2-1/2 inch heavy trackball. Buttons that don't click, they "thwack". The older models look like something from NORAD, but the newer ones...well, look like something from NORAD. Usually sold by companies that sell equipment for people with disabilities and the Armed Forces. With a standard mouse or trackball, you keep too many muscles rigid and tense to be precise. With a P+G, you just move the massive trackball. Remember Missile Command? Imagine it on your desk...oh, and around $300-$400 USD each. I've rebuilt one of mine twice (its from 19...87?).
Take the shape in one direction and you get an infinite number of facets; a smooth circle in practice. Here's a challenge: Go the other direction and reduce the number of sides. A triangle is easy; the hills on the road are a bit steeper and the curve is more pointed (quickly decreasing radius). Go one more side removed and you have...a two-sided wheel and.. triangular hills that are equal on a side to the "radius" of the "wheel". Now, that I'd like to see in action!
I was talkin' about convergence of phone and PC. Or, more accurately, phone and headless equipment, "Son". I want to reduce the stuff I carry with me, not add to it.
"Digital sounds" as you put it and mp3's are not the same. I'm talking about playback that does not have to squeeze through the frequency response of a piezo speaker the size of a pea.
If I'm hooking up to a box, I don't want to have to carry an irda converter, nor do I wish to equip every system I touch with one. USB serial and 9-pin serial are what you need. C'mon, would you really have a wireless back door on your router?
"Bluetooth" keyboard? Yeah, I want to have to balance and arrange a phone, a wireless keyboard AND your irda-converter while stretching behind some disorganized rack...
In short, if you think about the intended purpose, I want a TOOL, not a TOY...
Now I will be able to have my pda-cell-phone-mp3 player that will play an mp3 instead of a ringtone (hell, the DSP is already in the phone today, right?)...
Add an usb plug to one side and 128MB of flash and I will need nothing more...
I can just see the ad copy:
"It's my phone, my music, my OS"
How far can the convergence go? Where will it lead the hardware design? Once you go linux, you will of course need an RS-232 port so you can use your phone to set up a router or headless box...
Ah, the crux-question: Who will be the first to solve the cell-phone/QWERTY keyboard hardware design problem in such a way that users can actually create shell scripts without getting RSI in their stylus fingers?
Wasn't the IMP the equivelent of today's NIC, as a homogenized network interface for the PDP's, the Generals, etc. so that each location would have to figure out by themselves how to connect their computer to the IMP, but the IMPs would all have a standard interface to the network.
I dunno, it seems that if both shuttles failed due to materials failures -right on schedule- that the design worked just as it was supposed to. Also, if it was a materials failure, that means it wasn't a procedure nor operator nor process failure, so that's pretty good.
Topic-morph:
What really concerns me about the Shuttle program is that I can remember back when the first gliding tests were being done some of the aerospace and aviation journals mentioned two things that are *very* interesting but seemed to have been lost to time:
1. The Shuttle design has always included an ejectable crew compartment. The structure is still there in the skeleton. However, to save cost and increase the payload capacity slightly, the ejection mechanisms were removed and with them the capablity. Have you noticed that in both shuttle failures the crew cabin remained intact longer than anything else and were you disturbed by the reports of how long the crew survived after the disintegrations? Granted, the ejection was only for use during the slow part of takeoff, but if it could be used at any other unforeseeable time...
2. NASA released a facinating report regarding the solid rocket boosters. Each puts large amounts of...Chlorine into the upper atmosphere where it can do the most harm. But it's cheaper than liquid fuel. The estimates, by NASA at the time, were that the ozone layer could be depleted by booster-Chlorine alone after another 200 launches...guess how many there have been?
Tracking by product will be MOOT in five years
on
The Trouble with RFID
·
· Score: 1
...when, within five years, all state's driver's licenses and ID cards will have an RFID embedded in them (else they lose Federal Highway funds, etc., etc.) Think about it; the police will scan you without leaving their vehicle, reducing the threat to them by a large margin or at least helping to identify the level of threat. ID doesn't match the vehicle? Suspect car theft. Driver, but no RFID? Driving without a license (or worse). The safety it will bring the officers will trump individual rights to privacy.
Now, once all DL's and ID's have RFID's, there will be positive tracking of every individual, by name in our "free" nation. Oh, don't forget the new plan to give all to-be-legal immigrants "special" ID cards also. Don't want to leave anyone out, you know...
Mea Culpa, Mea Culpa, Mea Culpa. The 80386 was the first of the line to be 32-bit.
