"The Market" only works if buyers have a choise. The monopoly laws were passed in responce to J Paul Getrty's oild company at about the turn of the last century. WHat he did was if there was a shop selling oil he'd open a store right next to it and sell oil at a _much_ lower price. He's loose a ton of money but he'd put the old store out of bussiness. Once the first place was gone he's ise the price way up, higher then it was before he came to that town.
In my opinion Microsoft is only a monopoly because people are just plain stupid. People could have bought Apple, Linux or one of the others (SGI's Irix was one of the best systems I've ever used. Next was a lot like Mac OSX) But the stupis computer buys looked only at the initial purchace price and went for the botom.
Mr. Gates had one billient bit of insite that made him a bilionare. "There are more people who know absolutly nothing about coputers than there are computer experts. Why not sell the that larger group" and he did. If you think about it and are old enough to remember the pre-microsoft world. Cmputer marketing people targeted only compter professionals. Gates broke out of that and went after group that could read price tags did understand much else. Remember he got into bussines computing be puting the boxes in the home first.
No, I figure you'd use something like "shutdown -h now" or "init 0" to exit. The creen saver stops when the process terminates. The big news here is simply that someone got Linux to run as a Windows userland process. OK not 100% new but then they thought to call it a "screen saver". It's cool to be the first to think of something.
In a recent test, researchers attemped to drain the Pacific Ocean by caryying buckets loaded with seawater to the Arizona or Utah desert where water was dumpped on the dry sand. The ive gallon buckets were decleared 100% effective. Each bucket tested did in fact carry up to five gallons of water, exactly as designed. Due to the success of the test program plans are being made to procure 25 more buckets for use in 2006
THe title is completely misleading. Right in the summary it says.."decoded the mitochondrial DNA of..." mitochondrial DNA is a very small repcentage of the total.
Whouldn't it be great if slashdot implemented some kind of bullshit filter.
Sounds like you have come to a sound conclusion. The junk you have collected is just that junk. But it is "junk" because to __collected__ it. It was not ever really of value. But what if you had __created__ it. Let's say you are a photographer and dumb enough to keep all your work on one disk drive, 10 years worth of work. Or let's make it worse. Let's say you are running a bussenis and all you customer data, inventory and pending orders are of the disk? My point is that some people do have valuable and irreplaceable data even if many people have only "junk". If you spend much of your time writing, drawing, shotting video an editing then your data has value but if you are simply making a local copy of data you find on the Internet it diferent.
I can put a $$ value on my data easy. I spend maybe 2,000 hours a year writing and the company I work for spends about $120/hour to keep me here (no I don't get that much, they need to pay rent, for power and to keep the reastrooms) clean. So by definition I create about $240K worth of data per year, figure that there are about 2,000 people
like me that work here.... The cost of the pysical media or drives is trivial
THe rule is (or should be) that if you care about data it needs to exist in at least three places so that when one copy dies you still have some redundent storage. No matter how you store it, it wil eventually become unreadable.
100 years from now my kids grandchildren might want to see some old photos of when thier grandparents where 12 years old. THose photos are all digial now. In a few years only a few fine arts photographers and hobbiest will use film.
"Come on folks, this is junior high biology....." Yes it is but there is a high possability that many of the people posting these misinformed comments have not yet entered junior high school. After all kids are all on Winter break now. You really can't blame a 10 year old for not understanding 8th grade science.
It's like saying "what if Issac Newton were never born. Likely 20 years later a group of other people would have figured out clasical physics. With Calculus being invented at that time it was only a matter of a few years before it all fell into place. Same with Darwin. Biologists would have figured it out, but later.... What if there was life that beagal would have discovered had it not crashed? Same thing If it's there we find it later. Maybe 20 yers later but it it's there we find it.
"triangulation on that signal" Yes exactly. But I think that has already been done. The way to do it is to move the reciever and measure the direction to the signal. As luck would have it the reciever is mounted on a platform that is moving very fast and overs huge distances. The Earth is in orbit around the sun and it is spining too. So if you wait 12 hours your reciever has moved one Earth diameter. If you wait a few months it moves even further.
Faking the signal would require someone to place a transmitter in the line of sight between your antenna and mars. Mars moves in the sky in a complex way.
"how much more energy goes into refining the metal for the additional engine?" One way the figure this out is to look at the wholesale price of the metal. If steel sells for 60 cents a pound you might guess that less than 60 cents worth the energy was used to mine, transport and refine the steel.
