What i really need is a way to remember how i solved a particular programming problem six months ago.
Yep. More than once, I've written a small script to solve a problem. This tickled my memory, so I use find . -print | xargs grep -i FunctionName, and I find the same script I wrote months or years ago. Sometimes I've even used the same variable names and data structures.
Cool, Sys V Unix had a similar problem. New pids were created monotonically and stored in as 16 bit values (or 15). When the pids on a system rolled, things started going wonky.
An old 3B20 wasn't fast enough to roll the pids for any useful processes before the machine crashed or rebooted, but when the port of Sys V Unix to AIX was done, the new machines were much faster and pid allocation was rearranged so that they would be reused much quicker. That was 20 years ago, so I don't remember the code, but it was an easy fix.
There is actually only one real reason given and that is that if you reboot after some services ceased working, you might end up with a unbootable machine.
The other reason given, is that by using the bandaid of rebooting, you don't find the actual problem.
In the classroom, finding the "actual" problem is a useful exercise. In the real world, sometime business reasons trump finding the problem, a bandaid is exactly the most cost effective solution. Unfortunately, in the real world, many times you don't know which was the best solution until the postmortem.
I've found the pre-emptive reboot will trigger hidden system problems, but at a time when you're actually ready for them, rather than at a time when they happen spontaneously ( 2:30 in the morning ).
Right on the money, but what do you say to management that insists that reboots be done during non-business hours, like 2:30am?
Basically, our management is paranoid. They put controls in place, but never really trust those controls. Since the illusion of control is more important than actual control, real risk isn't mitigated.
"Party X should not be able to pay party Y to cause party Z to get less than what party Z paid for."
That makes the core anti-free-market agenda a bit more clear to me.
Yes, that's a better description.
I'm very much a "Truth in Advertising" "Free Market" type. (TAFM) Just like food manufacturers are required to tell you what's in food, ISPs should be required to tell you the minimum speed you're paying for, and how much it costs. (My parents tried to get a Verizon rep to give them a final cost for FIOS after fees, taxes, and expiration of the trial period. The rep couldn't do it.) You shouldn't have to buy a product to find out it's crap and not buy it again. You should always be able to compare what everyone's offering, and make an informed decision.
IT's a truth in advertising thing. If my ISP is actually selling me a 76Kbps connection that bursts to 768Kbps, make them sell it that way. (Low number first and foremost)
The ISPs should have to spell out exactly what they are selling and what it costs. Selling much, much more than you can deliver is bait and switch.
Today 1mbps is fine for light browsing, but if the local ISP sells that to everyone, then youtube adds HD video, and everyone tried to watch it at the same time, they're going to be glitches.
DSL is being advertised in some markets as being better than the higher speed cable because you're supposed to get that bandwidth all the time, unlike cable that's shared with your neighbors.
The key is that everyone should get what they pay for. If I pay for 768kbps, then I should get at least 768kbps. If google wants to pay extra, then I'm ok with google gettting to me at 2mbps, but not with google paying my ISP so that yahoo only comes to me at 250kbps.
I should get what I pay for.
Google should get what they pay for.
Party X should not be able to pay for party Y to get less than what has been paid for.
So, if I am a professional, and all my peers are professionals, I can't go to school?
Back when I was in school, my peer group were mostly paid student consultants, and we often got help from the professional staff. (And we helped them sometimes)
If students get help from a professor that's not the one teaching their class, are they cheating?
What are you using for comparison? When has any government been completely open?
Great project that would put the science in "Political Science". Divide a college campus into "States". Have one state run without secrets, and the other in total secrecy. See which does better. (Note that state secrets and personal privacy are different beasts)
This would make an excellent in class calculation for a Geophysics lecture. Just how big a set neodymium magnets (one north, one south) would it take to override the earth's magnetic field? If installed, would the core actually align with it? How bad would it be if you installed them backwards?
I can't actually try it because it's not tax time.
I use tax software to do my federal, so going to a separate web site to do the state seems inefficient, but it might be a good idea for other people.
Virginia even sent out a notice that it saves them beaucoup money when people use online tax filing. Unfortunately, they aren't willing to pay companies like Intuit or Kiplinger's so that the cost to the filer is zero. I always print and send my tax forms so that it maximizes their cost, and minimizes mine. As long as my refund isn't large, and I file early, my lost interest on the money doesn't exceed what it would cost to eFile using the tax software.
So, with that analogy in mind, if the government required you to let anyone and everyone use your bathroom - i.e. physically occupy it - and the requirement was permanent - i.e. anyone can use your bathroom, forever - then it would be a taking.
