Actually, counterfeiting is essentially more of a problem now than ever before. There are all the nice features of real bills, but seriously how often have you actually see someone hold a bill under a blacklight, or use a pen, or look for the watermark when accepting the bill. My answer: maybe a dozen times in the last couple years.
At the same time, printers are good enough that you can make a print that looks close enough if you're a clerk working at the cash register for the past 7 hours and there is a solid line.
Crumple them. New, creaseless bills could be hard to tell; crumple them and the fabric components of the real thing come through and feel a lot different.
On the contrary, I think a software company deciding is much better than a legislator or judge. At least in the former case, you can choose a different piece of software. In the latter you have to leave the country.
I don't see why they can't use 64 bit addresses. 32 bit machines seem to get by with it just fine.
For people who aren't familiar with computer architecture, I present a quick summary. If you are, skip to the last paragraph.
There are other ways of working around the instruction word problem to use the full 64 bits. I don't know if these are done in CPUs used in PCs, even 32-bit ones, but I'll briefly describe how the MIPS deals with 32-bit addresses with 32-bit words since the MIPS is what I know from my computer archetecture course:
Data memory accesses (standard loads and stores, lw and sw in ASM) are done by base address + offset. The base address is held in a 32-bit register, the offset is a 16-bit immediate field (sign extended to get 32-bits). The assumption is that you're probably not going to have much need to access memory more than 32K in either direction that much, and if you do, you can do some extra adds before the load/store. To get the base address into a register typically requires two instructions, one for the upper 16 bits and one for the lower 16 bits (lui followed by addi or ori or xori).
All memory access is done through a load or store word (byte, half-word, etc.) instruction. Thus we only have to deal with the instructions themselves.
MIPS (and I think most machines) has two types of goto-like instructions: branches and jumps. Branches are done conditionally and locally, jumps are unconditional and possibly long distance. Note that here I'm talking word addresses, not byte addresses. Instructions are word aligned, so storing the byte address wastes 2 bits.
The condition in the branches requires extra bits. For instance, there is a bne command that compares two registers and branches to the specified address if they are equal. The "specified address" is actually the sum of the current instruction (30-bits) and the 16-bit immediate. Thus the program can branch 32,768 instructions (131,072 bytes) in either direction.
For jumps, 26 bits of the instruction are used for the jump address. It is assumed that the program will be not jumping beyond the nearest multiple of 2^26 words. The high-order 4 bits of the current instruction are appended to the 26-bit immediate to get the resulting jump address.
In the case that these don't work, there are still ways of getting around it. With three instructions, it is possible to load a full 32-bit address (via lui and ori as above) into a register then jump to that address directly.
Now, these are all more annoying than having enough space in the instruction for the full address I'll agree, but to me the tradeoff of reducing address space for a slightly simpler time at programming isn't worth it. Now, maybe I'm just being silly here, after all, even I'll admit we probably won't see over 4TB memories in the same machine as these processors, but can someone explain why this tradeoff was really made?
It's an impressive difference from an engineering and "look what they did" standpoint, but it isn't that big of a difference from a "let's put this in my pocket and walk around" standpoint.
While the parent did get the IIS/ISS thing wrong, whether or not they have the same letters is up for interpretation.
I would tend toward the interpretation that they *do* have the same letters: "IIS" has the letters {I, S} "ISS" has the letters {I, S} The two sets are the same.
IMO, I think to get at what you're saying you would need to say they have the same letters with the same frequencies.
But this is just a matter of opinion, and I don't think my way is "right" and your way is "wrong".
With a leak that goes at a quarter PSI/month, probably would be too much trouble. You'd probably need to turn off equipment, seal off the compartment, let the air settle for a hour or two, light the match, then watch. Oh, and you can't actually go in, or you'll create air currents that will completely mask the leak's. Even breathing would probably be plenty. So you need to do the last two stoeoesAremotely.
A) Apollo 1 was pure O2, ISS is standard air mixture (~21% O2).
B) Part of the reason the fire on Apollo 1 was so bad was strong convection currents. In space, there's no convection. This is why NASA didn't have a problem running flights in pure O2 once they were out of the atmosphere. (The O2 N2 mixtures in the Apollo missions were replaced with pure O2 at ~5psi as the N2 leaked away.)
Because there are as many people using Linux as there are Windows?
