90's and 00's.
- breakthroughs in graphics
All designed in the 60's through 80's, but lacking in powerful enough hardware until the late 90's.
- breakthroughs in vision
???
- stunning advancements in computer architecture
Eh? What stunning advancements? Most of the architectures in use today go all the way back to the early 70's. They've merely become commercially available to the average Joe in recent years.
- fundamental advances in theory, algorithms, etc.
The whole purpose of long-term research isn't to bang out invention after invention. It's an investment in the future of the technology.
I understand that quite well. But I'm still not seeing amazing new algorithms that have future potential in many areas. AI seems to be the most promising, with most other areas of research trying to tackle the same sorts of problems without AI.
Beyond AI, I have a very difficult time coming up with CompSci advances in the last decade. The BWT algo, Bayesian Filters, and that's about where I run out.
To take just one example, this whole web thing of the 90s should not count for anything.
*Ahem* From your own link: The Web can be traced back to a project at CERN in 1989.
CS research is worthless
Didn't say that. I did say that there's not as much value as their used to be. The field is well saturated, and therefore is less likely to be much to be gained through expensive research. And as I also said, there's still research that's valuable, just far less overall.
real progress comes from companies like Google or Akamai. Oh wait... both came to us straight from the university (Stanford and MIT, respectively).
And how many millions of dollars did it take for PageRank to go from the start of research to an algorithm on paper? (Actually, I'd be quite interested to know. I'd expect that it probably wasn't more than a few thousand dollars.)
+5 insightful? People, this is *CompSci* we're talking about here. Think for a moment. What materials does a CompSci researcher need? A few thousand dollars worth of computing equipment? Maybe ten thousand a year in custom board manufacturing costs? Beyond that you're just talking about people's wages. This isn't chemistry or rocket science where rare and expensive materials are needed for experiments! This is computer science where 90-99% of the research is intellectual!
Just think for a moment here. If they've got massive multi-million dollar budgets, where is all the research money going?
The problem is that Computer Science hasn't advanced much since the 80's. All the core concepts have been long established, and precious little groundbreaking research has emerged. I hate to say it, but most of the valuable work being done today is at the commercial level. i.e. Building upon the CompSci foundations to create useful, real world products.
The biggest area that I see research being useful is in artificial intelligence. There's so much that we;re still trying to comprehend about emergent behaviors. Unfortunately, AI is very much like Fusion. It's only 20 years away (for the next century).:-) Not that I begrudge the AI research. It's fascinating stuff and deserves to be done. Just don't expect any sort of immediate results.
what is so special about the whole setup? Motherboards have been around for more than 20 years.
True, but what's up with the blue boards? They've been green for 20+ years, and yet in the past couple of years they've all been coming out blue. Are they using different materials of some sort, or is it just stylistic?
Yes, they were incompatible OSes. For example, the memory model used by Win32 was completely different from the Win3.1 memory model. Win32s was an attempt to remap the Win16 based OS into the Win32 design. And it actually worked.
WIn95 is a 32 bits GUI for MS-DOS
That's actually not true. Once Win95 started, DOS was kicked off into oblivion. The reason was that DOS couldn't actually handle protected mode, so once the CPU mode was changed DOS was effectively dead memory. Win95 actually emulated DOS inside of itself for old applications. The only reason why DOS was kept around is that it provided a convenient boot loader for Win95.
Yes [a separate Win32 API] was the case for MS-DOS+win (more or less - see above.) But this is not the case for WinNt.
Bullocks. The NT kernel and the Win32 API are separate, though closely related, entities. The kernel provides basic driver and I/O services which the Win32 API then remaps into a useful usermode API.
And taking into account that the crash of the GUI always crashes the OS your statement seems to be false.
More nonsense. This merely shows a lack of understanding of kernel design. There are two ways of producing graphics in an OS with memory protection and a clear separation between kernel and usermode. The slow way is to make most of the graphics processing happen in usermode, then jump to kernel space to perform the I/O to the video card. This is the safest route, but requires a lot of jumping in and out of kernel space. The alternative is to suck the entire graphics subsystem into a kernel mode driver. Since all processing happens inside kernel space, the graphics subsystem runs much faster. Unfortunately, any error generated in this space (e.g. GPF, buffer overflow, etc.) would be impossible to recover from without risking damage to the kernel's memory image.
