Perhaps I'm misreading the Netscape Article. I hope so. What do they mean by "a glitch prevented the use of all 65000 colors", when later on they talk about a 12-bit controller? Sounds to me like they got caught lying, never expecting consumers to notice. That's why I hope I'm wrong.
Accidentally designing all of your hardware around a 12-bit controller is not a glitch. Claiming a 12-bit controller can produce 65000 colors cannot entirely be an accident.
The only real advantage I've been able to see for design of this robot, is that it's already proven (it seems). If you made a specialized robot, there might be more uncertainty about reliability.
Careful with the POWER versus PowerPC, as they're not the same thing. The POWER architecture is really fast, even at low clocks. I remember that SpecFP used to have a 200MHz POWER3 chip in second place after the 500MHz Alpha 21264, ahead of the PA-RISC and everyone else. I think they've at least doubled the clocks by now, and I think they're in the POWER4 generation as well. They had a 64-bit version, the RS-II 64, and I believe they've also moved that ahead to the RS-III 64.
But I've never been able to really keep track of IBM's processors, it just seems tough to get info and keep it straight. Especially with the POWER versus PowerPC mixup.
In some numerical code (photometric redshift estimation), the Alpha 21264 consistently moves at least 3 times faster, but most often is closer to 5 times faster than similarly clocked P-IIIs.
Alphas really are damn fast, not just the chip but the whole architecture. If you're the kind of person that overbuilds everything you create (like me), then the Alpha will make you smile.
IIRC, Microsoft recommends giving the principal user of a machine root priviliges, in case the root password is forgotten. I read this on the MS website. Sort of an intersting suggestion, given the rest of this discussion.
"What RMS is really talking about is mandatory disclosure and a fixed $0.00 price tag on software."
I don't see where you get this from Stallman, and I don't think he'd agree with it. Mandatory disclosure, possibly. The right to reverse engineer and share, certainly. But a fixed $0.00 price tag has never been part of his philosophy, as far as I can tell.
These are subtle issues, so we need to be really careful when discussing them. Where RMS says, roughly, "if someone rejects my values, there's really nothing we can argue about," he is roughly describing any axiomatic system--for instance mathematics or formal logic, where an axiomatic system is used for clarity (well, some people think it is a religion, sullied by human issues such as clarity). I think he had such a well-thought-out answer for the 'values' bit because he is a careful, thoughtful, deliberate person.
So--mandatory disclosure doesn't jive with freedom, so I don't believe he'd advocate that exactly (IANRMS?). After all, the GPL doesn't propose to force full disclosure for Microsoft's code, unless this code is derived from GPL'd code (which is where copyright law comes into play, deciding what is a derivative, and whether the copyright holder has any power over the derivative). My right to analyze and share publicly distributed information, whether or not I paid to receive that information, is very much about my freedom. And the price bit is a red herring--the GPL doesn't require 'gratis' software.
RMS is trying to analyze and solve a really hard problem, accomodating freedom within a social structure. I don't think anyone has really solved this yet, especially the USA with it's rather inhumane economy (I am a US citizen, and I feel I can adequately justify this statements).
1) If you're looking for speed, ditch the x86 architecture all together. Outside that, try Evans and Sutherland. Sure it's closed source, but this is a company that has helped people get their _work_ done for years.
In response to
"Yes I do place that level of trust in nVidia."
I think this trust is misplaced. I think it is fair to say that it wasn't genius-level inspiration that allowed nVidia to double the performance of their Win32 TNT drivers between driver versions. The performance could double because the first drivers stunk.
I don't think you have any good reason to really trust nVidia. I've never heard anyone say anything good about their ability to write drivers, outside of your comments. Besides, what is there to place your trust in--that on your particular hardware, during a relatively short space of time, your Win32 machine didn't crash (very often) for the small percentage of commercial software that you were running? How many of those software vendors had to rewrite their code to work around bugs in nVidia's drivers? Will we ever know? Did you write your own code that targetted these drivers (indirectly)?
