The Whitechapel Bell Foundry started in 1570
on its current site. Later research indicated the company could trace itself back to 1420.
Some famous bells they have cast include the Liberty Bell (damaged in transit, they offered to repair it properly but a little war got in the way) and Big Ben. They recently cast a bell donated to the City of New York by the City of London on the first anniversary of 9/11.
We're going to have the same level of saturation in the next few years for the 2.4ghz band (the band that the current cordless phones AND 802.11 routers use).
I can just see the complaints being filed with the FCC as all of this wireless equipment we're buying starts going to pot on us because we have this giant radio signal "collision domain" that we're going to use up.
How long will it be before the techno-vandals start building pocket jammers to bring down public wireless networks? Even worse, microwave oven magnetrons put out over 600W of RF at 2.45GHz, damn close to the 802.11b/g frequency bands. A bit of creative retuning and you could have a really fun time during, say, final exams on a whole campus, or take down a complete office building's network.
My best memory of the EPA is seeing them in total body bunny suits to remove the asbestos from a locomotive in a park here, when all it takes is a simple wetting with a garden hose to keep any dust from going air born (besides its the cronic exposure that froms the risk).
As you say the problem with asbestos tends to be with repeated exposure. However you might like to consider the guys in the "bunny suits" do this for a living and are regularly exposed to the stuff. Wetting the dust to stop it from getting airborne doesn't help in the long term as it will dry out and become dangerous again. It needs to be collected, hence the big plastic tents and the guys in the bunny suits.
The only reason processors get so hot is overly high clock speed. With a well designed processor this isn't necessary. What kind of company would build a processor that needed to clock up to 3ghz just to competitive?
Motorola? (No proper DDR support, no Hyperthreading, poor FP implementation...)
I wonder if the 1.42GHz G4s used in Apple's Ultimate dualies are Golden Samples or genuine production devices?
Now I see that these are advertised as having DDR333. Can anyone elaborate on that? Do these PowerMacs make use of the advantages of DDR333 over regular SDRAM?
Apple are using 333DDR memory because there are no mass market manufacturers of 166MHz SDRAM. The Motorola G4 cannot talk to DDR properly but the new versions of the chip have managed to up their memory speed from 133MHz to 166MHz. SDRAM and DDR are now comparable in price so the the motherboard memory interface has been reworked to use DDR at half its total capacity. Any new-design chips like the 970 will use DDR (or RAMBUS) to its maximum capacity.
I have one of the Dual 1.25 GHZ G4's and from my experiences with it, even in dual processor capable apps, it seems as if most of the heavy lifting falls on one processor most of the time.
Most computation-intensive programs like Photoshop are multi-threaded, that is, they launch short processes (threads) that complete and return results back to a scheduler. A dual or multiprocessor machine with a suitable OS (like OSX or Win2k or XP Pro) can run a thread on each CPU in parallel.
This is a real advantage for some programs but they have to be designed to take advantage of this from the ground up. Disc-intensive or simple programs often aren't, but anything that uses lots of CPU power will benefit from running on two CPUs. The "new" Hyperthreading functionality of the Intel P4 CPUs gives them some of the same advantages as two CPUs but only when the OS and the applications being run can take advantage of it.
The Apple's PowerPC doesn't work very well in a multi-CPU design due to restrictions on its memory bus and internal cache control; that's one reason for the very large L3 caches in the current range of Apple desktop dual-CPU boxes.
Apple is working on better CPU's from all the rumors and such going around and if they keep with the dual processor game, the Mac will be set to overtake PC's in short order.
Apple isn't. It buys CPUs from other manufacturers. There seem to be plans for Apple to start using the 970 CPU from IBM, but this CPU will not be available until the end of 2003 at the earliest (unless a miracle happens). When it comes out it will be about as fast as the current fastest P4 is now. The Intel roadmap indicates the P4 family will reach at least 3.6GHz by 2003Q4, another 20% faster.
As for "bang per buck", the only rumours I've heard on the price of IBM's 970 part suggest it will cost about 2000 dollars each. IBM do not plan on making a lot of them since their only mass-market customer is likely to be Apple. This means the development and production costs will be high per chip. The current price for the top-end P4 is about 600 dollars.
Repeat after me, children, "MHz does not equal performance."
Pity that CPU speeds are measured in GHz these days, isn't it? Been standing a little too deep in the Reality Distortion Field again, huh?
Over 15 years ago, I went to a good friend who is an electrical engineer for advice on what computer I should buy. His response is still valid today, "Well, what do you want to do with it ?"