FYI, NT 3.51 used the former method. This caused it to generally be more stable than 4.0 which switched to using the later method for performance. This tended to make 4.0 somewhat unstable for graphics intensive apps. (I used to kill it with GIMP all the time.) Win2K added additional checks that made many of these issues go away.
Ok, bad news. The "architect" has just confirmed it to be a prank. He did mention that it wasn't far fetched, though, and what they really need is the source code to the Internix(STU) project to make it happen. He's speculated that it may be possible to get Microsoft to "Shared Source" the library.:-)/ME stops happy dance and sighs
This is a cheap trick to disguise advertising as a "joke". You're supposed to go to ThinkGeek, see the "funny" products (haha), then notice that you can get a free TShirt if you buy $10. Offer good today only! So then like a sheep, you go and buy stuff that you otherwise couldn't quite justify.
But the new Trek producers can cut corners by utilizing existing props and sets from "Team America." So don't be surprised if Captain Archer looks suspiciously like Alec Baldwin, Trip Tucker like Matt Damon, Travis Mayweather like Samuel L. Jackson, T'Pol like Janeane Garofalo, and Malcolm Reed like Susan Sarandon.
That's the funniest shit I've seem all day! Kudos to StarTrek.com!:-D
Slashdot Mirror
Here's a conspiracy theory for you:
Maui software actually was "inspired" by PearPC and hired a Pakistani company to create a similar product. The Pakistani company secretly stole PearPC to deliver the product on time (partly because they had no idea what they were doing) and then Maui compounded the issue by covering the fact that they'd used an overseas company. (They didn't want to drive customers away by admitting that.) So now Maui has a hot (as in stolen) product on their hands, and is trying like hell to get rid of it.
How's that for a conspiracy theory? Considering everything we know, quite plausable too.
Did you hear what the geneticist said when he peered into the dark forest?
;-))
"Gee! Gnomes!"
Thank you, I'm here all week. (Actually, I'm not. But I am feeling cheeky today.
Ah yes. Only on Slashdot does the first post get modded "redundant". If I'd instead reposted my text of the first post from the original article, that some mod would have given it a +1 Informative. Perhaps the mod in question should rethink his/her strategy?
Does anyone have a clue as to the (very approximate) size of the /. crowd?
See that number next to your name? That's your userid. Those userids *are* sequential, and started at (wait for it...) 1. With that in mind, run around and find the highest userid you can (I've see over 800,000 myself) and then you'll have a "very approximate" idea of how large the Slashdot crowd is.
Glad to see the duplicate checking code is in good working order. NOT! You guys really need to reverse the polarity on that thing, or something.
You can't really sell it easily, and who would you sell it to?
I don't know about your company, but ours uses eBay.
I can't say that I'd ever consider leasing unless it turned out to be MUCH cheaper. Old hardware can always be reused and/or sold off for temporary budget increases. Not to mention that high-wear equipment like laptops tend to break, thus requiring you to pay for the damage.
That being said, it does have certain political advantages. Having your equipment on lease ensures that the company *must* allow you to upgrade the equipment or go without.
It makes sense to register *both* bob-wilson.com and bobwilson.com, but if you can't get bobwilson.com, it does NOT make sense to register bob-wilson.com. It might also get you in trouble with the other domain owner.
On a sidenote, your karma is now good enough to get an automatic +2. Use the "No Karma Bonus" checkbox when you feel like shouting might get you in trouble. Congratulations, and welcome to the good posters club.
I mean, they are both distributed filesystems with the same name. What are the odds? ;)
Considering that it's in vogue to name file systems with one letter in front of "FS"? About 1 in 26. The odds are even better if you discount commonly used file systems such as XFS, UFS, FFS, NFS, and JFS.