There really isn't any reason to trust most commercial vendors, especially publicly-owned companies. This is just a part of capitalism. Don't kid yourself, Corporate America doesn't have your best interests at heart.
When in Kauaii, I just looked for a local copy shop that would let me install the drivers for my camera on one of their computers. Found one, no problem. I had brought a zip disk with me, and the shop of course had zip drives available. Worked out fine.
I wouldn't bother asking at Kinkos, though. Look for something independent.
I think it is clear that your choice in cameras depends on several quantifiable aspects, like price and storage medium, and harder-to-quantify aspects like battery life (this varies widely between users) and image quality.
The two best review sites I've seen are the Digital Camera Resource and the Imaging Resource. The former has shorter reviews and better reader comments, while the latter has very comprehensive reviews and image comparisons.
I typically use the Digital Camera resource to narrow down the search, and then use the Imaging Resource to compare image quality. Imaging Resrouce also measures how long between pressing the shutter release and when the picture is taken, something I think is important to consider if you photograph moving subjects.
I've bought/helped buy 4 cameras in the last couple years. For myself I have a Kodak DC-260, and I've suggested one Nikon Coolpix 950, a Kodak DC-215, and a Kodak DC-280. The Kodak cameras I've mentioned here all have amazing image quality, and the DC-215 is rather inexpensive as it is old and produces 1152x864 images. The Nikon also takes great pictures, and has a lot of features. I think all of these cameras work fine with linux, my DC-260 certainly does. These cameras were bought at different times with different people's needs in mind. A friend has an Olympus D-340L, which takes good pictures (though it's color saturation isn't as good as the Kodaks') and is fabulous indoors without a flash (something the Kodaks aren't very good at). Note that the Kodak's aren't known for rapid picture-taking ability.
The Casio 2000UX lost to the DC-280 on image quality, but it had an interesting feature set. The Coolpix 950 beat the Kodak DC-265 in features, and had comparable image quality. The DC-215 beat the comparable Olympus cameras on price, and especially beat the Olympus D-400, which has complaints about lack of lavender hues. I've never liked the image quality on Sony's older cameras, and I think the floppy disks are too small or else the lossy image compression is too aggressive (don't know which).
Memory format isn't that big of an issue, and neither is USB capability, because of dedicated card readers. Time between shutter release and the picture being taken will affect every picture you take. Low-light ability may be important to you. Image quality reigns supreme for me, and is an area where Kodak does well (except perhaps the DC-240). Point-and-click versus configurable f-stops, etc, will make a difference for some people. And if you've got $5000 to blow, check out the high end 6 megapixel SLR digital cameras from Kodak!
One last bit of advice--try to get a 'satisfaction guaranteed' return policy. If the camera's pictures come out a little to red, is that a defect? Better safe than sorry.
Incidently, parodies such as those done by Weird Al Yankovic don't really need permission. Parodies have long been protected by law from recrimination by the original authors. And the world is better off for it.
Also, I think the Slashdot editors' attitudes can be just as easily thought of as 'defensive' instead of condescending and righteous. They've been attacked in a big way. Also, they've never really acted condescending in the past (occasionaly righteous, perhaps).
You write "But here is something you CAN quote me on" followed by "I do not give permission to any entity to use my post..."
So which is it? And where does the Slashdot forum end? Can I tell my wife what you posted? Can I print your post and put it on my fridge? What if I'm discussing Slashdot with a friend over breakfast, is that part of the Slashdot forum?
I'm starting to think that much of the trolling and other stupid posts come from people who believe that Slashdot protects their idiocy from being seen. When you look like an idiot on Slashdot, your idiocy is broadcast with a speed and scope heretofore unimaginable. And Slashdot is a more-or-less permanent record, too!
`Engage brain before opening mouth' applies to public posts. For instance, it applies to posting on Slashdot!
1) Don't post to Slashdot if you don't want people to read your post. The Slashdot community doesn't end at the list of user accounts. This book extends the Slashdot community, using a new media for Slashdot -- print.