I'd also asume this "steem" engine" could last the life of the car like and automatic transision. They can pick the fluids so they are non-corosive. There would be no cumbustion products inside
COmpared to electric it saves a lot. No battery so it is lighter and maybe even cheaper. And you can use a smaler gas engine and recover some cost by loosing a couple cylinders there.
People want quick of the line short term acceleration. Storing energy as steam could do that. Also electric hybred does not help at all for long term freeway driving. Steam does.
It happens a lot. Someone who knows very little about computers or software thinks "If I just buy one more gadget or one ome program then I can do XXX" Same with people who don't understand how to fish "I I can only get that extrat fancy reel then I'll catch fish". Or the wantabe photographer who thinks if only he had a better camera he could take beter pictures.
I reality a true artist could take carcal from the backyard BBQ and draw on sidewalk something that others might like while tho wantabe artist has a closet of supplies and produces junk.
Marine VHF. It's the common medium range (up to about 25 miles) radio found on all boats. Other bands might be closer to the 107-148 range but none would have transmitters closer to the animals.
Of course this is silly. Norwall tusks are not antenna
Re:you Do need to have a cut off date
on
Merck's Deleted Data
·
· Score: 2, Insightful
No. I'm saying you need to selct a cut off date. How you choose it depends on many factors. They did set a date, we don't know how the picked it but we do know they picked the date before the study started, No one is claimming it wa picked after the fact to hide lateer data. So far nothing wrong. The controversy is what to do with late post cutoff data. The technical legalistic answer was "We are reporting on a study, not on what hapend after a study" I think anything that happened after should have gone into some appendix simply for public relations or polical reasons to revent this kind of public fallout.
I'm not arguing one side ot the other. My point is simply that there are two points of view and it is not so black and white.
Let's say we do a small study. We let 100 people drink 8 oz. of name brand bottled water. Now if we find that 50 of them are dead seconds afterward we might conclude something is wrong with the water. But if all them are dead 200 years after drinking the water we don't conclude anyhting about the water. So at what point between two seconds and 200 years do you stop watching the water drinkers?
The answer is "It doesn't mater but you DO have to select a time limit
What happened here was that a drug company did not publish information about what heppend after the time cut off. My opinion is that this "out of bounds data" should have gone into a footnote, apendix
or whatever but they have a point that what happens after the study period is over should not be included in a published account of the study.
"why do we spend $795 million to bring back space rocks"
Another way to word this question is "Why did each America contribute four bucks to the space rock project?"
Figure there are 250,000,000+ people. At four bucks each you get space rocks and change back.
I'd much rather read about the rocks for $4 than go see a movie for $8.
And we didn't realy _spend_ the money rere-cycled it. The money is not gone it was dumped
rigt back into the economy beacue almost all of the cost is labor(payroll checks) it is not
like if you dumped $795M worth of gold into a volcano
"...Pity, because the world really does need an alternative to Photoshop."
Aperature was NOT intended to replace Photoshop. Aperature's job is to streamline the digital workflow. This is a "Big Deal" for people who shoot hundreds of images a day. Just ty it. Download 200+ images every day, day after day. Just previewing those images and ranking them for quality, deciding which to keep and with to toss, doing minor crops and color corections take _hours_ Profesionals are looking for workflow automation it would be worth much more than $500 if post shoot time could be cust by even 20%
Aperature WAS intended in integrate with Photoshop. You can set up Aperature so that Photoshop is the default image editor and I figure that almost _everyone_ does this.
"You're using a $500 software product with a $300 camera?"
No you are thinking backwards. He is using a $500 software product to manage $10,000 worth of images. The fact that those images where shot with an inexpensive camera does not mater.
I use a Canon A95 also. But for example a couple weeks ago I spent two days of my time and $2,000 worth of SCUBA equipment and waterproof housing and pay for a charter boat to take me to the back side of an island of the California coast. I shoot for three hours and get about 200 images. What did those images cost me? The $300 I paid for the camera is meaningless in that calculation.
Let's say I go in vacation to Hawaii and take the famly or I take the camera with me to my daughter's Halloween party and so on and so on. OK I've actually done all that and have 2500 images all shot wuith the $300 A95.
and now I'm thinking of how I'm going to scan a few thousand slides and negatives I shoot with 35mm and 120 size film. What are these images worth? The cost of the equipment they were shot with? The cost of the time effort and money spent shooting them? IIf so then they are wortth FAR more then the cost of a $500 software aplication AND the Quad Core Power Mac.
