It is possible to reliably terminate aramid (Kevlar) fibres at around 99% efficiency, in production.
I know, since my workmates developed the idea, and I did the calculations for the necessary hardware.
You need careful control of the way that the hardware and the fibres move under load. For the electrical engineers here, it is essentially an impedance mismatch problem.
We replaced a technique that could reliably hit 65%, in the same space.
Well, one reason to choose a nom de plume like that is so people can make jokes.
Oddly enough there was nothing wrong with the connection, it was all my PC. My point was that their helpdesk script could never get them tot he point of identifying the real problem.
Now they'll be using Linux themselves do you think they'll support Linux at BigPond?
Nothing. As I said, the PC had two mistakes in the comms/modem setup - I can't remeber what one was, the other was that about 4 menus deep there was a limit on max throughput speed.
My point was that after a total of 5 hours on the phone (and probably 10-15 hours all up) EVERY suggestion telstra made was completely irrelevant to the actual problem.
The true cost of a landline is reputedly more than that. A more sensible billing arrangement given their real cost structure would be a higher landline rental, and then a flat fee per hour for accessing the network.
T2 was the gummint, although Telstra's projections must have been pretty rosy to support that valuation. Hmm, given the fairly hefty dividend payouts, and the franking rate, I suspect that even T2 investors might at least be close to breaking even, if you ignore the cost of capital.
I think the standard bribe is about 100 bucks for your mobile contract, same for ISP, same for phone.
That's not bad, since my bill runs at less than 100/month all up (no mobile).
Optus' bribe was effectively free local calls, and better capped rates for IDD. Not as attractive in the short term as the Telstra one, but I went with it after a few run-ins with Telstra's help desk.
Broadband? They can't even manage a dial up account.
I tried for one month to get decent (better than 1kbytes/second) speed out of my modem, and was on the phone to the Testra Bigpond helpdesk for perhaps 5 hours all up.
I basically had two settings incorrect on my PC, they did not offer any useful advice on finding them, they were obsessed with getting me to reinstall windows, explorer, their cd and modem strings. Which I did, many times. Finally I found what was wrong, predictably enough it was none of the above.
So at the end of the first month I cancelled my account and switched all my telecom stuff to Optus, who have been fine. d/l speed is 5k, as you'd expect.
People seem to gloss over the office suites. OO is NOT a functional replacement for Excel. It provides the core, but the migration path for most heavy users would be uncertain. The problem is that Excel is used in such wildly different ways that if you asked 10 different users I expect you'd get 10 different feature sets that they use occasionally.
Now, a whole separate issue is should people in an engineering organisation even be analysing using unmaintainable code written for Excel (whether VBA, or just cell commands)? I don't think so.
If I were running the joint I would insist that all design calculations are done using a proper mathematical worksheet, like Mathcad (or Scilab, I guess).
I run W2k at work, as do many thousands of other people on our network.
Our sysadmins must have done something right - we had NO downtime (from a user's perspective) due to the recent fun and games. I did have to reboot my machine with a new fix or virus definition, so that was 5 minutes lost at the start of one day.
So I think this 2 or 3 days a year lost time is an overestimate, by a factor of 10 or so.
To be honest spam is a bigger concern on our network than virusses. They took the filters off our email service because it was (reputedly) blocking real messages. In a Catch 22 scenario, I manually delete many messages from the summary page, thereby deleting legitimate messages by hand!
My home machine runs NT, and using free versions of commercial software (Agnitum Outpost and AVG anti-virus) I have had no problems with sobig, msblaster, or any other worm or virus. That I know of. Touch wood. It may be that my ISP has a good handle on this, I don't know, as a user the problem is more or less invisible to me although I am seeing a lot of ICMP connection requests which are bounced by my firewall.
I've been using Mozilla 1.3 and 1.4 under NT for 6 months. With the exception of the lack of the googlebar, I can't think of a reason to switch back to IE. I haven't had a single crash in the mail client, and the browser seems at least as stable as IE did. It might close down unexpectedly once a month or so, but will happily restart without a reboot.
I hadn't really thought about this in a while - once the stability and the features are there, you stop noticing the browser.
Oh, one really irritating thing, ctrl shift o for open a new web location. Hate it.
"How about we spend that much money and figure out how to actually educate our children, feed our hungry, and stop messing up all the time?"
Have you any evidence that putting money into those projects would do any good?
Experience to date suggests that the best way to make these things happen is to have a succesful capitalist democratic society. The Japanese may well be doing the right thing, by jump-starting a new industry. Good on 'em.
Oh, I though his explanation of how a turbo 'worked' took the biscuit.
"For fifteen thousand dollars, Norwood rigged the Ferrari with a turbo system that would activate when Carmack floored the gas pedal. It was a ballsy bit of hacking, and Carmack immediately felt a kinship with the veteran racing man. "
Also, I don't quite think paying $15k to someone else to do the dirty work is quite my definition of a cool hack.
...and that seems to be a good way of building a cluster with the performance of a 2 GHz P4.