2) Perhaps we better appreciate why RMS dislikes the term Intellectual Property, two words which he believes don't belong together. Comment ownership is a really strange idea--once you say something verbally, or post it on Slashdot, it's out in the open. So how can you 'own' it? In particular, I think the 'comment ownership' clause should be read as "don't sue Slashdot because of something you read in a post".
I think the above question is very interesting. In particular, what questions does SCO want answered when determining their business direction? What answers did they receive that made Linux attractive?
I'm curious because I'm unsure how to interpret Gartner Group (for example) research I read on/., The Register, CNET, etc.. Too often, this 'research' is made public by the company which paid for the research. We can all imagine the conflicts of interest here, especially after reading some of the Gartner Group's fine print at the bottom of the page. But if a (sensible) company is determining its future direction, they (hopefully) want the most brutally honest research they can get.
try a 'man loadkeys', and look at the bottom for related utilities. or try 'man -k keyboard' and sift through the entries.
On my RedHat 6.1 system, keyboard-related files are under/usr/lib/kbd/keymaps. There's all kinds of goodies there. I looked into this because I wanted right-alt + f2 to go to virtual console two, not just left-alt + f2.
I expect that the 'UNIX' tools you used under NT were the GNU tools. Had you used the GNU tools on Solaris and HP, you'd be better off for compatibility.
Not to mention that HP breaks just about all code of any sort that works on other UNIXs. HPUX is my worst nightmare, I'm not sure that I wouldn't rather use the Win95 CLI. At least I would expect Win95's CLI to be useless ahead of time.
HP can make some darn good chips, but the sooner they ditch HSUX, er HPUX, the better. I may even be interested in their hardware again once the Puffin group gets Linux going on the PA.
I can't believe I didn't see this while scanning the comments--maybe I missed it. A huge part of UNIX is the filesystem. The UNIX filesystem has got to be among the most beautiful device abstraction layers ever built. I don't need to know how many drives are in the machine--or I can add extra drives, change the physical location of data, and keep the same paths and filenames.
I can access my printer as a file, my serial ports as a file, my memory as a file. Heck, I can mkfs/dev/ram and make filesystems in files on my filesystem. Then there are links, especially symlinks. And networked filesystem types that look local to the user (slow with lots of latency, and some race conditions, but local). You can even access the kernel via the filesystem.
NTFS and VFAT don't have symlinks and the hardware is up-front and ugly (ever try moving an app from C: to D:?). I'm sure there's a lot more, but these things alone are horrible. What happened to device abstraction, that I have to track my hardware when naming files?
At one time, not too long ago, the Power 3 architecture was rated (by some) as the second fasted floating point to the Alpha 21264 500MHz. The punchline is that the Power chip was running at 200 MHz!
In the past, complications with multiprocessor computers has prevented their supremacy of single cpu architectures. I'd love to see IBM succeed with their multicpu chips, as I believe this technology may solve the nagging parallel problems with processor interconnect. And the Power architecture is very nice.
Does anyone know if the PowerPC and Power architectures will finally become one with this product, as was expected with previous Power revisions? Somehow, I really don't expect to see it ever happen, with the way Motorola and IBM have gotten along.
I think you are glossing over years of interesting history and computer architecture. IBM's mainframe division might have lost mililions because of PCs, but not because PCs were superior computing platforms for big data.
PCs won for the same reason that we have traffic problems on our highways--everybody wants to be a driver and doesn't want to share. This was, it seems, part of the rationale for the federally funded development of the internet (well, ARPANET): getting scientists to share computing resources bought with federal money (c.f. A History of Modern Computing, Ceruzzi 1998, p 296). This is of course the same phenomenon that drove minicomputers, which were also replaced by PCs. I'm not saying this is bad, well, not in the case of computing anyway.
But PCs still suck at any number of computing tasks, and aren't really improving in areas that can't be mass marketed. That's why my lab bought a very expensive dual Alpha machine instead of spending that money on the 5-to-20 equivalently clocked P-IIIs (these numbers come from real computations). Not to mention a farm of PCs can only handle embarassingly parallel computations at the same speed has the Alpha, and require more programming effort than the Alpha. The Alpha isn't even close to a mainframe, either.