If you are a profesional it's easy to know tha value of your images: It's whatever a client is willing to pay you for them. The cost of your time and your equipment does not even come into the calaculation. If you images are paying the rent and you areliving a midle class lifestyle then a few years worth of image are worth a LOT more then a Quad Core Powwer Mac and a RAID disk system and the $500 cost of Aperature would be in the noise.
All that said I'm seriouly looking to buy a DSLR soon. My A95 is just to darn slow and the image quality is not up to 35mm film standards and I think the new Nikon D200 will do better. OK it's a $2K camera but it will shoot tens of thosands of images in it's lifetime - A nickle a shot maybe.
You don't need special software to take advantage of muliple CPUs. All you need to more threads or processes. Certainly for servers this is not a problem for the desktop if you run multiple programs at once multiple CPUs can be used. However most users don't stress even one CPU. Unless you are running a game or rendering video.
One thing, I looked at iTunes and found it had 11 threads active. THese were all running on one litle 1.25Ghz G4 processor (without stressing it in the least) So even if I had bought a quad core power Mac, iTunes would surely have used those four cres but I'd see absolutly zero increase in speed. If the single CPU is not at 100% utilization then adding another CPU means that both will run below 50%.
OK, so my point is you need BOTH an applacation sweet that has multiple threads/processes AND a big enough job to get done such that the CPU is used past 100% Many uses just don't do this kind of work Only if you play games or create digital content.
THere _has_ to be some level of bureaucraty. You can't just let any low-level employee release a patched version of IE without some kind of review and because 18 bzillion other programs depend of IE's DLLs you can't d a design change to a DLL without asking 18 bzillion other programmers if the change will effect them.
You say "good programmers make their software modular". That's right but MS has an interrest in making cross dependent code. One example is Front Page generates broken HTML and IE can deal with HTML that is broken in just that way. Or MS Office making undocumented system calls. "Lock in" is the secrect of Microsoft's monopoly. Modular code with well documented interfaces would not help them.
About "grep/path/to/codebase/*" Rumor has it that parania runs so deep that no one organization within MS is allowed to have the whole code base. They are comparmentalized.
Have you ever actually tried to fix a bug in a large software system. When I say "large" I define that as a system of such size that no one single person could hope to understand how it works So you have specialists. ON truely large systems I would not know who would know something so I spend time asking around. That's the trouble with having thousands of programmers, they don't and can't know each other
This means that even if you have 1,000 programmers working for you they are not interchangable and you have maybe 10 that can work in any one area without starting from zero.
There's more. Because of the need to specialize in narrow areas of the code a "file system guy" would not know the implications of a design change in his area to the "SQL Server guys" and would want to hold a meeting and review changes with others. What this means in large systems is that any simple change needs to get writtren up and reviewed and aproved by some "change control board" and then coded and then you need to test to seethat you have not broken something unrelated. It's a slow process one that is frustrating to developers but design reviews, change
control borads and regression testing are the only way to insure quality
There is a way out but it requires that you design your big system in such a way that it is not so interrelated and the parts are more "stand alone" Basically you make hard rules that one subsystem just absolutly can not enven "know" much less depend on the design of another subsystem. There are many ways to do this. The problem is that Windows is a big interrelated "house of cards" and they know it. It must be very hard to fix one thing without introducing risk that something else is not broken. The standard method for managing this kind of risk is to only do widly spaced releases of the software
In my own work I have to always resist to tempation to fix a bug _now_ and get it out. I want to look responsive to user needs but no. I just have to say "it will be fixed in the next decimal point release, in about 6 or 8 weeks.
And then again from Microsoft's point of view, users don't pay for bug fixes, and will stand in line at night to buy whatever junk is offered. Bug fixes are "money down a rat hole" why bother?
Of course as many have pointed out "Asterisk" is the answer. The poblem is someone who runs, say a shoe store will not be able to set up an Asterisk server himself and will need to hire a consultent. But this is not different from seting up a point of sale cash register application or heck, even painting the office interior. For those who don't like Asterisk there is a new "fork" called "Open PBX" that usesAsisk cose but making it truely open and "pure GPL'd" While this gets you removed from Digium's comercialium it makes seting up a PBX even more of a do it yourself project. Open sourse does not mean "free as in free beer" you will need either your time to learn and set it up or you pay someone else.
The good news is that nerds with free time really can have a hugely complex phone system with all the bells and wistles for just the time it takes them to implement it. Even for a single guy living in a small apertment Asterisk is usfull. It's nice to get voice messages emailed to me. I hate going through the answering machine's messages one at a time. It is worth instaling Asterisk just so I can see the messages inside a web browser and click on the ones I want to hear without need to leasten to them all. Also access to the messages from any phone or PC world wide is worth it too., I hate those little tape machines.