Some municipalities do regulate which businesses (mostly food) must have public restrooms.
This is good news for AIs all over the web. A web site that is ADA compliant is much easier for a program to navigate than one that requires screen scraping and OCR. The bad news is that botnets can also run AIs, and making government information more available will make things easier for scammers.
These guys, http://www.the-impossible-project.com/, purchased some old Polaroid manufacturing equipment and are making new film, mostly for artists. Perhaps someone will do the same for Kodachrome.
My great-grandfather used a camera that put negatives onto glass plates. (4x6 or 5x7) My mother had these contact printed. The resolution is astounding.
The more rural the area, the fewer route choices, and thusly the less importance a GPS due to the lack of choice.
That's true for the great plains states, but in places like rural West Virginia, you can easily get lost in the Mountains, which are full of twisty, winding roads. (Cue "Dueling Banjos")
If you password is long enough, you could allow a certain amount of errors, and still have a strong authentication system. I.e. allow one dropped space, two swapped charaters or two switched capitalizations, wouldn't give the attacker too much help. i.e. Is a search space of 52^35th practially any better than 52^42nd?
You could freeze everyone's account with the newly banned password, but I can't imagine the users would be too keen on that either.
or, add it to a special list, then only move the bad passwords from the special list to the banned list when no-one is using them. That would keep the bad passwords in the system too long.
The service simply counts how many times any user on the service chooses a given password. When more than a small number of users pick a password, the password is banned and no one else is allowed to choose it.
This system was thought of, and rejected many years ago. If you let user x know that a given password is in use, he now has a password that he can try against everyone else's account.
You could try to randomly reject "good" passwords as well, but that would piss off your users.
Slashdot Mirror
What i really need is a way to remember how i solved a particular programming problem six months ago.
Yep. More than once, I've written a small script to solve a problem. This tickled my memory, so I use find . -print | xargs grep -i FunctionName, and I find the same script I wrote months or years ago. Sometimes I've even used the same variable names and data structures.
It was a Russian Inwention. (:-)
http://www.supermemo.com/articles/power.htm
Dr. Piotr A. Wozniak (Actually he's Polish)
Cool, Sys V Unix had a similar problem. New pids were created monotonically and stored in as 16 bit values (or 15). When the pids on a system rolled, things started going wonky.
An old 3B20 wasn't fast enough to roll the pids for any useful processes before the machine crashed or rebooted, but when the port of Sys V Unix to AIX was done, the new machines were much faster and pid allocation was rearranged so that they would be reused much quicker. That was 20 years ago, so I don't remember the code, but it was an easy fix.
There is actually only one real reason given and that is that if you reboot after some services ceased working, you might end up with a unbootable machine.
The other reason given, is that by using the bandaid of rebooting, you don't find the actual problem.
In the classroom, finding the "actual" problem is a useful exercise. In the real world, sometime business reasons trump finding the problem, a bandaid is exactly the most cost effective solution. Unfortunately, in the real world, many times you don't know which was the best solution until the postmortem.
If that were true, you'd reboot after every change in /etc/rc* (or whatever's appropriate for your system)
Instead, most places just have a scheduled reboot. (Which, frankly, works, but still seems like a bad idea)
I've found the pre-emptive reboot will trigger hidden system problems, but at a time when you're actually ready for them, rather than at a time when they happen spontaneously ( 2:30 in the morning ).
Right on the money, but what do you say to management that insists that reboots be done during non-business hours, like 2:30am?
Basically, our management is paranoid. They put controls in place, but never really trust those controls. Since the illusion of control is more important than actual control, real risk isn't mitigated.
For that price I hope he at least gets a crown and a scepter to go along with the wicked-ass title
Well, yes, if he builds them out of Lego. (:-)
"Party X should not be able to pay party Y to cause party Z to get less than what party Z paid for."
That makes the core anti-free-market agenda a bit more clear to me.
Yes, that's a better description.
I'm very much a "Truth in Advertising" "Free Market" type. (TAFM) Just like food manufacturers are required to tell you what's in food, ISPs should be required to tell you the minimum speed you're paying for, and how much it costs. (My parents tried to get a Verizon rep to give them a final cost for FIOS after fees, taxes, and expiration of the trial period. The rep couldn't do it.) You shouldn't have to buy a product to find out it's crap and not buy it again. You should always be able to compare what everyone's offering, and make an informed decision.
