Radio telescopes tuning in keep hearing murmurings about how the Galaxy was not taking the Star's ideas seriously, so it's going off to start its own Galaxy.
Hey, I shop at Ikea. The stuff isn't event assembled. It's a flat box full of precision cut boards with bolts and one of those funky allen keys.
Getting back to your point, there is still a market for hand-coders. With most consumer electronics, I'm talking kid's toys, alarm clocks, talking dolls, you try to shave off every penny you can in manufacturing costs. Plus, once you start a product line, you run it out for years.
In that case, of high volume and low cost, it is easy to absorb the cost of a $100,000 hand coder. Especially if he can save you $0.10 a unit on lines where volume is measured in the millions of units.
Besides, most of the "hand coders" I know work more in the $36,000 dollar range.
The HP-UX/Tru64 merger should again be taken up, probably favoring Tru64. "TruHP 12 UNIX" should target PA-RISC, Itanium, and Opteron. GNU Userland wherever possible, RPM packager, AdvFS, 128-cpu scalability, binary compatibility with previous OS implmentations on their respective platforms.
Um, why? If they had been developing it all along, I'd say sure. But they haven't. If you are going to use an external package manager, you might as well f' it and use another software distro.
Linux and BSD have ports for your architecture. Your clients are going to be screaming for compadibility with GNU anyway. Just swallow your pride, pick one (or both) and be content to let those with a geniune interest in operating systems write operating systems.
IBM has grudgingly accepted Linux. Sun is slowly backing away from Solaris. For God's sack, Apple even runs BSD under the hood. Even Microsoft uses chunks of BSD in Windows (the network stack.) Unless you are doing something radical with the OS, there is not excuse for not working with something off the shelf. Especially when the source and APIs are available for you to write custom device drivers.
Oh hell no. Do not use high-performance workstations and "Opteron" in the same sentence. Are they fast? Yes. Would I build a supercomputer out of them? No. If I could build an operating system from scratch, would I choose them? No.
The Opteron does kick the legs out of anything else for running x86 instructions. But if you are building new systems odds are your are going to go PPC. It's out. It's understood. It blew the doors off the x86 architecture back in '91. It is still blowing the doors off the x86 architecture today. You aren't working around a pile of backward compadibility crap. You have an assload more registers to work with. And finally, it uses big-endian addressing.
I will offer one suggestion, as one who was also at one point 3 years from an EE degree.
Forget everything you are told about X being optimal, and Y being old hat. Computer architectures come and go like bell bottoms and short skirts.
Branch prediction is a workaround. It is not a radical performance enhancing technology. It is there to keep the CPU busy when it would otherwise be starved for instructions and data. Branch prediction is simply there to allow the CPU to operate at an insanely high clock speed as compared to the memory bus. And it only works well when you have a relatively fixed target to optimize for (namely Windows.) Branch prediction is also needed because later generations of the i686 processor have insanely long pipelines.
And it flows completely counter to what I was taught back in the late 90s about compiler theory. The idea then was to make instructions small and simple, and let an automated system figure out the optimal way to arrange instructions for maximum throughput.
I full heartedly agree in the concept of Capitalism.
I would argue though that Capitalism hasn't existed in this country since the late 19th century. Ah the joys of natural monopolies, limited liability entities, and holding companies.
What looks like a "market" is in fact more like a casino. If you play the high stakes you have a pretty good chance of winning. Most of the market, though, is over playing the slots.
Re:computers for the Masses not the classes
on
The Sub-$100 Laptop?
·
· Score: 1
And that "psychological" figure was $50 back in the day, and the computer you mentioned was actually $400.
I remember one story about a guy who had his house broken into. The thieves stole everything but his IBM PC. The thieves couldn't figure out what it was. Ironically, of all the things they could have stolen, the PC was the most expensive.
I never feel right asking of cash, or walking out for cash. I get paid all day long for people not to listen to me. And the few times money did change hands... well, no one listened to me.