THere are other questions to be answered even after you deside to use Asterix/Open PBX. For exampleyou will need to select one or more VOIP service providers, these are companies that coonect VOIP to the PSTN
I used one of the "real-time modified" linux kernels in my work too. I can see why LM selected it. For those who don't understand "real time" is does not mean "runs fast" it means that if I need a task to run once every 100 miliseconds it does just that. If you are doing something like controling a radar transmitter you need _exact_ timming. For jobs like running a web browser a real-time OS may seem unresponsive and "not smooth"
They will likely need access to the kernel source code if they are fielding a military system. These systems have long (25 years plus) lifetimes and you need the ability to repair the OS 20 years from now. I used a real time Linux inside an astronomical CCD camera to generate the waveforms that shift the charges acros the rows and out to the amps and digitiziers. the camera moved relative to the target and the charge was shifted in sync with the moving image. Linuix was great. The hard real time stuff could be done with interrupts disabled inside a kernel level driver and the data was written to disk by a user land process. Standard tols and debugers could be used to develope the camera controller. I'm not surprized at all that LM went this way. But the "Linux" they are using is NOT the "linux" you see when you get Ret Hat or Debian.
All these arguments about how an ET signal could not harm a current model computers are correct. They would have to know something about our computers, the instruction sets operating systems and so on. Even if they did manage to get a virus in, so what? We know how to manages viruses. In the worst case the computer catches fire and melts. Big deal buy a new one. Even a million dead PCs is not a "Big Deal"
But what they _Could_ do is send a series of instructions teling us the basics of how one of _thier_ computers work. Reading these instructions would be irresistable to us. Then they would send a program. The program could be a simulation of one of them, a (to us) very advanced inteligent being. We could talk to this simulation and learn a lot about the ETs without the long speed of light delay. Everyone would want to talk to these simulated ETs, it would be widely copied and distributed
Remember the story of the Trojan Horse. It was a very atractive gift that the Trojans welcomed into thier city, only later did they discover the bad side effects.
I really do expect that a REAL ET transmition would include a simulation of the ET's best technical experts, performers and artists along with vertual reality tours of their world. They would not be sending us text books. That is "so 1980's" They would send us copies (or simulations) of themselves.
Now we ask "what happens when two cultures make contact when one of them is much more advanced?"
WHat do you meand "new life". Computers don't slow down as they age. Whatever the computer did backl when it was new it can still do today.
I have a 144Mhz Penitum notebook with 80MB RAM. I installed Linux on this machine the first day I owned it. I never even booted Windows even once. The PC worked well then and continues to work well today. I used the machine for software development which means running four copies of a terminal emulator. The computer still performs the functions it did when new just as well now as then.
But the problem is that now days people do different things with thier PCs than they did 10 years ago. Today
I want a notebok that will alow me to edit my digital photos and even my video from my mini DV camera. No matter what software I load on the 144Mhz Pentium it is marginal at editing muli-megapixel images and usless at video. For that use even a low-end Apple iBook blows it away.
But I agree with the parrent. Basically it says "If you have a 1990's vintage PC, it wil still perform quite well if you go and get some 1990's vintage software for it.
Backing up a multi-tarabyte server is no as hard as you make it out to be for several reasons:
1) You do not need to backup the whole serverevery day. All you need to backup are the changes made since the last backup. So the question is NOT the size of the server's file system but the number of bytes that change each day.
2) Slowing down the server? Yes every write to the server disk means a write tio the disk, a read back and a write to the backup media. But on a typical PC server what is the write vs. read ratio? Mostly reads I'd bet.
3) You can set the backup to run at a low priority so it will run whent he server has nothing else to do
I know some people _DO_ want powerfull notebook computers. I for one would love a notebook that would let me edit HD Video. Wouldn't it be great if that notbook could be re-configured for either 6 hours of batery life or high compute power. I know you can't have both at the same time One way to build a dual use notebook would be if you could power down 1/2 the CPUs and 1/2 of the RAM
My point was that having multiple cores allows the laptop builder to dynamically adjust both power use and CPU power
There is _always_ a use for more compute power. For example in Photoshop on current machines (Win or Mac) I have to enter filter settings and then click "apply" if the machine were 20 times faster we could loose the "apply" botton and make the filter adjustments in real tim with slider type controls. Other aplications become practical too with a 20X faster machine, such as "google like" searches on recorded voicemail folders.