Ok, I looked it up. You're confusing Sistina's (now Red Hat) Global File System with the Google File System. The two ARE NOT THE SAME.
g hemawat.pdf (PDF): www.cs.rochester.edu/sosp2003/papers/p125-ghemawat .pdf+Google+File+System&hl=en&client=safari (HTML)
Here's Red Hat:
http://www.redhat.com/software/rha/gfs/
Here's Google:
http://www.cs.rochester.edu/sosp2003/papers/p125-
http://64.233.161.104/search?q=cache:m0TMQYgIlIoJ
At the time OpenMosix had their own distributed filesystem called MFS. But it's proved inadequate, which is why they are switching to GFS
I'm sorry, did I miss the point at which Google made an open source implementation of GFS? Last I knew, the only docs for GFS were the papers that Google published on the concept. And those papers (unfortunately) seemed to lack a few of the finer details of implementation.
IE is not really IE without COM.
Funny you should mention that. It just so happens that COM was a technology originally developed for Unix, but Microsoft purchased the company that developed it and made it into their basis for Windows components.
Hi KFG!
;-)
I haven't seen a commercial product that paid the slightest attention to CompSci foundations in years, leaving them with the sort of saleable "usefulness" that pleases the marketing department, but a bit lacking in the sort of usefulness that "gets shit done."
While I agree with your general complaint, allow me to point something out:All modern computer products are based upon the CompSci foundations laid out by researchers years ago. They don't have to pay much attention to CompSci theory, because the APIs, hardware, OSes, and Virtual Machines do all the work for them.
That being said, there are a lot of idiots in the field who cheated or slept their way through CompSci. (Or perhaps they were taught the "marketable" brand of "Comp[Not]Sci") That, however, is a separate problem from scientific research.
It could move CompSci research back to an academic field conducted in the universities, if the universities themselves hadn't already forgotten what CompSci was and devolved into Java trade schools, because Java is "useful."
The sad truth, however, is that it's happening in ALL fields. For example, most of the crack aerospace and nuclear engineers I've talked to have iterated the same complaint as you. The only difference is that they're speaking about their own field instead of CompSci. Feel safer about flying yet?
- parallel computing and supercomputing
Experimented with and designed in the 70's and 80's. Commercially available in the 90's.
- the Web
Experimented with and designed in the 80's. Commercially available in the 90's.
- scalable clusters and Internet services
Experimented with and designed in the 80's. Commercially available in the 90's.
- mobile computing
Commercially available since the 80's. Lowering costs of commercial hardware made mobile devices more popular in the 90's and 00's.
- breakthroughs in graphics
All designed in the 60's through 80's, but lacking in powerful enough hardware until the late 90's.
- breakthroughs in vision
???
- stunning advancements in computer architecture
Eh? What stunning advancements? Most of the architectures in use today go all the way back to the early 70's. They've merely become commercially available to the average Joe in recent years.
- fundamental advances in theory, algorithms, etc.
*What* fundamental advances? Name them!
The whole purpose of long-term research isn't to bang out invention after invention. It's an investment in the future of the technology.
I understand that quite well. But I'm still not seeing amazing new algorithms that have future potential in many areas. AI seems to be the most promising, with most other areas of research trying to tackle the same sorts of problems without AI.
Beyond AI, I have a very difficult time coming up with CompSci advances in the last decade. The BWT algo, Bayesian Filters, and that's about where I run out.
To take just one example, this whole web thing of the 90s should not count for anything.
*Ahem* From your own link: The Web can be traced back to a project at CERN in 1989.
CS research is worthless
Didn't say that. I did say that there's not as much value as their used to be. The field is well saturated, and therefore is less likely to be much to be gained through expensive research. And as I also said, there's still research that's valuable, just far less overall.
real progress comes from companies like Google or Akamai. Oh wait... both came to us straight from the university (Stanford and MIT, respectively).
And how many millions of dollars did it take for PageRank to go from the start of research to an algorithm on paper? (Actually, I'd be quite interested to know. I'd expect that it probably wasn't more than a few thousand dollars.)
+5 insightful? People, this is *CompSci* we're talking about here. Think for a moment. What materials does a CompSci researcher need? A few thousand dollars worth of computing equipment? Maybe ten thousand a year in custom board manufacturing costs? Beyond that you're just talking about people's wages. This isn't chemistry or rocket science where rare and expensive materials are needed for experiments! This is computer science where 90-99% of the research is intellectual!