Bartering is the way to go. It is an exchange of equals. Money is and exchange of buyer to seller (and the seller is always lower on the totem pole.)
My wife actually does run a for pay business, going out and giving computer lessons. In the process, she does also perform repairs, and file recovery, and new system burn in.
Since the lessons are the primary reason she is there, she is treated more like a piano teacher than a plumber. The few times I have tried to work at the computer repair for hire, I nearly lost my mind.
I can do it, but people are just plain... human. They do think money/upgrades/new computers are supposed to fix their problem, and are offended to be told that a change in behavior would be needed.
Um, you don't have DMA on flash media because you address it just like RAM. It's just slower and non-volitile.
DMA is needed on hard drives because seek times are measured in milleseconds. (1/1000th of a second), but once you are there, you know you are going to be reading the file off the drive like a ribbon at megabytes per second. All the data comes in a burst, and the computer usually caches large chunks of the file system in RAM so it doesn't have to access the device to find it.
Re:Spoiled kids these days...
on
iPod Shuffle RAID
·
· Score: 4, Funny
You had cards?
Damn you had it easy. We just had punches back in my day. Our fists would get bloody entering in the simplest of instructions. And then, after every operation, we had to punch everything in again to check it.
Yeah, and you'd also trash both flash media in a relatively short amount of time.
Why not simply use a key with a hard-coded key that is two parts and requires the presence of both to generate a hash.
Or simpler, wire two household deadbolt locks (keyed differently, of course) to complete a circuit when in the latched position, and design your application with an NOR gate that detects when both circuits are open.
(Whistle blow... flag down)
on
iPod Shuffle RAID
·
· Score: 1, Troll
We have unnecissary geekness, and roughing a consumer product. iPods are basically external firewire drives, the hard part has alreadybeen engineered for you.
Repeat of down, half the distance to the goal.
Re:computers for the Masses not the classes
on
The Sub-$100 Laptop?
·
· Score: 1
(Cough) Apple (Cough).
Seriously, with inflation, $200 in 1980 is $484 (2003 dollars) today.
Look it up yourself.
The Apple ][ was $1298 in 1977. ($4009 in 2003 dollars.)
The IBM PC was $3000 in 1981. ( $6389.32 in 2003 dollars.)
I was there. Computers were REALLY expensive. You used to collect soup labels to get discounts. Schools held bake sales for computers. They were NOT cheap.
People who wish to prove me, and the laws of economics, wrong, are invited to go ahead and actually try it.
On the contrary. I would like to relate a case from the brewery industry. They used to throw out the spent grain that was used brewing beer. Then brewers noticed that farmers were using the spent grain to feed livestock.
Today, what was "waste material" is now a comodity, traded on markets and peddled in bulk like oil, lumber, and corn.
Now I've always been under the impression that by using a voltage divider you create a new "ground" which is the center point where your dropping resisitors meet.
In my case I am thinking of a circuit where to generate +/- 6V from 12V. You would take two identical resistors in series. The +6 would be from the + of the source to the connection between the resistors. The -6 would be from the connection to the negative pole of the source.
5 and 3.3 would be generated by putting 2 pair of resistors in series:
Where R1=R4, and R2=R3, and the ratio between R1 and R2 are wieghted to produce either 5V or 3.3V.
I realize the circuit would be a poor choice for any high power application, and the components would have to be reasonably tolerant of what power input they would take.
How often are you popping keyboards out because they are just plain filthy? And working at a science museum, with a lot of interactives, I can tell you from first hand experience that the general public is VERY hard on keyboards. (And equally brutal on buttons, switches, touch screens...)
We actually tried out a keyboard like the one I described made by . It worked for a good long time before it died. Manufacturer link. There is some resistance by the adult public to using them, but only because your typical American office drone is already accustomed to how a keyboard "should" work. Kids just pick up the concept and run with it.