I think we will see 8 and 16 core notebooks in 10 years and two core notebooks next year.
and they will have the option of powering up of down as many as you need.
Today some of us are starting to think in "Terrabytes" of disk space and giga bytes of RAM. Not long ago it was gigabytes and megabytes. How long untill we hear of "Kilo-CPUs" anyone want to bet?
Slashdot Mirror
"The Market" only works if buyers have a choise. The monopoly laws were passed in responce to J Paul Getrty's oild company at about the turn of the last century. WHat he did was if there was a shop selling oil he'd open a store right next to it and sell oil at a _much_ lower price. He's loose a ton of money but he'd put the old store out of bussiness. Once the first place was gone he's ise the price way up, higher then it was before he came to that town. In my opinion Microsoft is only a monopoly because people are just plain stupid. People could have bought Apple, Linux or one of the others (SGI's Irix was one of the best systems I've ever used. Next was a lot like Mac OSX) But the stupis computer buys looked only at the initial purchace price and went for the botom. Mr. Gates had one billient bit of insite that made him a bilionare. "There are more people who know absolutly nothing about coputers than there are computer experts. Why not sell the that larger group" and he did. If you think about it and are old enough to remember the pre-microsoft world. Cmputer marketing people targeted only compter professionals. Gates broke out of that and went after group that could read price tags did understand much else. Remember he got into bussines computing be puting the boxes in the home first.
No, I figure you'd use something like "shutdown -h now" or "init 0" to exit. The creen saver stops when the process terminates. The big news here is simply that someone got Linux to run as a Windows userland process. OK not 100% new but then they thought to call it a "screen saver". It's cool to be the first to think of something.
In a recent test, researchers attemped to drain the Pacific Ocean by caryying buckets loaded with seawater to the Arizona or Utah desert where water was dumpped on the dry sand. The ive gallon buckets were decleared 100% effective. Each bucket tested did in fact carry up to five gallons of water, exactly as designed. Due to the success of the test program plans are being made to procure 25 more buckets for use in 2006
THe title is completely misleading. Right in the summary it says .."decoded the mitochondrial DNA of..." mitochondrial DNA is a very small repcentage of the total.
Whouldn't it be great if slashdot implemented some kind of bullshit filter.
Sounds like you have come to a sound conclusion. The junk you have collected is just that junk. But it is "junk" because to __collected__ it. It was not ever really of value. But what if you had __created__ it. Let's say you are a photographer and dumb enough to keep all your work on one disk drive, 10 years worth of work. Or let's make it worse. Let's say you are running a bussenis and all you customer data, inventory and pending orders are of the disk? My point is that some people do have valuable and irreplaceable data even if many people have only "junk". If you spend much of your time writing, drawing, shotting video an editing then your data has value but if you are simply making a local copy of data you find on the Internet it diferent. I can put a $$ value on my data easy. I spend maybe 2,000 hours a year writing and the company I work for spends about $120/hour to keep me here (no I don't get that much, they need to pay rent, for power and to keep the reastrooms) clean. So by definition I create about $240K worth of data per year, figure that there are about 2,000 people like me that work here.... The cost of the pysical media or drives is trivial THe rule is (or should be) that if you care about data it needs to exist in at least three places so that when one copy dies you still have some redundent storage. No matter how you store it, it wil eventually become unreadable. 100 years from now my kids grandchildren might want to see some old photos of when thier grandparents where 12 years old. THose photos are all digial now. In a few years only a few fine arts photographers and hobbiest will use film.
"Come on folks, this is junior high biology....." Yes it is but there is a high possability that many of the people posting these misinformed comments have not yet entered junior high school. After all kids are all on Winter break now. You really can't blame a 10 year old for not understanding 8th grade science.
It's like saying "what if Issac Newton were never born. Likely 20 years later a group of other people would have figured out clasical physics. With Calculus being invented at that time it was only a matter of a few years before it all fell into place. Same with Darwin. Biologists would have figured it out, but later. ... What if there was life that beagal would have discovered had it not crashed? Same thing If it's there we find it later. Maybe 20 yers later but it it's there we find it.
"triangulation on that signal" Yes exactly. But I think that has already been done. The way to do it is to move the reciever and measure the direction to the signal. As luck would have it the reciever is mounted on a platform that is moving very fast and overs huge distances. The Earth is in orbit around the sun and it is spining too. So if you wait 12 hours your reciever has moved one Earth diameter. If you wait a few months it moves even further. Faking the signal would require someone to place a transmitter in the line of sight between your antenna and mars. Mars moves in the sky in a complex way.