Gotta disagree with you on the PPro's 16-bit performance, though. Back in the day (1996? 1998?), a 200Mhz Pentium with 256k cache would consistantly outperform a PPro-200 / 256k on 16-bit applications, and the benchmarks confirmed this. The problem was compounded by anyone using Win 9.x which was partly still 16-bit. I always assumed that it was due to the CISC-to-RISC translation being optimized for 32-bit, as per Intel's official statements, but I believe there is another posting in this thread that points to an actual bug in the prediction algorithms that was fixed along with the "reoptimization".
Curiosity question for slashdotters: Have the compromises in optimization for 16-bit performance ever been dropped from the Pentium line, now that 16-bit is rarely seen? Or was the slowdown completely due to the bug?
The Pentium Pro line was abandoned in a sense, but you are correct that the lineage and the CPU core lived on in the PII, etc. The CPU+L2 on the same die format was dropped due to high losses -they couldn't be tested until they were completed and at that point it was too late. That was solved by the SEC and a whole new name. Smart move on Intel's part, in hindsight.
The 80286, '386, '486, Pentium, PPro, PII, PIII and P4 are all 32 bit
AMD's latest CPU's are 64 bit
...all are in the x86 family. The later CPU's have a few extra instructions, math copro's, larger L1 cache, on-die L2, more address lines and a few other bells and whistles but include (almost) all of the instructions of the very first x86 chip and can run code written for the first IBM PC 5150.
Intel's 64-bit Itanium, which has been a bit of a marketing disappointment for them, on the other hand, uses an entirely new instruction set. Might as well be a PowerPC, Transmeta or Motorola CPU as far as the base of installed and legacy software is concerned. Does anyone remember the Pentium Pro? Similar tactic; it's 16-bit performance was awful as it essentially had to emulate a 16-bit processor set internally. There was still too much old code in use and the PPro line was abandoned after only one or two clock increases (introduced at 150 or 160Mhz, killed off at only 200Mhz). Intel had to backpedal on that one and the Pentuim II was a PPro with MMX and adequate 16-bit performance.
Watch for the Intel x86 64-bit CPU sometime in the next year!
They've been doing these tricks for decades, using tape, to delay while recording, record while playing, etc. "Tape delay", I believe it is called. Gives the broadcast manager control over what actually goes out over the air as a "live" broadcast. Slightly more manual process, but the process (and all the things they can do with it) are the same. Take it a step further and think back to the "time delayed" Olympic broadcasts. They did everything that ReplayTV and the others can do, but they did it slower and it was between the camera and the viewer (who viewed it "live", but time delayed), not the cable box stream and the viewer.
Seems to me, and I am no lawyer, that just about every time the word "video" was mentioned in the text of AB301, the word "television" preceded it. In the few instances that it did not, it was referring to a television video screen, used to show full screen images of scenes. Could we be confusing what we consider video (VGA, QVGA, SVGA *and* NTSC or PAL) and what normal everyday people consider video (Broadcast, DVD and VCS)?
Off-the-cuff curiosity question and yes, at risk of appearing ignorant, but what better way to start out the new year? Is it possible that for 28 years we added a second even though only, say,.93 second was needed and now, for five years, no full second is needed? A second is a pretty big full unit of time for the calculation to always come out even.
During part of the late '90's, I was installing custom accounting software on clients' systems and the one thing we did to vastly increase the reliability of their shiny new windows 98 systems was to turn Active Desktop *off*. Only some systems showed the problem, not all, but of those that started crashing with seemingly random-module 0E errors, turning it off permanently stopped the problem. Wicked little bastard that one was to correlate. The closest I got to finding the exact cause before leaving the company was that it didn't happen until C or VB libraries were installed as part of an application or driver install. Mismatched set of IE libs + Active Desktop (IE based) = 0E. At my next job I hunted it down. It was the "auto packaging" part of VB that was only including *some* of the IE.dll's, not the complete set (only those that were actually called), and by chance they often did not match those already on the system. I insisted that if any IE.dll's were to be included in our SETUP.EXE's, a complete version-matched set had to go. Worked like a charm, but we also had the luxury of insisting in our licensing that the systems be for our software's use only. Due to the clients' security practices, they were not to be networked, so possibly killing IE was not a problem. That way if an 0E problem cropped up due to someone else's install, a reinstall of our app including the.dll's solved it 98% of the time at little support cost. The other 2% were due to printer driver app installs; we usually backed them down to a more "generic" driver included with the OS so that the fancy printer applications were not also installed. Long and short of it is that Active Destop + VB-autopackaged installs pretty much ensured a flakey system. May.dll-hell rest in peace and never again walk this earth.