Way before power, you would have to deal with literacy. These laptops are not designed for the chunks of the world that still living a subsistanance agrarian culture.
This system is designed for the chunks of the world that already have electricity and water and phones (at least in civic structures), but where $1200 is a fortune, and where a unit costing $100 that could replace 10+ textbooks costing $10, while providing some added functions, is economically advantageous.
Well first off, no child is going to be lugging these laptops around. The target here is to set up workstations in schools and government facilities. Second, you don't want to make a "premium" version. You defeat all of the advantages of mass production. Finally, as far as suppliers go you have a ready market. Most of the chips that go into consumer electronics items, like Cell phones and calculators and game consoles, are available by the millions, are cheap, and have more than enough processor power for the application.
Slashdot Mirror
Radio telescopes tuning in keep hearing murmurings about how the Galaxy was not taking the Star's ideas seriously, so it's going off to start its own Galaxy.
Getting back to your point, there is still a market for hand-coders. With most consumer electronics, I'm talking kid's toys, alarm clocks, talking dolls, you try to shave off every penny you can in manufacturing costs. Plus, once you start a product line, you run it out for years.
In that case, of high volume and low cost, it is easy to absorb the cost of a $100,000 hand coder. Especially if he can save you $0.10 a unit on lines where volume is measured in the millions of units.
Besides, most of the "hand coders" I know work more in the $36,000 dollar range.
Um, why? If they had been developing it all along, I'd say sure. But they haven't. If you are going to use an external package manager, you might as well f' it and use another software distro.
Linux and BSD have ports for your architecture. Your clients are going to be screaming for compadibility with GNU anyway. Just swallow your pride, pick one (or both) and be content to let those with a geniune interest in operating systems write operating systems.
IBM has grudgingly accepted Linux. Sun is slowly backing away from Solaris. For God's sack, Apple even runs BSD under the hood. Even Microsoft uses chunks of BSD in Windows (the network stack.) Unless you are doing something radical with the OS, there is not excuse for not working with something off the shelf. Especially when the source and APIs are available for you to write custom device drivers.
Oh hell no. Do not use high-performance workstations and "Opteron" in the same sentence. Are they fast? Yes. Would I build a supercomputer out of them? No. If I could build an operating system from scratch, would I choose them? No.
The Opteron does kick the legs out of anything else for running x86 instructions. But if you are building new systems odds are your are going to go PPC. It's out. It's understood. It blew the doors off the x86 architecture back in '91. It is still blowing the doors off the x86 architecture today. You aren't working around a pile of backward compadibility crap. You have an assload more registers to work with. And finally, it uses big-endian addressing.
Forget everything you are told about X being optimal, and Y being old hat. Computer architectures come and go like bell bottoms and short skirts.
Branch prediction is a workaround. It is not a radical performance enhancing technology. It is there to keep the CPU busy when it would otherwise be starved for instructions and data. Branch prediction is simply there to allow the CPU to operate at an insanely high clock speed as compared to the memory bus. And it only works well when you have a relatively fixed target to optimize for (namely Windows.) Branch prediction is also needed because later generations of the i686 processor have insanely long pipelines.
And it flows completely counter to what I was taught back in the late 90s about compiler theory. The idea then was to make instructions small and simple, and let an automated system figure out the optimal way to arrange instructions for maximum throughput.
Well it sure beats admitting that the rest of your family are id10ts.
I would argue though that Capitalism hasn't existed in this country since the late 19th century. Ah the joys of natural monopolies, limited liability entities, and holding companies.
What looks like a "market" is in fact more like a casino. If you play the high stakes you have a pretty good chance of winning. Most of the market, though, is over playing the slots.
I remember one story about a guy who had his house broken into. The thieves stole everything but his IBM PC. The thieves couldn't figure out what it was. Ironically, of all the things they could have stolen, the PC was the most expensive.
Thanx.
Bartering is the way to go. It is an exchange of equals. Money is and exchange of buyer to seller (and the seller is always lower on the totem pole.)