And I haven't even gotten to bandwidth issues that sponsered this thread (well, they're part of the 5-to-20 figure above, in some ways). IBM lost on mainframes because they dealt _only_ with mainframes. DEC lost with minicomputers because they too were arrogant/ignorant about PCs. And while Intel seems to acknowledge the information appliance ideas, they're x86 tech will only go so far (we can hope, can't we?). But just as information appliances aren't the best choice for PC-type tasks, PCs aren't aren't the best choice for mainframe-sized loads.
Oversimplification is a marketing tool. It has no place in intelligent discussion, where flippant remarks are better replaced by _questions_.
Newspapers try to be everything to everyone, and of course fail. I assume that this least-common-denominator approach explains the lack of depth, uninteresting writing, and generic selection of stories.
I do think that the comics section is still pretty okay, though. I don't know what Katz was talking about there. This is nearly a 'killer app' for the Sunday newspaper, along with the weekend ads.
But I can find what I want, when I want online. And I can email the author with questions or criticisms. It's niche marketing again, just like John Katz with his Mountain book. I take an active role in finding products, including information, rather than hoping in vain it will come to me. Cheaper for the publisher, better for me. Furthermore, I can easily archive the interesting bits. As soon as information is relegated to print, it dies. It's too hard to mess with printed data.
We still need reporters and editors. There is a profound lack of strong journalism in tech areas on the web. This applies to both writing ability and verification of facts. But these areas are quickly improving.
The best reason to make a transition away from printed words soon is environmental. The points above are fringe benefits. Our world would be better off with less printed material. If you recycle newspapers, you know how fast they accumulate. Now multiply that by the NYT subscription base, as well as all the other newspapers. This is one of the easiest ways to make a huge improvement in resource consumption and waste reduction, with few (if any) real drawbacks.
In the US, universities are becoming extensions of high school. However, not everyone needs a 4-year liberal education in order to discover and specialize their niche. In fact, I'll bet most people _don't_ want/need a 4-year degree.
Vocational schools are often disdained in the US. However, I believe that many people who have a strong non-research interest would be best served by vocational schools. Learn what you need, and quickly go to work on your career.
This unfortunate situation has persisted for many years. Hopefully the new economy, which is encouraging technically-oriented students to bail out of universities, will encourage the acceptance of job-specific vocational schools.
Insane people like me can stick it out in the universities for the rest of our lives, to be owned by the rich people who bailed. That's why only insane people consider a life of research.
I'd like to emphasize the memory effect. Cache performance in SMP architectures can have a huge impact on performance. For larger processor counts in shared memory machines, things like main memory latency, size and location are similarly important. The interconnect has to handle these memory requests too, as well as coordination messages. It probably isn't worth parallelizing on an SMP machine if you can't get good cache behavior.
Clusters are shared-nothing architectures, great when you don't want to share anything. The SGI, Cray (I know), IBM, et. al, supercomputers are meant for problems that are hard to parallelize. They cost a lot because sharing resources is _hard_ to do well.
We aren't getting something for free with Beowulf and other clusters. What we're doing is more niche marketing, just like John Katz with his book announcements on/.. Clusters are helping people with a particular kind of parallel problem, and they arrived because appropriate hardware and software has been commoditized. In particular, high speed networking and cheap `nodes' (in this case, whole PCs), as well as Free (in both senses) software.
Slashdot Mirror
Perhaps I'm misreading the Netscape Article. I hope so. What do they mean by "a glitch prevented the use of all 65000 colors", when later on they talk about a 12-bit controller? Sounds to me like they got caught lying, never expecting consumers to notice. That's why I hope I'm wrong.
Accidentally designing all of your hardware around a 12-bit controller is not a glitch. Claiming a 12-bit controller can produce 65000 colors cannot entirely be an accident.