Did you not read about the improvements to SMP and the last minute rewrite of the "patch" to IDE-SCSI? How much of your work have YOU given away this holliday season?
Slashdot Mirror
2-1/2 inch heavy trackball. Buttons that don't click, they "thwack". The older models look like something from NORAD, but the newer ones...well, look like something from NORAD. Usually sold by companies that sell equipment for people with disabilities and the Armed Forces. With a standard mouse or trackball, you keep too many muscles rigid and tense to be precise. With a P+G, you just move the massive trackball. Remember Missile Command? Imagine it on your desk...oh, and around $300-$400 USD each. I've rebuilt one of mine twice (its from 19...87?).
Take the shape in one direction and you get an infinite number of facets; a smooth circle in practice. Here's a challenge: Go the other direction and reduce the number of sides. A triangle is easy; the hills on the road are a bit steeper and the curve is more pointed (quickly decreasing radius). Go one more side removed and you have...a two-sided wheel and.. triangular hills that are equal on a side to the "radius" of the "wheel". Now, that I'd like to see in action!
"Digital sounds" as you put it and mp3's are not the same. I'm talking about playback that does not have to squeeze through the frequency response of a piezo speaker the size of a pea.
If I'm hooking up to a box, I don't want to have to carry an irda converter, nor do I wish to equip every system I touch with one. USB serial and 9-pin serial are what you need. C'mon, would you really have a wireless back door on your router?
"Bluetooth" keyboard? Yeah, I want to have to balance and arrange a phone, a wireless keyboard AND your irda-converter while stretching behind some disorganized rack...
In short, if you think about the intended purpose, I want a TOOL, not a TOY...
Now I will be able to have my pda-cell-phone-mp3 player that will play an mp3 instead of a ringtone (hell, the DSP is already in the phone today, right?)... Add an usb plug to one side and 128MB of flash and I will need nothing more... I can just see the ad copy: "It's my phone, my music, my OS" How far can the convergence go? Where will it lead the hardware design? Once you go linux, you will of course need an RS-232 port so you can use your phone to set up a router or headless box... Ah, the crux-question: Who will be the first to solve the cell-phone/QWERTY keyboard hardware design problem in such a way that users can actually create shell scripts without getting RSI in their stylus fingers?
How much memory does your NIC have?
I dunno, it seems that if both shuttles failed due to materials failures -right on schedule- that the design worked just as it was supposed to. Also, if it was a materials failure, that means it wasn't a procedure nor operator nor process failure, so that's pretty good. Topic-morph: What really concerns me about the Shuttle program is that I can remember back when the first gliding tests were being done some of the aerospace and aviation journals mentioned two things that are *very* interesting but seemed to have been lost to time: 1. The Shuttle design has always included an ejectable crew compartment. The structure is still there in the skeleton. However, to save cost and increase the payload capacity slightly, the ejection mechanisms were removed and with them the capablity. Have you noticed that in both shuttle failures the crew cabin remained intact longer than anything else and were you disturbed by the reports of how long the crew survived after the disintegrations? Granted, the ejection was only for use during the slow part of takeoff, but if it could be used at any other unforeseeable time... 2. NASA released a facinating report regarding the solid rocket boosters. Each puts large amounts of...Chlorine into the upper atmosphere where it can do the most harm. But it's cheaper than liquid fuel. The estimates, by NASA at the time, were that the ozone layer could be depleted by booster-Chlorine alone after another 200 launches...guess how many there have been?
...when, within five years, all state's driver's licenses and ID cards will have an RFID embedded in them (else they lose Federal Highway funds, etc., etc.) Think about it; the police will scan you without leaving their vehicle, reducing the threat to them by a large margin or at least helping to identify the level of threat. ID doesn't match the vehicle? Suspect car theft. Driver, but no RFID? Driving without a license (or worse). The safety it will bring the officers will trump individual rights to privacy. Now, once all DL's and ID's have RFID's, there will be positive tracking of every individual, by name in our "free" nation. Oh, don't forget the new plan to give all to-be-legal immigrants "special" ID cards also. Don't want to leave anyone out, you know...
Gotta disagree with you on the PPro's 16-bit performance, though. Back in the day (1996? 1998?), a 200Mhz Pentium with 256k cache would consistantly outperform a PPro-200 / 256k on 16-bit applications, and the benchmarks confirmed this. The problem was compounded by anyone using Win 9.x which was partly still 16-bit. I always assumed that it was due to the CISC-to-RISC translation being optimized for 32-bit, as per Intel's official statements, but I believe there is another posting in this thread that points to an actual bug in the prediction algorithms that was fixed along with the "reoptimization".