Seriously, I remember that for general purpose cpu limited tasks computing power varies with the square root of the number of processors, and directly with the speed of the processors.
1 vs 2 processors, on an Octane (note that a multi cpu mobo has both advantages and disadvantages compared with two networked single cpu computers)
If your task can be easily chunked then you might see a 7 GHz equivalent machine, but if it cannot be paralleled efficiently you might be stuck with performance that is worse than a single 750 machine.
And watch them struggle to dial out on the modem in order to get on the web the second time.
Dull show!
Go on, find one clue in the Mandrake 9.1 distro that you need to use kddd.
Getting onto the Internet should be a single click on the desktop or at worst two menus down in the "start" menu. Anything else is just plain silly. IMO.
Well they'd probably hit the "Save s Web Page" command in the file menu.
If you don't mind the size it is pretty painless, and often works.
Re:Multi-Channel motion control
on
Mirror, Mirror
·
· Score: 1
I wonder if he considered using a solenoid (coil) to get the same effect? If the ball were sprung loaded then the position of the ball would be determined solely by the voltage - no need for fancy control systems.
If you have fuel cell in your house, and electrolyse hydrogen to power it you will need MORE electricity than you do today, since neither the electrolysis or the fuel cell are 100% efficient.
Convergence - I like the idea, but, given a non-fractal programming language, how can you compress more and more information (code) into a given number of bits? There has to be some sort of Shannon's law equivalent that says ultimately that if you have x inputs and y outputs you need a certain size of code just to contain, or generate, all the possible i/o states.
I'm not saying that current code is space-optimised by the way.
I know you are correct, but even a geek should know that tanks were invented/during/ WW1, and were called tanks at the the time, as a security measure. 1915 would be a good date, depending on your exact definition - most of the early problems were to do with tracks, engines, and transmissions. Once they had a suffciently mobile steel box sticking the guns in was a no-brainer.
Is there any intrinsic reason why Linux would be less prone to virusses and worms than Windows? Or is it just that nobody bothers to write them for Linux because the population of Linux computers is too small to matter, like Macs?
Slashdot Mirror
It is possible to reliably terminate aramid (Kevlar) fibres at around 99% efficiency, in production.
I know, since my workmates developed the idea, and I did the calculations for the necessary hardware.
You need careful control of the way that the hardware and the fibres move under load. For the electrical engineers here, it is essentially an impedance mismatch problem.
We replaced a technique that could reliably hit 65%, in the same space.
Well, one reason to choose a nom de plume like that is so people can make jokes.
Oddly enough there was nothing wrong with the connection, it was all my PC. My point was that their helpdesk script could never get them tot he point of identifying the real problem.
Now they'll be using Linux themselves do you think they'll support Linux at BigPond?
Nothing. As I said, the PC had two mistakes in the comms/modem setup - I can't remeber what one was, the other was that about 4 menus deep there was a limit on max throughput speed.
My point was that after a total of 5 hours on the phone (and probably 10-15 hours all up) EVERY suggestion telstra made was completely irrelevant to the actual problem.
The true cost of a landline is reputedly more than that. A more sensible billing arrangement given their real cost structure would be a higher landline rental, and then a flat fee per hour for accessing the network.
T2 was the gummint, although Telstra's projections must have been pretty rosy to support that valuation. Hmm, given the fairly hefty dividend payouts, and the franking rate, I suspect that even T2 investors might at least be close to breaking even, if you ignore the cost of capital.
I think the standard bribe is about 100 bucks for your mobile contract, same for ISP, same for phone.
That's not bad, since my bill runs at less than 100/month all up (no mobile).
Optus' bribe was effectively free local calls, and better capped rates for IDD. Not as attractive in the short term as the Telstra one, but I went with it after a few run-ins with Telstra's help desk.
Broadband? They can't even manage a dial up account.
I tried for one month to get decent (better than 1kbytes/second) speed out of my modem, and was on the phone to the Testra Bigpond helpdesk for perhaps 5 hours all up.
I basically had two settings incorrect on my PC, they did not offer any useful advice on finding them, they were obsessed with getting me to reinstall windows, explorer, their cd and modem strings. Which I did, many times. Finally I found what was wrong, predictably enough it was none of the above.
So at the end of the first month I cancelled my account and switched all my telecom stuff to Optus, who have been fine. d/l speed is 5k, as you'd expect.
If you read the link he was fined $10k, and had to do 400 hours community service.
People seem to gloss over the office suites. OO is NOT a functional replacement for Excel. It provides the core, but the migration path for most heavy users would be uncertain. The problem is that Excel is used in such wildly different ways that if you asked 10 different users I expect you'd get 10 different feature sets that they use occasionally.
Now, a whole separate issue is should people in an engineering organisation even be analysing using unmaintainable code written for Excel (whether VBA, or just cell commands)? I don't think so.
If I were running the joint I would insist that all design calculations are done using a proper mathematical worksheet, like Mathcad (or Scilab, I guess).
I run W2k at work, as do many thousands of other people on our network.