-Paul Komarek
The only real advantage I've been able to see for design of this robot, is that it's already proven (it seems). If you made a specialized robot, there might be more uncertainty about reliability.
-Paul Komarek
Careful with the POWER versus PowerPC, as they're not the same thing. The POWER architecture is really fast, even at low clocks. I remember that SpecFP used to have a 200MHz POWER3 chip in second place after the 500MHz Alpha 21264, ahead of the PA-RISC and everyone else. I think they've at least doubled the clocks by now, and I think they're in the POWER4 generation as well. They had a 64-bit version, the RS-II 64, and I believe they've also moved that ahead to the RS-III 64.
But I've never been able to really keep track of IBM's processors, it just seems tough to get info and keep it straight. Especially with the POWER versus PowerPC mixup.
-Paul Komarek
In some numerical code (photometric redshift estimation), the Alpha 21264 consistently moves at least 3 times faster, but most often is closer to 5 times faster than similarly clocked P-IIIs.
Alphas really are damn fast, not just the chip but the whole architecture. If you're the kind of person that overbuilds everything you create (like me), then the Alpha will make you smile.
=-) =-) (we have a dual alpha)
-Paul
I think a little precision is in order on this contentious issue. They didn't say "our content", they said "public content". Quite a big difference.
-Paul Komarek
IIRC, Microsoft recommends giving the principal user of a machine root priviliges, in case the root password is forgotten. I read this on the MS website. Sort of an intersting suggestion, given the rest of this discussion.
-Paul Komarek
"What RMS is really talking about is mandatory disclosure and a fixed $0.00 price tag on software."
I don't see where you get this from Stallman, and I don't think he'd agree with it. Mandatory disclosure, possibly. The right to reverse engineer and share, certainly. But a fixed $0.00 price tag has never been part of his philosophy, as far as I can tell.
These are subtle issues, so we need to be really careful when discussing them. Where RMS says, roughly, "if someone rejects my values, there's really nothing we can argue about," he is roughly describing any axiomatic system--for instance mathematics or formal logic, where an axiomatic system is used for clarity (well, some people think it is a religion, sullied by human issues such as clarity). I think he had such a well-thought-out answer for the 'values' bit because he is a careful, thoughtful, deliberate person.
So--mandatory disclosure doesn't jive with freedom, so I don't believe he'd advocate that exactly (IANRMS?). After all, the GPL doesn't propose to force full disclosure for Microsoft's code, unless this code is derived from GPL'd code (which is where copyright law comes into play, deciding what is a derivative, and whether the copyright holder has any power over the derivative). My right to analyze and share publicly distributed information, whether or not I paid to receive that information, is very much about my freedom. And the price bit is a red herring--the GPL doesn't require 'gratis' software.
RMS is trying to analyze and solve a really hard problem, accomodating freedom within a social structure. I don't think anyone has really solved this yet, especially the USA with it's rather inhumane economy (I am a US citizen, and I feel I can adequately justify this statements).
-Paul Komarek
1) If you're looking for speed, ditch the x86 architecture all together. Outside that, try Evans and Sutherland. Sure it's closed source, but this is a company that has helped people get their _work_ done for years.
In response to
"Yes I do place that level of trust in nVidia."
I think this trust is misplaced. I think it is fair to say that it wasn't genius-level inspiration that allowed nVidia to double the performance of their Win32 TNT drivers between driver versions. The performance could double because the first drivers stunk.
I don't think you have any good reason to really trust nVidia. I've never heard anyone say anything good about their ability to write drivers, outside of your comments. Besides, what is there to place your trust in--that on your particular hardware, during a relatively short space of time, your Win32 machine didn't crash (very often) for the small percentage of commercial software that you were running? How many of those software vendors had to rewrite their code to work around bugs in nVidia's drivers? Will we ever know? Did you write your own code that targetted these drivers (indirectly)?
There really isn't any reason to trust most commercial vendors, especially publicly-owned companies. This is just a part of capitalism. Don't kid yourself, Corporate America doesn't have your best interests at heart.