That's electrical engineers for ya -- a day late and a dollar short. I ask three questions:
"What do you want to do with it?"
"What do you think you'll be wanting to do with it in eighteen months time?"
"How much can you afford?"
That last question tends to knock Apples out of the running unless the potential buyer is a believer in style over substance and also pretty well off.
Why even go forward with this? By starting at levels of chips FIVE years old, are they going to continue forward in such a way indefinitely? Their products will be running 2002 tech levels in 2007?
I'd guess, if they continue with the design, by 2007 they'll be making 2005-level hardware. They won't have gone down so many blind alleys in the process and the design software and manufacturing equipment available to them will be much better than today's. They have to start somewhere, and they've chosen an achievable target design which has the potential to be useful from day one as a general-purpose low-end desktop CPU, as well as perhaps being embedded as a microcontroller core.
Actors also get screwed over with the gross/net point deal. Sir Alec Guinness got something like 1.5% of the gross from the first Star Wars movie, as he had worked in Hollywood before and had been screwed over previously by the studio system's attack accountants. They needed hydraulic jacks to get the smile off his face when the millions started rolling in.
It may be Urban Legend but I have heard it said that the Star Wars films haave yet to show a profit -- on paper, at least.
One scratch could already kill your disc. The red-book CD-DA standard DOES NOT provide ANY EDC or ECC data on the disc, merely RAW digitised audio.
Trust me.
Err, have you actually read the Red Book? If you had you might have noticed that the data is broken into frames. Each data frame (there are also control frames, which is what the copy protection screws around with) contains the audio data for the two stereo channels. This is encoded in an eight-to-fourteen pattern on the disc and the data frames themselves are in Cross-Interleaved Reed Solomon (CIRC) codes as well to allow some error recovery from, say, scratches. It is pretty simple-minded but easy to build decoders for, an important requirement because of the cost of 1980's electronics when the spec was laid out.
I wonder if this is not just a ploy to find out who the bad apples of the bunch are. Maybe the record industry knows this is futile battle and hence wants to combat it somehow.
If they wanted to find out which reviewers were releasing previews, the easiest way would be to send out a few hundred gold CD-Rs with each one individually steganographically encoded. When the album appears on KaZaa or wherever, look for the codes and backtrack the gold disc. Bad reviewer! No freebie cruises for you!
didn't they film all three films at once? this would mean that they are NOT in post production but just waiting to be realease.
That's like saying they wrote all the code, so it's just waiting to be released. Forget about debugging, test, customer acceptance trials, documentation etc.
The LOTR team keep bringing actors back to reshoot scenes for the second and third movies because the director changed his mind, or something didn't work right, or continuity blew it (sorta like debugging). They resynch all the dialogue in a recording studio, compose and record the music, create and tie in the special effects, edit all the film they shot into a coherent whole and effectively turn what they've got into a sellable product. This takes longer than principal photography, although it doesn't cost as much.
New Line don't have the next two movies sitting in a vault under lock and key somewhere. They are works in progress.
The first one, The Fellowship Of the Ring, was also leaked well ahead of time. The copy was a rip of a screening DVD, and was distributed in DVD quality.
If I recall correctly (and if I don't, I expect I will be politely corrected...) the rip of FOTR came from an Academy (read: Oscars) DVD that was circulated to possible voters. It came out quite a while after the cinema release of the movie itself; the first FOTR rip I saw was at a party in February, and that was from a camcorder.
Right now there is no complete TTT movie to send to Academy voters on DVD. There *might* be a rough-cut (no SFX, duff music, gaps with a whiteboard reading "big battle scene here") but that's all there is. Peter Jackson is still fine-tuning the release version (come on guys, you know what it's like trying to get finished code out the door...)
Peter Jackson is still cutting and chopping at the version of The Two Towers movie he's planning to release in December. He's well known as a fiddler, unable to finish the damn thing and push it out the door. Whatever's on the Net right now isn't going to be what's on the screen in December.
I'm also suspicious that nobody in the Drudge article claims to have actually seen this download. There *are* copies of the camcordered TTT trailer circulating; this has never been officially released, and was only shown in cinemas around March-April this year. Some of it (but not all) appears on the extras DVD of the Fellowship Of The Ring. Maybe that's what's causing all the panic. We'll see.
Photovoltaics are simply a dual layer of silicon. Using them doesn't degrade the material as far as I've learned, so I don't see why power generated should diminish.
PV cell performance degrades on long-term exposure to sunlight -- basically the cell surface fades like paint or cloth does. This is a bit of a disadvantage for photovotaics, you must admit. The electrochemistry of the PV surface is complex; too much heat is a problem, as well as mechanical fracture defects from repeated heating/cooling cycles (like in a Canadian winter...)