The motivation to port to Windows comes so that the enormous market share Windows has can be tapped. I don't know figures, but I wouldn't be at all surprised if Apple has a larger market share on desktops than Linux, so there's actually not much motivation there at all.
Well, yes. If MS knows that it's on there and thinks they should be open, they'll be open. Other Windows services -- MS update for instance -- will be open too. They don't know that you have, for instance, Mozilla installed, so they can't really open it. (I guess they could have a big long list of programs that if they are on there would be open by default, but that just seems silly, even to me.)
"That's because the high-gain antenna wasn't deployed until last night, and these pictures were sent before then."
Yeah, but the high gain was deployed shortly after midnight, and they are still waiting until when Odyssey passes over as a relay around 3:30am tonight. I know why the first photos aren't the pancams, but I still don't get why we don't have them yet. (I'm impatient:-p)
I wouldn't be surprised if this is changed... It's too clunky, even for MS.
Most likely it's that they have been working on the actual code for the firewall itself and have so far built just enough of a UI that it is operational and usable for people who know what they are doing.
These are just preliminary. The good stuff--super hires color photos--are sitting in memory on the rover waiting for a chance to send. The next chance will come at about 3:30am EST tonight when Odyssey makes another overhead pass to act as a relay.
I'm not quite sure why they can't send it directly via Spirit's high-gain antenna, but my guess is that it has to do with the fact that the transmission speed is about that of a 33.6 modem.
You can go ahead and make fun of our probes that crash. Especially the metric/imperial one. That deserves to be made fun of. I've done it many times myself.
*BUT*
Don't you DARE make fun of or insult ANYONE whose lives are lost in ANY space program ANYWHERE.
The "problem" with opening the source to Win98 (as seen from MS's position) is twofold;
1) Much of the source in 98 is probably used elsewhere; you're essentially opening the source to a fair portion of later versions, something they don't want.
2) It would strengthen the OS movement in general, something they don't want.
Slashdot Mirror
What? We're talking about detecting counterfeits, not making them...
People rarely check.
But there are several features of the bill that will light up under UV, like the security strip.
Actually, counterfeiting is essentially more of a problem now than ever before. There are all the nice features of real bills, but seriously how often have you actually see someone hold a bill under a blacklight, or use a pen, or look for the watermark when accepting the bill. My answer: maybe a dozen times in the last couple years.
At the same time, printers are good enough that you can make a print that looks close enough if you're a clerk working at the cash register for the past 7 hours and there is a solid line.
Crumple them. New, creaseless bills could be hard to tell; crumple them and the fabric components of the real thing come through and feel a lot different.
On the contrary, I think a software company deciding is much better than a legislator or judge. At least in the former case, you can choose a different piece of software. In the latter you have to leave the country.
I don't see why they can't use 64 bit addresses. 32 bit machines seem to get by with it just fine.
For people who aren't familiar with computer architecture, I present a quick summary. If you are, skip to the last paragraph.
There are other ways of working around the instruction word problem to use the full 64 bits. I don't know if these are done in CPUs used in PCs, even 32-bit ones, but I'll briefly describe how the MIPS deals with 32-bit addresses with 32-bit words since the MIPS is what I know from my computer archetecture course:
Data memory accesses (standard loads and stores, lw and sw in ASM) are done by base address + offset. The base address is held in a 32-bit register, the offset is a 16-bit immediate field (sign extended to get 32-bits). The assumption is that you're probably not going to have much need to access memory more than 32K in either direction that much, and if you do, you can do some extra adds before the load/store. To get the base address into a register typically requires two instructions, one for the upper 16 bits and one for the lower 16 bits (lui followed by addi or ori or xori).
All memory access is done through a load or store word (byte, half-word, etc.) instruction. Thus we only have to deal with the instructions themselves.
MIPS (and I think most machines) has two types of goto-like instructions: branches and jumps. Branches are done conditionally and locally, jumps are unconditional and possibly long distance. Note that here I'm talking word addresses, not byte addresses. Instructions are word aligned, so storing the byte address wastes 2 bits.
The condition in the branches requires extra bits. For instance, there is a bne command that compares two registers and branches to the specified address if they are equal. The "specified address" is actually the sum of the current instruction (30-bits) and the 16-bit immediate. Thus the program can branch 32,768 instructions (131,072 bytes) in either direction.
For jumps, 26 bits of the instruction are used for the jump address. It is assumed that the program will be not jumping beyond the nearest multiple of 2^26 words. The high-order 4 bits of the current instruction are appended to the 26-bit immediate to get the resulting jump address.