-Paul Komarek
When in Kauaii, I just looked for a local copy shop that would let me install the drivers for my camera on one of their computers. Found one, no problem. I had brought a zip disk with me, and the shop of course had zip drives available. Worked out fine.
I wouldn't bother asking at Kinkos, though. Look for something independent.
-Paul
I think it is clear that your choice in cameras depends on several quantifiable aspects, like price and storage medium, and harder-to-quantify aspects like battery life (this varies widely between users) and image quality.
The two best review sites I've seen are the Digital Camera Resource and the Imaging Resource. The former has shorter reviews and better reader comments, while the latter has very comprehensive reviews and image comparisons.
I typically use the Digital Camera resource to narrow down the search, and then use the Imaging Resource to compare image quality. Imaging Resrouce also measures how long between pressing the shutter release and when the picture is taken, something I think is important to consider if you photograph moving subjects.
I've bought/helped buy 4 cameras in the last couple years. For myself I have a Kodak DC-260, and I've suggested one Nikon Coolpix 950, a Kodak DC-215, and a Kodak DC-280. The Kodak cameras I've mentioned here all have amazing image quality, and the DC-215 is rather inexpensive as it is old and produces 1152x864 images. The Nikon also takes great pictures, and has a lot of features. I think all of these cameras work fine with linux, my DC-260 certainly does. These cameras were bought at different times with different people's needs in mind. A friend has an Olympus D-340L, which takes good pictures (though it's color saturation isn't as good as the Kodaks') and is fabulous indoors without a flash (something the Kodaks aren't very good at). Note that the Kodak's aren't known for rapid picture-taking ability.
The Casio 2000UX lost to the DC-280 on image quality, but it had an interesting feature set. The Coolpix 950 beat the Kodak DC-265 in features, and had comparable image quality. The DC-215 beat the comparable Olympus cameras on price, and especially beat the Olympus D-400, which has complaints about lack of lavender hues. I've never liked the image quality on Sony's older cameras, and I think the floppy disks are too small or else the lossy image compression is too aggressive (don't know which).
Memory format isn't that big of an issue, and neither is USB capability, because of dedicated card readers. Time between shutter release and the picture being taken will affect every picture you take. Low-light ability may be important to you. Image quality reigns supreme for me, and is an area where Kodak does well (except perhaps the DC-240). Point-and-click versus configurable f-stops, etc, will make a difference for some people. And if you've got $5000 to blow, check out the high end 6 megapixel SLR digital cameras from Kodak!
One last bit of advice--try to get a 'satisfaction guaranteed' return policy. If the camera's pictures come out a little to red, is that a defect? Better safe than sorry.
-Paul KomarekIncidently, parodies such as those done by Weird Al Yankovic don't really need permission. Parodies have long been protected by law from recrimination by the original authors. And the world is better off for it.
Also, I think the Slashdot editors' attitudes can be just as easily thought of as 'defensive' instead of condescending and righteous. They've been attacked in a big way. Also, they've never really acted condescending in the past (occasionaly righteous, perhaps).
-Paul Komarek
You write
"But here is something you CAN quote me on"
followed by
"I do not give permission to any entity to use my post..."
So which is it? And where does the Slashdot forum end? Can I tell my wife what you posted? Can I print your post and put it on my fridge? What if I'm discussing Slashdot with a friend over breakfast, is that part of the Slashdot forum?
I'm starting to think that much of the trolling and other stupid posts come from people who believe that Slashdot protects their idiocy from being seen. When you look like an idiot on Slashdot, your idiocy is broadcast with a speed and scope heretofore unimaginable. And Slashdot is a more-or-less permanent record, too!
`Engage brain before opening mouth' applies to public posts. For instance, it applies to posting on Slashdot!
-Paul Komarek
1) Don't post to Slashdot if you don't want people to read your post. The Slashdot community doesn't end at the list of user accounts. This book extends the Slashdot community, using a new media for Slashdot -- print.