"how much more energy goes into refining the metal for the additional engine?" One way the figure this out is to look at the wholesale price of the metal. If steel sells for 60 cents a pound you might guess that less than 60 cents worth the energy was used to mine, transport and refine the steel. I'd also asume this "steem" engine" could last the life of the car like and automatic transision. They can pick the fluids so they are non-corosive. There would be no cumbustion products inside COmpared to electric it saves a lot. No battery so it is lighter and maybe even cheaper. And you can use a smaler gas engine and recover some cost by loosing a couple cylinders there. People want quick of the line short term acceleration. Storing energy as steam could do that. Also electric hybred does not help at all for long term freeway driving. Steam does.
It happens a lot. Someone who knows very little about computers or software thinks "If I just buy one more gadget or one ome program then I can do XXX" Same with people who don't understand how to fish "I I can only get that extrat fancy reel then I'll catch fish". Or the wantabe photographer who thinks if only he had a better camera he could take beter pictures. I reality a true artist could take carcal from the backyard BBQ and draw on sidewalk something that others might like while tho wantabe artist has a closet of supplies and produces junk.
Marine VHF. It's the common medium range (up to about 25 miles) radio found on all boats. Other bands might be closer to the 107-148 range but none would have transmitters closer to the animals. Of course this is silly. Norwall tusks are not antenna
I'm not arguing one side ot the other. My point is simply that there are two points of view and it is not so black and white.
Let's say we do a small study. We let 100 people drink 8 oz. of name brand bottled water. Now if we find that 50 of them are dead seconds afterward we might conclude something is wrong with the water. But if all them are dead 200 years after drinking the water we don't conclude anyhting about the water. So at what point between two seconds and 200 years do you stop watching the water drinkers? The answer is "It doesn't mater but you DO have to select a time limit What happened here was that a drug company did not publish information about what heppend after the time cut off. My opinion is that this "out of bounds data" should have gone into a footnote, apendix or whatever but they have a point that what happens after the study period is over should not be included in a published account of the study.
"why do we spend $795 million to bring back space rocks" Another way to word this question is "Why did each America contribute four bucks to the space rock project?" Figure there are 250,000,000+ people. At four bucks each you get space rocks and change back. I'd much rather read about the rocks for $4 than go see a movie for $8. And we didn't realy _spend_ the money rere-cycled it. The money is not gone it was dumped rigt back into the economy beacue almost all of the cost is labor(payroll checks) it is not like if you dumped $795M worth of gold into a volcano
Aperature was NOT intended to replace Photoshop. Aperature's job is to streamline the digital workflow. This is a "Big Deal" for people who shoot hundreds of images a day. Just ty it. Download 200+ images every day, day after day. Just previewing those images and ranking them for quality, deciding which to keep and with to toss, doing minor crops and color corections take _hours_ Profesionals are looking for workflow automation it would be worth much more than $500 if post shoot time could be cust by even 20%
Aperature WAS intended in integrate with Photoshop. You can set up Aperature so that Photoshop is the default image editor and I figure that almost _everyone_ does this.
No you are thinking backwards. He is using a $500 software product to manage $10,000 worth of images. The fact that those images where shot with an inexpensive camera does not mater.
I use a Canon A95 also. But for example a couple weeks ago I spent two days of my time and $2,000 worth of SCUBA equipment and waterproof housing and pay for a charter boat to take me to the back side of an island of the California coast. I shoot for three hours and get about 200 images. What did those images cost me? The $300 I paid for the camera is meaningless in that calculation.
Let's say I go in vacation to Hawaii and take the famly or I take the camera with me to my daughter's Halloween party and so on and so on. OK I've actually done all that and have 2500 images all shot wuith the $300 A95. and now I'm thinking of how I'm going to scan a few thousand slides and negatives I shoot with 35mm and 120 size film. What are these images worth? The cost of the equipment they were shot with? The cost of the time effort and money spent shooting them? IIf so then they are wortth FAR more then the cost of a $500 software aplication AND the Quad Core Power Mac.
If you are a profesional it's easy to know tha value of your images: It's whatever a client is willing to pay you for them. The cost of your time and your equipment does not even come into the calaculation. If you images are paying the rent and you areliving a midle class lifestyle then a few years worth of image are worth a LOT more then a Quad Core Powwer Mac and a RAID disk system and the $500 cost of Aperature would be in the noise.