Another point: We are at a point where just over a year ago companies were racing to hit that magical '1.0 Ghz' mark for CPUs...and now we have CPU speeds close to 3Ghz! A question that needs to be asked, does anyone need that kind of power? I mean seriously?
Perhaps you recall Bill Gates having a new asshole surgically installed when he claimed "640k should be enough for anyone?"
Work expands to fill the resources made available to it. You say you're a gamer -- have you seen the current hardware spec required to run DOOM 3 in anything other than a limping manner? It takes more than enhanced GPU performance to model and manipulate the pocket universe most hi-reality FPS games present to the user. That's where all those clock cycles go, and that's why the speed race is continuing. Unfortunately for Apple, they hitched their wagon to PPC architecture a while back, and the wheels have fallen off. Their only solution, I think, would be to licence the PPC architecture and commission their own G5/G6 designs, but this would cost serious time (two years, maybe three) and money (at least a billion, maybe two). Apple users tend to be rich types but even they might cough at having to pay an extra five hundred bucks per box because the cost of the new CPUs has gone waaaay up.
I could easily believe Apple have a small project team whose job it is to make ports of OS/X to target x86 hardware *just in case*. The current G4s are only made in a couple of fabs; bad luck or an Enron could leave Apple with lotsa boxes and no CPUs to install in them. That doesn't mean that they would ever release an x86 port unless they really really had to, to save the company.
we showed how overclockable the Athlon XP 2600+, which officially runs at 2133 MHz is. We were easily able to overclock it to a stable 2400 GHz using a conventional CPU cooler.
There are two ways of killing a modern Main Battle Tank -- the first is to use a DU penetrator, but this requires a Big Gun, up to 120mm bore and it is usually found on another MBT. It hits from the side, front or back and that is where most of the MBT's armour is concentrated, to try and stop the penetrator getting through. Electrical discharge would probably not stop this kind of kinetic attack; for one thing DU takes a lot of energy to melt it.
The second form of attack is from above, by smart air-launched or ground-launched missiles which fire light self-forging projectiles down through the lighter-armoured top and turret of the tank. The MBT designers can't fit heavier armour there as the extra weight would cripple the tank. The electric armour would be best placed here, as these self-forged projectiles seem to be vulnerable to disruption by electrical discharge.
APCs and other lighter-skinned vehicles might benefit from this kind of armour to defend against missile attacks, but they are still dead meat for any kind of large DU penetrator round.
AFAIK there are no industrial I/O subsystems that can be easily hooked up to a Mac, but plenty for the PC. Integrators have been building rack-mounted PCs and IP-65+ x86 boxes for industrial control of conveyors and other manufacturing systems for decades, and there's a large software base out there too with specialist tools for sequencing, failure recovery etc. Macintoshes are cute and pretty, but that doesn't cut it in a refrigerated food distribution plant or an iron foundry (a couple of places I've installed PC control/monitoring systems in the recent past).
The boxes themselves don't really care what they run -- for the iron foundry's soak-pit monitor (measuring and controlling the cooling of 100-tonne castings from 1000 deg C down to room temp over a period of up to four weeks) I ran DOS 5.0 because it didn't need anything smarter. For IBM's shop-floor control they needed something more complex with networking capability, but I think you'd find that the Linux image plus apps each individual box ran would probably be tiny. No use including the GUI kitchen sink if the box is only going to be doing a couple of simple jobs. Going with Windows in that kind of application doesn't make a lot of sense to start with, but thaat's not what it was designed for.
Slashdot Mirror
The Whitechapel Bell Foundry started in 1570 on its current site. Later research indicated the company could trace itself back to 1420.
Some famous bells they have cast include the Liberty Bell (damaged in transit, they offered to repair it properly but a little war got in the way) and Big Ben. They recently cast a bell donated to the City of New York by the City of London on the first anniversary of 9/11.
http://www.whitechapelbellfoundry.co.uk/
I can just see the complaints being filed with the FCC as all of this wireless equipment we're buying starts going to pot on us because we have this giant radio signal "collision domain" that we're going to use up.
How long will it be before the techno-vandals start building pocket jammers to bring down public wireless networks? Even worse, microwave oven magnetrons put out over 600W of RF at 2.45GHz, damn close to the 802.11b/g frequency bands. A bit of creative retuning and you could have a really fun time during, say, final exams on a whole campus, or take down a complete office building's network.