IT's a truth in advertising thing. If my ISP is actually selling me a 76Kbps connection that bursts to 768Kbps, make them sell it that way. (Low number first and foremost)
The ISPs should have to spell out exactly what they are selling and what it costs. Selling much, much more than you can deliver is bait and switch.
Today 1mbps is fine for light browsing, but if the local ISP sells that to everyone, then youtube adds HD video, and everyone tried to watch it at the same time, they're going to be glitches.
DSL is being advertised in some markets as being better than the higher speed cable because you're supposed to get that bandwidth all the time, unlike cable that's shared with your neighbors.
The key is that everyone should get what they pay for. If I pay for 768kbps, then I should get at least 768kbps. If google wants to pay extra, then I'm ok with google gettting to me at 2mbps, but not with google paying my ISP so that yahoo only comes to me at 250kbps.
I should get what I pay for.
Google should get what they pay for.
Party X should not be able to pay for party Y to get less than what has been paid for.
So, if I am a professional, and all my peers are professionals, I can't go to school?
Back when I was in school, my peer group were mostly paid student consultants, and we often got help from the professional staff. (And we helped them sometimes)
If students get help from a professor that's not the one teaching their class, are they cheating?
Prove it.
What are you using for comparison? When has any government been completely open?
Great project that would put the science in "Political Science". Divide a college campus into "States". Have one state run without secrets, and the other in total secrecy. See which does better. (Note that state secrets and personal privacy are different beasts)
Quick! Is there a Geophysicist in the house???
This would make an excellent in class calculation for a Geophysics lecture. Just how big a set neodymium magnets (one north, one south) would it take to override the earth's magnetic field? If installed, would the core actually align with it? How bad would it be if you installed them backwards?
It seems to still be there: https://www.individual.tax.virginia.gov/VTOL/IndLogin.seam
I can't actually try it because it's not tax time.
I use tax software to do my federal, so going to a separate web site to do the state seems inefficient, but it might be a good idea for other people.
Virginia even sent out a notice that it saves them beaucoup money when people use online tax filing. Unfortunately, they aren't willing to pay companies like Intuit or Kiplinger's so that the cost to the filer is zero. I always print and send my tax forms so that it maximizes their cost, and minimizes mine. As long as my refund isn't large, and I file early, my lost interest on the money doesn't exceed what it would cost to eFile using the tax software.
So, with that analogy in mind, if the government required you to let anyone and everyone use your bathroom - i.e. physically occupy it - and the requirement was permanent - i.e. anyone can use your bathroom, forever - then it would be a taking.
Some municipalities do regulate which businesses (mostly food) must have public restrooms.
This is good news for AIs all over the web. A web site that is ADA compliant is much easier for a program to navigate than one that requires screen scraping and OCR. The bad news is that botnets can also run AIs, and making government information more available will make things easier for scammers.
These guys, http://www.the-impossible-project.com/, purchased some old Polaroid manufacturing equipment and are making new film, mostly for artists. Perhaps someone will do the same for Kodachrome.
My great-grandfather used a camera that put negatives onto glass plates. (4x6 or 5x7) My mother had these contact printed. The resolution is astounding.
It was just an idle story, but http://idle.slashdot.org/story/10/07/14/1235220/Given-Truth-the-Misinformed-Believe-Lies-More?art_pos=1 " Idle: Given Truth, the Misinformed Believe Lies More"
Propaganda is extremely effective, and the truth doesn't stop it. Reverse propaganda might help, but then you become the enemy.
The more rural the area, the fewer route choices, and thusly the less importance a GPS due to the lack of choice.
That's true for the great plains states, but in places like rural West Virginia, you can easily get lost in the Mountains, which are full of twisty, winding roads. (Cue "Dueling Banjos")
If you password is long enough, you could allow a certain amount of errors, and still have a strong authentication system. I.e. allow one dropped space, two swapped charaters or two switched capitalizations, wouldn't give the attacker too much help. i.e. Is a search space of 52^35th practially any better than 52^42nd?
You could freeze everyone's account with the newly banned password, but I can't imagine the users would be too keen on that either.
or, add it to a special list, then only move the bad passwords from the special list to the banned list when no-one is using them. That would keep the bad passwords in the system too long.
The service simply counts how many times any user on the service chooses a given password. When more than a small number of users pick a password, the password is banned and no one else is allowed to choose it.
This system was thought of, and rejected many years ago. If you let user x know that a given password is in use, he now has a password that he can try against everyone else's account.
You could try to randomly reject "good" passwords as well, but that would piss off your users.
You might be able to find an anti-glare overlay similar to the screen protectors used for handhelds.
Much more interesting are the few two legged dogs that manage to get around.