In the case that these don't work, there are still ways of getting around it. With three instructions, it is possible to load a full 32-bit address (via lui and ori as above) into a register then jump to that address directly.
Now, these are all more annoying than having enough space in the instruction for the full address I'll agree, but to me the tradeoff of reducing address space for a slightly simpler time at programming isn't worth it. Now, maybe I'm just being silly here, after all, even I'll admit we probably won't see over 4TB memories in the same machine as these processors, but can someone explain why this tradeoff was really made?
Leave it at +5, troll ;-)
Funny?
It's an impressive difference from an engineering and "look what they did" standpoint, but it isn't that big of a difference from a "let's put this in my pocket and walk around" standpoint.
"Don't you think 2504 for 4Go is MUCH better ?"
Let's see, two and a half times the rumored price for twice the space... I think my answer to that has to be "NO, I don't think it's much better."
While the parent did get the IIS/ISS thing wrong, whether or not they have the same letters is up for interpretation.
I would tend toward the interpretation that they *do* have the same letters:
"IIS" has the letters {I, S}
"ISS" has the letters {I, S}
The two sets are the same.
IMO, I think to get at what you're saying you would need to say they have the same letters with the same frequencies.
But this is just a matter of opinion, and I don't think my way is "right" and your way is "wrong".
With a leak that goes at a quarter PSI/month, probably would be too much trouble. You'd probably need to turn off equipment, seal off the compartment, let the air settle for a hour or two, light the match, then watch. Oh, and you can't actually go in, or you'll create air currents that will completely mask the leak's. Even breathing would probably be plenty. So you need to do the last two stoeoesAremotely.
A) Apollo 1 was pure O2, ISS is standard air mixture (~21% O2).
B) Part of the reason the fire on Apollo 1 was so bad was strong convection currents. In space, there's no convection. This is why NASA didn't have a problem running flights in pure O2 once they were out of the atmosphere. (The O2 N2 mixtures in the Apollo missions were replaced with pure O2 at ~5psi as the N2 leaked away.)
Because there are as many people using Linux as there are Windows?
The motivation to port to Windows comes so that the enormous market share Windows has can be tapped. I don't know figures, but I wouldn't be at all surprised if Apple has a larger market share on desktops than Linux, so there's actually not much motivation there at all.
Well, yes. If MS knows that it's on there and thinks they should be open, they'll be open. Other Windows services -- MS update for instance -- will be open too. They don't know that you have, for instance, Mozilla installed, so they can't really open it. (I guess they could have a big long list of programs that if they are on there would be open by default, but that just seems silly, even to me.)
"That's because the high-gain antenna wasn't deployed until last night, and these pictures were sent before then."
:-p)
Yeah, but the high gain was deployed shortly after midnight, and they are still waiting until when Odyssey passes over as a relay around 3:30am tonight. I know why the first photos aren't the pancams, but I still don't get why we don't have them yet. (I'm impatient
I wouldn't be surprised if this is changed... It's too clunky, even for MS.
Most likely it's that they have been working on the actual code for the firewall itself and have so far built just enough of a UI that it is operational and usable for people who know what they are doing.
Probably just like any other program; denied until it is told otherwise.
These are just preliminary. The good stuff--super hires color photos--are sitting in memory on the rover waiting for a chance to send. The next chance will come at about 3:30am EST tonight when Odyssey makes another overhead pass to act as a relay.
I'm not quite sure why they can't send it directly via Spirit's high-gain antenna, but my guess is that it has to do with the fact that the transmission speed is about that of a 33.6 modem.
You can go ahead and make fun of our probes that crash. Especially the metric/imperial one. That deserves to be made fun of. I've done it many times myself.
*BUT*
Don't you DARE make fun of or insult ANYONE whose lives are lost in ANY space program ANYWHERE.
There are good explanations here, but a picture is worth a thousand words, so... http://www.vision3d.com/3views.html
"8th RULE: If this is your first night at FIGHT CLUB, your robot HAS to fight."
What's fight club? I thought I was at robot club.
The "problem" with opening the source to Win98 (as seen from MS's position) is twofold;
1) Much of the source in 98 is probably used elsewhere; you're essentially opening the source to a fair portion of later versions, something they don't want.
2) It would strengthen the OS movement in general, something they don't want.
No one would make $300 million dollar films.
They aren't; see this comment and, more specifically, its replies.