Just think for a moment here. If they've got massive multi-million dollar budgets, where is all the research money going?
The problem is that Computer Science hasn't advanced much since the 80's. All the core concepts have been long established, and precious little groundbreaking research has emerged. I hate to say it, but most of the valuable work being done today is at the commercial level. i.e. Building upon the CompSci foundations to create useful, real world products.
:-) Not that I begrudge the AI research. It's fascinating stuff and deserves to be done. Just don't expect any sort of immediate results.
The biggest area that I see research being useful is in artificial intelligence. There's so much that we;re still trying to comprehend about emergent behaviors. Unfortunately, AI is very much like Fusion. It's only 20 years away (for the next century).
what is so special about the whole setup? Motherboards have been around for more than 20 years.
True, but what's up with the blue boards? They've been green for 20+ years, and yet in the past couple of years they've all been coming out blue. Are they using different materials of some sort, or is it just stylistic?
They weren't incompatible OS's.
Yes, they were incompatible OSes. For example, the memory model used by Win32 was completely different from the Win3.1 memory model. Win32s was an attempt to remap the Win16 based OS into the Win32 design. And it actually worked.
WIn95 is a 32 bits GUI for MS-DOS
That's actually not true. Once Win95 started, DOS was kicked off into oblivion. The reason was that DOS couldn't actually handle protected mode, so once the CPU mode was changed DOS was effectively dead memory. Win95 actually emulated DOS inside of itself for old applications. The only reason why DOS was kept around is that it provided a convenient boot loader for Win95.
Yes [a separate Win32 API] was the case for MS-DOS+win (more or less - see above.) But this is not the case for WinNt.
Bullocks. The NT kernel and the Win32 API are separate, though closely related, entities. The kernel provides basic driver and I/O services which the Win32 API then remaps into a useful usermode API.
And taking into account that the crash of the GUI always crashes the OS your statement seems to be false.
More nonsense. This merely shows a lack of understanding of kernel design. There are two ways of producing graphics in an OS with memory protection and a clear separation between kernel and usermode. The slow way is to make most of the graphics processing happen in usermode, then jump to kernel space to perform the I/O to the video card. This is the safest route, but requires a lot of jumping in and out of kernel space. The alternative is to suck the entire graphics subsystem into a kernel mode driver. Since all processing happens inside kernel space, the graphics subsystem runs much faster. Unfortunately, any error generated in this space (e.g. GPF, buffer overflow, etc.) would be impossible to recover from without risking damage to the kernel's memory image.
FYI, NT 3.51 used the former method. This caused it to generally be more stable than 4.0 which switched to using the later method for performance. This tended to make 4.0 somewhat unstable for graphics intensive apps. (I used to kill it with GIMP all the time.) Win2K added additional checks that made many of these issues go away.
Ok, bad news. The "architect" has just confirmed it to be a prank. He did mention that it wasn't far fetched, though, and what they really need is the source code to the Internix(STU) project to make it happen. He's speculated that it may be possible to get Microsoft to "Shared Source" the library. :-) /ME stops happy dance and sighs
This is a cheap trick to disguise advertising as a "joke". You're supposed to go to ThinkGeek, see the "funny" products (haha), then notice that you can get a free TShirt if you buy $10. Offer good today only! So then like a sheep, you go and buy stuff that you otherwise couldn't quite justify.
Well, I've emailed the architect to get the real story. With any luck, we'll know the truth shortly.
But the new Trek producers can cut corners by utilizing existing props and sets from "Team America." So don't be surprised if Captain Archer looks suspiciously like Alec Baldwin, Trip Tucker like Matt Damon, Travis Mayweather like Samuel L. Jackson, T'Pol like Janeane Garofalo, and Malcolm Reed like Susan Sarandon.
:-D
That's the funniest shit I've seem all day! Kudos to StarTrek.com!
1. No, I'm not joking.
2. Look at their last modified date on the website. 2 days ago.
3. That's one hell of a lot of effort for just a joke.