As you say the problem with asbestos tends to be with repeated exposure. However you might like to consider the guys in the "bunny suits" do this for a living and are regularly exposed to the stuff. Wetting the dust to stop it from getting airborne doesn't help in the long term as it will dry out and become dangerous again. It needs to be collected, hence the big plastic tents and the guys in the bunny suits.
The only reason processors get so hot is overly high clock speed. With a well designed processor this isn't necessary. What kind of company would build a processor that needed to clock up to 3ghz just to competitive?
Motorola? (No proper DDR support, no Hyperthreading, poor FP implementation...)
I wonder if the 1.42GHz G4s used in Apple's Ultimate dualies are Golden Samples or genuine production devices?
Apple are using 333DDR memory because there are no mass market manufacturers of 166MHz SDRAM. The Motorola G4 cannot talk to DDR properly but the new versions of the chip have managed to up their memory speed from 133MHz to 166MHz. SDRAM and DDR are now comparable in price so the the motherboard memory interface has been reworked to use DDR at half its total capacity. Any new-design chips like the 970 will use DDR (or RAMBUS) to its maximum capacity.
Most computation-intensive programs like Photoshop are multi-threaded, that is, they launch short processes (threads) that complete and return results back to a scheduler. A dual or multiprocessor machine with a suitable OS (like OSX or Win2k or XP Pro) can run a thread on each CPU in parallel.
This is a real advantage for some programs but they have to be designed to take advantage of this from the ground up. Disc-intensive or simple programs often aren't, but anything that uses lots of CPU power will benefit from running on two CPUs. The "new" Hyperthreading functionality of the Intel P4 CPUs gives them some of the same advantages as two CPUs but only when the OS and the applications being run can take advantage of it.
The Apple's PowerPC doesn't work very well in a multi-CPU design due to restrictions on its memory bus and internal cache control; that's one reason for the very large L3 caches in the current range of Apple desktop dual-CPU boxes.
Apple is working on better CPU's from all the rumors and such going around and if they keep with the dual processor game, the Mac will be set to overtake PC's in short order.
Apple isn't. It buys CPUs from other manufacturers. There seem to be plans for Apple to start using the 970 CPU from IBM, but this CPU will not be available until the end of 2003 at the earliest (unless a miracle happens). When it comes out it will be about as fast as the current fastest P4 is now. The Intel roadmap indicates the P4 family will reach at least 3.6GHz by 2003Q4, another 20% faster.
As for "bang per buck", the only rumours I've heard on the price of IBM's 970 part suggest it will cost about 2000 dollars each. IBM do not plan on making a lot of them since their only mass-market customer is likely to be Apple. This means the development and production costs will be high per chip. The current price for the top-end P4 is about 600 dollars.
Pity that CPU speeds are measured in GHz these days, isn't it? Been standing a little too deep in the Reality Distortion Field again, huh?
Over 15 years ago, I went to a good friend who is an electrical engineer for advice on what computer I should buy. His response is still valid today, "Well, what do you want to do with it ?"
That's electrical engineers for ya -- a day late and a dollar short. I ask three questions:
"What do you want to do with it?"
"What do you think you'll be wanting to do with it in eighteen months time?"
"How much can you afford?"
That last question tends to knock Apples out of the running unless the potential buyer is a believer in style over substance and also pretty well off.
I'd guess, if they continue with the design, by 2007 they'll be making 2005-level hardware. They won't have gone down so many blind alleys in the process and the design software and manufacturing equipment available to them will be much better than today's. They have to start somewhere, and they've chosen an achievable target design which has the potential to be useful from day one as a general-purpose low-end desktop CPU, as well as perhaps being embedded as a microcontroller core.
I'd like mine in the TV remote please, so I can find the damn furgling thing.
Where do I send the ten bucks, and when do I start?
It may be Urban Legend but I have heard it said that the Star Wars films haave yet to show a profit -- on paper, at least.
Yeah, but imagine what happens when the shit hits the fan...
2. ???
3. Profit!!!
Trust me.
Err, have you actually read the Red Book? If you had you might have noticed that the data is broken into frames. Each data frame (there are also control frames, which is what the copy protection screws around with) contains the audio data for the two stereo channels. This is encoded in an eight-to-fourteen pattern on the disc and the data frames themselves are in Cross-Interleaved Reed Solomon (CIRC) codes as well to allow some error recovery from, say, scratches. It is pretty simple-minded but easy to build decoders for, an important requirement because of the cost of 1980's electronics when the spec was laid out.
Maggie Thatcher was a chemical engineer before she moved into politics.