My wife actually does run a for pay business, going out and giving computer lessons. In the process, she does also perform repairs, and file recovery, and new system burn in.
Since the lessons are the primary reason she is there, she is treated more like a piano teacher than a plumber. The few times I have tried to work at the computer repair for hire, I nearly lost my mind.
I can do it, but people are just plain... human. They do think money/upgrades/new computers are supposed to fix their problem, and are offended to be told that a change in behavior would be needed.
You have to watch it though. The one relative who actually let me talk him into using Linux has practically started a cult.
I get calls from my siblings and in-laws all the time asking me to come out and fix this, and help them fix a problem with a driver over the phone...
For Windows ME...
While I'm in a meeting...
And they don't have the CD...
See lessons from history:
DMA is needed on hard drives because seek times are measured in milleseconds. (1/1000th of a second), but once you are there, you know you are going to be reading the file off the drive like a ribbon at megabytes per second. All the data comes in a burst, and the computer usually caches large chunks of the file system in RAM so it doesn't have to access the device to find it.
Damn you had it easy. We just had punches back in my day. Our fists would get bloody entering in the simplest of instructions. And then, after every operation, we had to punch everything in again to check it.
Why not simply use a key with a hard-coded key that is two parts and requires the presence of both to generate a hash.
Or simpler, wire two household deadbolt locks (keyed differently, of course) to complete a circuit when in the latched position, and design your application with an NOR gate that detects when both circuits are open.
Repeat of down, half the distance to the goal.
Seriously, with inflation, $200 in 1980 is $484 (2003 dollars) today. Look it up yourself.
And for the record, the C64 was $429 when it came out in 1982 ($930.69 2003 dollars).
The Apple ][ was $1298 in 1977. ($4009 in 2003 dollars.)
The IBM PC was $3000 in 1981. ( $6389.32 in 2003 dollars.)
I was there. Computers were REALLY expensive. You used to collect soup labels to get discounts. Schools held bake sales for computers. They were NOT cheap.
On the contrary. I would like to relate a case from the brewery industry. They used to throw out the spent grain that was used brewing beer. Then brewers noticed that farmers were using the spent grain to feed livestock.
Today, what was "waste material" is now a comodity, traded on markets and peddled in bulk like oil, lumber, and corn.
In my case I am thinking of a circuit where to generate +/- 6V from 12V. You would take two identical resistors in series. The +6 would be from the + of the source to the connection between the resistors. The -6 would be from the connection to the negative pole of the source.
5 and 3.3 would be generated by putting 2 pair of resistors in series:
Where R1=R4, and R2=R3, and the ratio between R1 and R2 are wieghted to produce either 5V or 3.3V.
I realize the circuit would be a poor choice for any high power application, and the components would have to be reasonably tolerant of what power input they would take.
How often are you popping keyboards out because they are just plain filthy? And working at a science museum, with a lot of interactives, I can tell you from first hand experience that the general public is VERY hard on keyboards. (And equally brutal on buttons, switches, touch screens...)
We actually tried out a keyboard like the one I described made by . It worked for a good long time before it died. Manufacturer link. There is some resistance by the adult public to using them, but only because your typical American office drone is already accustomed to how a keyboard "should" work. Kids just pick up the concept and run with it.
This system is designed for the chunks of the world that already have electricity and water and phones (at least in civic structures), but where $1200 is a fortune, and where a unit costing $100 that could replace 10+ textbooks costing $10, while providing some added functions, is economically advantageous.
Well first off, no child is going to be lugging these laptops around. The target here is to set up workstations in schools and government facilities. Second, you don't want to make a "premium" version. You defeat all of the advantages of mass production. Finally, as far as suppliers go you have a ready market. Most of the chips that go into consumer electronics items, like Cell phones and calculators and game consoles, are available by the millions, are cheap, and have more than enough processor power for the application.