All that said I'm seriouly looking to buy a DSLR soon. My A95 is just to darn slow and the image quality is not up to 35mm film standards and I think the new Nikon D200 will do better. OK it's a $2K camera but it will shoot tens of thosands of images in it's lifetime - A nickle a shot maybe.
You don't need special software to take advantage of muliple CPUs. All you need to more threads or processes. Certainly for servers this is not a problem for the desktop if you run multiple programs at once multiple CPUs can be used. However most users don't stress even one CPU. Unless you are running a game or rendering video. One thing, I looked at iTunes and found it had 11 threads active. THese were all running on one litle 1.25Ghz G4 processor (without stressing it in the least) So even if I had bought a quad core power Mac, iTunes would surely have used those four cres but I'd see absolutly zero increase in speed. If the single CPU is not at 100% utilization then adding another CPU means that both will run below 50%. OK, so my point is you need BOTH an applacation sweet that has multiple threads/processes AND a big enough job to get done such that the CPU is used past 100% Many uses just don't do this kind of work Only if you play games or create digital content.
THere _has_ to be some level of bureaucraty. You can't just let any low-level employee release a patched version of IE without some kind of review and because 18 bzillion other programs depend of IE's DLLs you can't d a design change to a DLL without asking 18 bzillion other programmers if the change will effect them. You say "good programmers make their software modular". That's right but MS has an interrest in making cross dependent code. One example is Front Page generates broken HTML and IE can deal with HTML that is broken in just that way. Or MS Office making undocumented system calls. "Lock in" is the secrect of Microsoft's monopoly. Modular code with well documented interfaces would not help them. About "grep /path/to/codebase/*" Rumor has it that parania runs so deep that no one organization within MS is allowed to have the whole code base. They are comparmentalized.
Have you ever actually tried to fix a bug in a large software system. When I say "large" I define that as a system of such size that no one single person could hope to understand how it works So you have specialists. ON truely large systems I would not know who would know something so I spend time asking around. That's the trouble with having thousands of programmers, they don't and can't know each other This means that even if you have 1,000 programmers working for you they are not interchangable and you have maybe 10 that can work in any one area without starting from zero. There's more. Because of the need to specialize in narrow areas of the code a "file system guy" would not know the implications of a design change in his area to the "SQL Server guys" and would want to hold a meeting and review changes with others. What this means in large systems is that any simple change needs to get writtren up and reviewed and aproved by some "change control board" and then coded and then you need to test to seethat you have not broken something unrelated. It's a slow process one that is frustrating to developers but design reviews, change control borads and regression testing are the only way to insure quality There is a way out but it requires that you design your big system in such a way that it is not so interrelated and the parts are more "stand alone" Basically you make hard rules that one subsystem just absolutly can not enven "know" much less depend on the design of another subsystem. There are many ways to do this. The problem is that Windows is a big interrelated "house of cards" and they know it. It must be very hard to fix one thing without introducing risk that something else is not broken. The standard method for managing this kind of risk is to only do widly spaced releases of the software In my own work I have to always resist to tempation to fix a bug _now_ and get it out. I want to look responsive to user needs but no. I just have to say "it will be fixed in the next decimal point release, in about 6 or 8 weeks. And then again from Microsoft's point of view, users don't pay for bug fixes, and will stand in line at night to buy whatever junk is offered. Bug fixes are "money down a rat hole" why bother?
Of course as many have pointed out "Asterisk" is the answer. The poblem is someone who runs, say a shoe store will not be able to set up an Asterisk server himself and will need to hire a consultent. But this is not different from seting up a point of sale cash register application or heck, even painting the office interior. For those who don't like Asterisk there is a new "fork" called "Open PBX" that usesAsisk cose but making it truely open and "pure GPL'd" While this gets you removed from Digium's comercialium it makes seting up a PBX even more of a do it yourself project. Open sourse does not mean "free as in free beer" you will need either your time to learn and set it up or you pay someone else. The good news is that nerds with free time really can have a hugely complex phone system with all the bells and wistles for just the time it takes them to implement it. Even for a single guy living in a small apertment Asterisk is usfull. It's nice to get voice messages emailed to me. I hate going through the answering machine's messages one at a time. It is worth instaling Asterisk just so I can see the messages inside a web browser and click on the ones I want to hear without need to leasten to them all. Also access to the messages from any phone or PC world wide is worth it too., I hate those little tape machines. THere are other questions to be answered even after you deside to use Asterix/Open PBX. For exampleyou will need to select one or more VOIP service providers, these are companies that coonect VOIP to the PSTN
I used one of the "real-time modified" linux kernels in my work too. I can see why LM selected it. For those who don't understand "real time" is does not mean "runs fast" it means that if I need a task to run once every 100 miliseconds it does just that. If you are doing something like controling a radar transmitter you need _exact_ timming. For jobs like running a web browser a real-time OS may seem unresponsive and "not smooth" They will likely need access to the kernel source code if they are fielding a military system. These systems have long (25 years plus) lifetimes and you need the ability to repair the OS 20 years from now. I used a real time Linux inside an astronomical CCD camera to generate the waveforms that shift the charges acros the rows and out to the amps and digitiziers. the camera moved relative to the target and the charge was shifted in sync with the moving image. Linuix was great. The hard real time stuff could be done with interrupts disabled inside a kernel level driver and the data was written to disk by a user land process. Standard tols and debugers could be used to develope the camera controller. I'm not surprized at all that LM went this way. But the "Linux" they are using is NOT the "linux" you see when you get Ret Hat or Debian.