Our sysadmins must have done something right - we had NO downtime (from a user's perspective) due to the recent fun and games. I did have to reboot my machine with a new fix or virus definition, so that was 5 minutes lost at the start of one day.
So I think this 2 or 3 days a year lost time is an overestimate, by a factor of 10 or so.
To be honest spam is a bigger concern on our network than virusses. They took the filters off our email service because it was (reputedly) blocking real messages. In a Catch 22 scenario, I manually delete many messages from the summary page, thereby deleting legitimate messages by hand!
My home machine runs NT, and using free versions of commercial software (Agnitum Outpost and AVG anti-virus) I have had no problems with sobig, msblaster, or any other worm or virus. That I know of. Touch wood. It may be that my ISP has a good handle on this, I don't know, as a user the problem is more or less invisible to me although I am seeing a lot of ICMP connection requests which are bounced by my firewall.
I've been using Mozilla 1.3 and 1.4 under NT for 6 months. With the exception of the lack of the googlebar, I can't think of a reason to switch back to IE. I haven't had a single crash in the mail client, and the browser seems at least as stable as IE did. It might close down unexpectedly once a month or so, but will happily restart without a reboot.
I hadn't really thought about this in a while - once the stability and the features are there, you stop noticing the browser.
Oh, one really irritating thing, ctrl shift o for open a new web location. Hate it.
Hi Jody,
Goalseek and Solver are very different in their approach, frankly I'd find it hard to use a spreadsheet that offered less than Solver, these days.
I realise that Solver is not perfect (anyone who needs to can buy a better Solver from the original developer) http://www.frontsys.com/premplus.htm.
"How about we spend that much money and figure out how to actually educate our children, feed our hungry, and stop messing up all the time?"
Have you any evidence that putting money into those projects would do any good?
Experience to date suggests that the best way to make these things happen is to have a succesful capitalist democratic society. The Japanese may well be doing the right thing, by jump-starting a new industry. Good on 'em.
Oh, I though his explanation of how a turbo 'worked' took the biscuit.
"For fifteen thousand dollars, Norwood rigged the Ferrari with a turbo system that would activate when Carmack floored the gas pedal. It was a ballsy bit of hacking, and Carmack immediately felt a kinship with the veteran racing man. "
Also, I don't quite think paying $15k to someone else to do the dirty work is quite my definition of a cool hack.
...and that seems to be a good way of building a cluster with the performance of a 2 GHz P4.
n 1v s2cpu.html
n ce .html
. ht ml
Seriously, I remember that for general purpose cpu limited tasks computing power varies with the square root of the number of processors, and directly with the speed of the processors.
1 vs 2 processors, on an Octane (note that a multi cpu mobo has both advantages and disadvantages compared with two networked single cpu computers)
http://futuretech.mirror.vuurwerk.net/spec95oct
If your task can be easily chunked then you might see a 7 GHz equivalent machine, but if it cannot be paralleled efficiently you might be stuck with performance that is worse than a single 750 machine.
http://coeweb.eb.uah.edu/licos/spectra/performa
has some figures indicating that doubling the size of the cluster reduced run times by around 30%-40%, for some benchmarks (by eyeball)
http://www.climate.unibe.ch/cluster/performance
Has some fun graphs showing what can happen if your network protocol absorbs too much CPU time.
And watch them struggle to dial out on the modem in order to get on the web the second time.
Dull show!
Go on, find one clue in the Mandrake 9.1 distro that you need to use kddd.
Getting onto the Internet should be a single click on the desktop or at worst two menus down in the "start" menu. Anything else is just plain silly. IMO.
Well they'd probably hit the "Save s Web Page" command in the file menu.
If you don't mind the size it is pretty painless, and often works.
I wonder if he considered using a solenoid (coil) to get the same effect? If the ball were sprung loaded then the position of the ball would be determined solely by the voltage - no need for fancy control systems.
Who modded this up?
If you have fuel cell in your house, and electrolyse hydrogen to power it you will need MORE electricity than you do today, since neither the electrolysis or the fuel cell are 100% efficient.
Yes, they'll have a handy doorstop to keep the doors open and let some air in.
Which bit of the concept 'electricity' don't you understand?
IANAP
Convergence - I like the idea, but, given a non-fractal programming language, how can you compress more and more information (code) into a given number of bits? There has to be some sort of Shannon's law equivalent that says ultimately that if you have x inputs and y outputs you need a certain size of code just to contain, or generate, all the possible i/o states.
I'm not saying that current code is space-optimised by the way.
ffs
/during/ WW1, and were called tanks at the the time, as a security measure. 1915 would be a good date, depending on your exact definition - most of the early problems were to do with tracks, engines, and transmissions. Once they had a suffciently mobile steel box sticking the guns in was a no-brainer.
I know you are correct, but even a geek should know that tanks were invented
I read at -10
Is there any intrinsic reason why Linux would be less prone to virusses and worms than Windows? Or is it just that nobody bothers to write them for Linux because the population of Linux computers is too small to matter, like Macs?
/.!=usenet
How did you train a cockroach to play fetch?