2) Perhaps we better appreciate why RMS dislikes the term Intellectual Property, two words which he believes don't belong together. Comment ownership is a really strange idea--once you say something verbally, or post it on Slashdot, it's out in the open. So how can you 'own' it? In particular, I think the 'comment ownership' clause should be read as "don't sue Slashdot because of something you read in a post".
-Paul Komarek
I think the above question is very interesting. In particular, what questions does SCO want answered when determining their business direction? What answers did they receive that made Linux attractive?
/., The Register, CNET, etc.. Too often, this 'research' is made public by the company which paid for the research. We can all imagine the conflicts of interest here, especially after reading some of the Gartner Group's fine print at the bottom of the page. But if a (sensible) company is determining its future direction, they (hopefully) want the most brutally honest research they can get.
I'm curious because I'm unsure how to interpret Gartner Group (for example) research I read on
-Paul Komarek
I'd like to see this question answered. I'm writing a reply since I'm not moderating!
-Paul Komarek
I don't believe redshift is available for the
majority of the survey.
-Paul Komarek
try a 'man loadkeys', and look at the bottom for related utilities. or try 'man -k keyboard' and sift through the entries.
/usr/lib/kbd/keymaps. There's all kinds of goodies there. I looked into this because I wanted right-alt + f2 to go to virtual console two, not just left-alt + f2.
On my RedHat 6.1 system, keyboard-related files are under
-Paul Komarek
I expect that the 'UNIX' tools you used under NT were the GNU tools. Had you used the GNU tools on Solaris and HP, you'd be better off for compatibility.
Not to mention that HP breaks just about all code of any sort that works on other UNIXs. HPUX is my worst nightmare, I'm not sure that I wouldn't rather use the Win95 CLI. At least I would expect Win95's CLI to be useless ahead of time.
HP can make some darn good chips, but the sooner they ditch HSUX, er HPUX, the better. I may even be interested in their hardware again once the Puffin group gets Linux going on the PA.
I can't believe I didn't see this while scanning the comments--maybe I missed it. A huge part of UNIX is the filesystem. The UNIX filesystem has got to be among the most beautiful device abstraction layers ever built. I don't need to know how many drives are in the machine--or I can add extra drives, change the physical location of data, and keep the same paths and filenames.
/dev/ram and make filesystems in files on my filesystem. Then there are links, especially symlinks. And networked filesystem types that look local to the user (slow with lots of latency, and some race conditions, but local). You can even access the kernel via the filesystem.
I can access my printer as a file, my serial ports as a file, my memory as a file. Heck, I can mkfs
NTFS and VFAT don't have symlinks and the hardware is up-front and ugly (ever try moving an app from C: to D:?). I'm sure there's a lot more, but these things alone are horrible. What happened to device abstraction, that I have to track my hardware when naming files?
-Paul Komarek
Heck, I just noticed it was the bus running at 500MHz, with the cpu better than double this. Now I'm really impressed!
-Paul Komarek
At one time, not too long ago, the Power 3 architecture was rated (by some) as the second fasted floating point to the Alpha 21264 500MHz. The punchline is that the Power chip was running at 200 MHz!
In the past, complications with multiprocessor computers has prevented their supremacy of single cpu architectures. I'd love to see IBM succeed with their multicpu chips, as I believe this technology may solve the nagging parallel problems with processor interconnect. And the Power architecture is very nice.
Does anyone know if the PowerPC and Power architectures will finally become one with this product, as was expected with previous Power revisions? Somehow, I really don't expect to see it ever happen, with the way Motorola and IBM have gotten along.
I think you are glossing over years of interesting history and computer architecture. IBM's mainframe division might have lost mililions because of PCs, but not because PCs were superior computing platforms for big data.
PCs won for the same reason that we have traffic problems on our highways--everybody wants to be a driver and doesn't want to share. This was, it seems, part of the rationale for the federally funded development of the internet (well, ARPANET): getting scientists to share computing resources bought with federal money (c.f. A History of Modern Computing, Ceruzzi 1998, p 296). This is of course the same phenomenon that drove minicomputers, which were also replaced by PCs. I'm not saying this is bad, well, not in the case of computing anyway.