All these arguments about how an ET signal could not harm a current model computers are correct. They would have to know something about our computers, the instruction sets operating systems and so on. Even if they did manage to get a virus in, so what? We know how to manages viruses. In the worst case the computer catches fire and melts. Big deal buy a new one. Even a million dead PCs is not a "Big Deal" But what they _Could_ do is send a series of instructions teling us the basics of how one of _thier_ computers work. Reading these instructions would be irresistable to us. Then they would send a program. The program could be a simulation of one of them, a (to us) very advanced inteligent being. We could talk to this simulation and learn a lot about the ETs without the long speed of light delay. Everyone would want to talk to these simulated ETs, it would be widely copied and distributed Remember the story of the Trojan Horse. It was a very atractive gift that the Trojans welcomed into thier city, only later did they discover the bad side effects. I really do expect that a REAL ET transmition would include a simulation of the ET's best technical experts, performers and artists along with vertual reality tours of their world. They would not be sending us text books. That is "so 1980's" They would send us copies (or simulations) of themselves. Now we ask "what happens when two cultures make contact when one of them is much more advanced?"
WHat do you meand "new life". Computers don't slow down as they age. Whatever the computer did backl when it was new it can still do today. I have a 144Mhz Penitum notebook with 80MB RAM. I installed Linux on this machine the first day I owned it. I never even booted Windows even once. The PC worked well then and continues to work well today. I used the machine for software development which means running four copies of a terminal emulator. The computer still performs the functions it did when new just as well now as then. But the problem is that now days people do different things with thier PCs than they did 10 years ago. Today I want a notebok that will alow me to edit my digital photos and even my video from my mini DV camera. No matter what software I load on the 144Mhz Pentium it is marginal at editing muli-megapixel images and usless at video. For that use even a low-end Apple iBook blows it away. But I agree with the parrent. Basically it says "If you have a 1990's vintage PC, it wil still perform quite well if you go and get some 1990's vintage software for it.
Backing up a multi-tarabyte server is no as hard as you make it out to be for several reasons: 1) You do not need to backup the whole serverevery day. All you need to backup are the changes made since the last backup. So the question is NOT the size of the server's file system but the number of bytes that change each day. 2) Slowing down the server? Yes every write to the server disk means a write tio the disk, a read back and a write to the backup media. But on a typical PC server what is the write vs. read ratio? Mostly reads I'd bet. 3) You can set the backup to run at a low priority so it will run whent he server has nothing else to do
I know some people _DO_ want powerfull notebook computers. I for one would love a notebook that would let me edit HD Video. Wouldn't it be great if that notbook could be re-configured for either 6 hours of batery life or high compute power. I know you can't have both at the same time One way to build a dual use notebook would be if you could power down 1/2 the CPUs and 1/2 of the RAM My point was that having multiple cores allows the laptop builder to dynamically adjust both power use and CPU power There is _always_ a use for more compute power. For example in Photoshop on current machines (Win or Mac) I have to enter filter settings and then click "apply" if the machine were 20 times faster we could loose the "apply" botton and make the filter adjustments in real tim with slider type controls. Other aplications become practical too with a 20X faster machine, such as "google like" searches on recorded voicemail folders. I think we will see 8 and 16 core notebooks in 10 years and two core notebooks next year. and they will have the option of powering up of down as many as you need. Today some of us are starting to think in "Terrabytes" of disk space and giga bytes of RAM. Not long ago it was gigabytes and megabytes. How long untill we hear of "Kilo-CPUs" anyone want to bet?