If they wanted to find out which reviewers were releasing previews, the easiest way would be to send out a few hundred gold CD-Rs with each one individually steganographically encoded. When the album appears on KaZaa or wherever, look for the codes and backtrack the gold disc. Bad reviewer! No freebie cruises for you!
Compgeeks were selling reconditioned T221s for about 2500 bucks a few weeks ago.
That's like saying they wrote all the code, so it's just waiting to be released. Forget about debugging, test, customer acceptance trials, documentation etc.
The LOTR team keep bringing actors back to reshoot scenes for the second and third movies because the director changed his mind, or something didn't work right, or continuity blew it (sorta like debugging). They resynch all the dialogue in a recording studio, compose and record the music, create and tie in the special effects, edit all the film they shot into a coherent whole and effectively turn what they've got into a sellable product. This takes longer than principal photography, although it doesn't cost as much.
New Line don't have the next two movies sitting in a vault under lock and key somewhere. They are works in progress.
If I recall correctly (and if I don't, I expect I will be politely corrected...) the rip of FOTR came from an Academy (read: Oscars) DVD that was circulated to possible voters. It came out quite a while after the cinema release of the movie itself; the first FOTR rip I saw was at a party in February, and that was from a camcorder.
Right now there is no complete TTT movie to send to Academy voters on DVD. There *might* be a rough-cut (no SFX, duff music, gaps with a whiteboard reading "big battle scene here") but that's all there is. Peter Jackson is still fine-tuning the release version (come on guys, you know what it's like trying to get finished code out the door...)
I'm also suspicious that nobody in the Drudge article claims to have actually seen this download. There *are* copies of the camcordered TTT trailer circulating; this has never been officially released, and was only shown in cinemas around March-April this year. Some of it (but not all) appears on the extras DVD of the Fellowship Of The Ring. Maybe that's what's causing all the panic. We'll see.
PV cell performance degrades on long-term exposure to sunlight -- basically the cell surface fades like paint or cloth does. This is a bit of a disadvantage for photovotaics, you must admit. The electrochemistry of the PV surface is complex; too much heat is a problem, as well as mechanical fracture defects from repeated heating/cooling cycles (like in a Canadian winter...)
Perhaps you recall Bill Gates having a new asshole surgically installed when he claimed "640k should be enough for anyone?"
Work expands to fill the resources made available to it. You say you're a gamer -- have you seen the current hardware spec required to run DOOM 3 in anything other than a limping manner? It takes more than enhanced GPU performance to model and manipulate the pocket universe most hi-reality FPS games present to the user. That's where all those clock cycles go, and that's why the speed race is continuing. Unfortunately for Apple, they hitched their wagon to PPC architecture a while back, and the wheels have fallen off. Their only solution, I think, would be to licence the PPC architecture and commission their own G5/G6 designs, but this would cost serious time (two years, maybe three) and money (at least a billion, maybe two). Apple users tend to be rich types but even they might cough at having to pay an extra five hundred bucks per box because the cost of the new CPUs has gone waaaay up.
I could easily believe Apple have a small project team whose job it is to make ports of OS/X to target x86 hardware *just in case*. The current G4s are only made in a couple of fabs; bad luck or an Enron could leave Apple with lotsa boxes and no CPUs to install in them. That doesn't mean that they would ever release an x86 port unless they really really had to, to save the company.
Dentistry?
http://www.tomshardware.com/cpu/02q3/020826/inde x.html
we showed how overclockable the Athlon XP 2600+, which officially runs at 2133 MHz is. We were easily able to overclock it to a stable 2400 GHz using a conventional CPU cooler.
I want one of *those*.
The second form of attack is from above, by smart air-launched or ground-launched missiles which fire light self-forging projectiles down through the lighter-armoured top and turret of the tank. The MBT designers can't fit heavier armour there as the extra weight would cripple the tank. The electric armour would be best placed here, as these self-forged projectiles seem to be vulnerable to disruption by electrical discharge.
APCs and other lighter-skinned vehicles might benefit from this kind of armour to defend against missile attacks, but they are still dead meat for any kind of large DU penetrator round.
The boxes themselves don't really care what they run -- for the iron foundry's soak-pit monitor (measuring and controlling the cooling of 100-tonne castings from 1000 deg C down to room temp over a period of up to four weeks) I ran DOS 5.0 because it didn't need anything smarter. For IBM's shop-floor control they needed something more complex with networking capability, but I think you'd find that the Linux image plus apps each individual box ran would probably be tiny. No use including the GUI kitchen sink if the box is only going to be doing a couple of simple jobs. Going with Windows in that kind of application doesn't make a lot of sense to start with, but thaat's not what it was designed for.