But PCs still suck at any number of computing tasks, and aren't really improving in areas that can't be mass marketed. That's why my lab bought a very expensive dual Alpha machine instead of spending that money on the 5-to-20 equivalently clocked P-IIIs (these numbers come from real computations). Not to mention a farm of PCs can only handle embarassingly parallel computations at the same speed has the Alpha, and require more programming effort than the Alpha. The Alpha isn't even close to a mainframe, either.
And I haven't even gotten to bandwidth issues that sponsered this thread (well, they're part of the 5-to-20 figure above, in some ways). IBM lost on mainframes because they dealt _only_ with mainframes. DEC lost with minicomputers because they too were arrogant/ignorant about PCs. And while Intel seems to acknowledge the information appliance ideas, they're x86 tech will only go so far (we can hope, can't we?). But just as information appliances aren't the best choice for PC-type tasks, PCs aren't aren't the best choice for mainframe-sized loads.
Oversimplification is a marketing tool. It has no place in intelligent discussion, where flippant remarks are better replaced by _questions_.
Newspapers try to be everything to everyone, and of course fail. I assume that this least-common-denominator approach explains the lack of depth, uninteresting writing, and generic selection of stories.
I do think that the comics section is still pretty okay, though. I don't know what Katz was talking about there. This is nearly a 'killer app' for the Sunday newspaper, along with the weekend ads.
But I can find what I want, when I want online. And I can email the author with questions or criticisms. It's niche marketing again, just like John Katz with his Mountain book. I take an active role in finding products, including information, rather than hoping in vain it will come to me. Cheaper for the publisher, better for me. Furthermore, I can easily archive the interesting bits. As soon as information is relegated to print, it dies. It's too hard to mess with printed data.
We still need reporters and editors. There is a profound lack of strong journalism in tech areas on the web. This applies to both writing ability and verification of facts. But these areas are quickly improving.
The best reason to make a transition away from printed words soon is environmental. The points above are fringe benefits. Our world would be better off with less printed material. If you recycle newspapers, you know how fast they accumulate. Now multiply that by the NYT subscription base, as well as all the other newspapers. This is one of the easiest ways to make a huge improvement in resource consumption and waste reduction, with few (if any) real drawbacks.
We still need reporters and editors.
In the US, universities are becoming extensions of high school. However, not everyone needs a 4-year liberal education in order to discover and specialize their niche. In fact, I'll bet most people _don't_ want/need a 4-year degree.
Vocational schools are often disdained in the US. However, I believe that many people who have a strong non-research interest would be best served by vocational schools. Learn what you need, and quickly go to work on your career.
This unfortunate situation has persisted for many years. Hopefully the new economy, which is encouraging technically-oriented students to bail out of universities, will encourage the acceptance of job-specific vocational schools.
Insane people like me can stick it out in the universities for the rest of our lives, to be owned by the rich people who bailed. That's why only insane people consider a life of research.
-Paul Komarek
I'd like to emphasize the memory effect. Cache performance in SMP architectures can have a huge impact on performance. For larger processor counts in shared memory machines, things like main memory latency, size and location are similarly important. The interconnect has to handle these memory requests too, as well as coordination messages. It probably isn't worth parallelizing on an SMP machine if you can't get good cache behavior.
/.. Clusters are helping people with a particular kind of parallel problem, and they arrived because appropriate hardware and software has been commoditized. In particular, high speed networking and cheap `nodes' (in this case, whole PCs), as well as Free (in both senses) software.
Clusters are shared-nothing architectures, great when you don't want to share anything. The SGI, Cray (I know), IBM, et. al, supercomputers are meant for problems that are hard to parallelize. They cost a lot because sharing resources is _hard_ to do well.
We aren't getting something for free with Beowulf and other clusters. What we're doing is more niche marketing, just like John Katz with his book announcements on
-Paul Komarek