Curiosity question for slashdotters: Have the compromises in optimization for 16-bit performance ever been dropped from the Pentium line, now that 16-bit is rarely seen? Or was the slowdown completely due to the bug?
The Pentium Pro line was abandoned in a sense, but you are correct that the lineage and the CPU core lived on in the PII, etc. The CPU+L2 on the same die format was dropped due to high losses -they couldn't be tested until they were completed and at that point it was too late. That was solved by the SEC and a whole new name. Smart move on Intel's part, in hindsight.
8088/8086's were 16 bit
The 80286, '386, '486, Pentium, PPro, PII, PIII and P4 are all 32 bit
AMD's latest CPU's are 64 bit
...all are in the x86 family. The later CPU's have a few extra instructions, math copro's, larger L1 cache, on-die L2, more address lines and a few other bells and whistles but include (almost) all of the instructions of the very first x86 chip and can run code written for the first IBM PC 5150.
Intel's 64-bit Itanium, which has been a bit of a marketing disappointment for them, on the other hand, uses an entirely new instruction set. Might as well be a PowerPC, Transmeta or Motorola CPU as far as the base of installed and legacy software is concerned. Does anyone remember the Pentium Pro? Similar tactic; it's 16-bit performance was awful as it essentially had to emulate a 16-bit processor set internally. There was still too much old code in use and the PPro line was abandoned after only one or two clock increases (introduced at 150 or 160Mhz, killed off at only 200Mhz). Intel had to backpedal on that one and the Pentuim II was a PPro with MMX and adequate 16-bit performance.
Watch for the Intel x86 64-bit CPU sometime in the next year!
They've been doing these tricks for decades, using tape, to delay while recording, record while playing, etc. "Tape delay", I believe it is called. Gives the broadcast manager control over what actually goes out over the air as a "live" broadcast. Slightly more manual process, but the process (and all the things they can do with it) are the same. Take it a step further and think back to the "time delayed" Olympic broadcasts. They did everything that ReplayTV and the others can do, but they did it slower and it was between the camera and the viewer (who viewed it "live", but time delayed), not the cable box stream and the viewer.
Seems to me, and I am no lawyer, that just about every time the word "video" was mentioned in the text of AB301, the word "television" preceded it. In the few instances that it did not, it was referring to a television video screen, used to show full screen images of scenes. Could we be confusing what we consider video (VGA, QVGA, SVGA *and* NTSC or PAL) and what normal everyday people consider video (Broadcast, DVD and VCS)?
1. Bootable USB port
2. Combined 802.11x and 128MB flash device
3. Linux configured as wireless router on the flash.
4. Plugin, reboot. Instant WiFi access bypassing the corporate firewalls.
Can you hear the SysAdmins screaming? Let's hope everyone remembers to disable booting from USB in the BIOS and then locks it down...
Off-the-cuff curiosity question and yes, at risk of appearing ignorant, but what better way to start out the new year? Is it possible that for 28 years we added a second even though only, say, .93 second was needed and now, for five years, no full second is needed? A second is a pretty big full unit of time for the calculation to always come out even.
During part of the late '90's, I was installing custom accounting software on clients' systems and the one thing we did to vastly increase the reliability of their shiny new windows 98 systems was to turn Active Desktop *off*. Only some systems showed the problem, not all, but of those that started crashing with seemingly random-module 0E errors, turning it off permanently stopped the problem. Wicked little bastard that one was to correlate. The closest I got to finding the exact cause before leaving the company was that it didn't happen until C or VB libraries were installed as part of an application or driver install. Mismatched set of IE libs + Active Desktop (IE based) = 0E. At my next job I hunted it down. It was the "auto packaging" part of VB that was only including *some* of the IE .dll's, not the complete set (only those that were actually called), and by chance they often did not match those already on the system. I insisted that if any IE .dll's were to be included in our SETUP.EXE's, a complete version-matched set had to go. Worked like a charm, but we also had the luxury of insisting in our licensing that the systems be for our software's use only. Due to the clients' security practices, they were not to be networked, so possibly killing IE was not a problem. That way if an 0E problem cropped up due to someone else's install, a reinstall of our app including the .dll's solved it 98% of the time at little support cost. The other 2% were due to printer driver app installs; we usually backed them down to a more "generic" driver included with the OS so that the fancy printer applications were not also installed. Long and short of it is that Active Destop + VB-autopackaged installs pretty much ensured a flakey system. May .dll-hell rest in peace and never again walk this earth.
Did you not read about the improvements to SMP and the last minute rewrite of the "patch" to IDE-SCSI? How much of your work have YOU given away this holliday season?