Its easy to pick three soft targets and talk about strategies to defeat them, but in practice any automated defeat only works on really large volume sites (signing up for gmail for example) where the method is rather static.
Employing multiple Captcha types randomly makes this harder. Employing consecutive captcha pages of differing types where the user must remember things from one page to the next is harder still.
Filtering snow out of images is fairly easy given the advancements in image processing and character recognition available today.
Recognizing word puzzles is quite a different thing. The "mental effort" captcha's magic is not actually in the effort required, but merely in the fact that it is difficult to write bots to answer the questions that appear in verbal form. Especially when the verbal form has a twist.
I've seen some fairly unique examples:
"Multiply the number of fingers on your right hand by the number of fingers on your left hand and enter the number of noses you have below."
"Add the number of feet in a mile to the number of inches in a foot and enter the gender of your birth mother below."
"Please select the Wrong answers from the following list instead of the Correct answers:" (followed by some question and answers to check.
These are not hard individually, but when there is a large number of them rotated thru a page only visual inspection by humans can reliably get them right.
Yes, you can pay some 3 cent an hour person to read and record them all, but new ones can be generated just as quickly.
>AT&T (and the rest) have either been way over-charging for their voice service for a very long time, or that they are way undercharging (taking a loss) on their data service.
I don't see that math when I look at my bill.
We have multiple lines of service with ATT. Each additional after the first costs 9 bucks. However we pay a 30 dollar data plan charge PER LINE for smartphone usage. (iPhone, Blackberry, etc).
So I pay much more for the data plan than I do for the minutes.
The research is largely paid for by the Australian tax payer, and the ownership of the fruits of their labor should also rest there, rather than beinb plowed back into building an ivory tower that suffers from out of control growth in the face of shrinking budgets,
In the US we are a little suspect of agencies like this, (although we do have some).
We tend to prefer the NASA, DARPA, TVA model of publicly funded organizations, where thousands of ideas are released to the general public and find their way quickly onto the market.
> On top of that, it would mandate the destruction of any wrongly obtained information discovered
Why not toss out any court case based on such wrongly discovered info? Why not criminal prosecutions of those issuing the letters from which information was wrongly discovered? Why not remove the muzzeling of anyone issued such letters? After all if they were improperly issued letters in the first place any inducement to STFU about such a letter must also be wrong.
Tune her radio to the BBC. They paid the fee already, and as a tax payer, she is part owner of the BBC like every other citizen. Let the PRS sort it out with the BBC.
It can not be that you have to pay a fee to transit the material by radio, and also a fee to convert the radio to sound waves without it ALSO requiring a fee to convert the sound waves to ear-pressure changes.
> It wouldn't affect the average home user, but think about the implications for I.T. Staff.
Said Staff would be rejoicing because proprietary company documents would stop walking out the door on every disgruntled employee's thumb drive.
I'm sure that corporate IT is not by far the largest thumb drive consumer, let alone cameras, SD cards, Cell phones, etc.
You can always format the thumb drive again with what ever file system you want.
Besides, you make a mountain out of a mole hill. Microsoft would distribute the driver for you, just like they do for hundreds of other third party drivers.
>The settlement neither implies that Microsoft patents are valid nor that TomTom's products were or are infringing.
These guys are whistling past the grave yard. They weren't party to the suit, just putting their oar in, too little too late.
As far as TomTom was concerned the patents were valid. Cry all you want about cost of a defense, the point remains that a patent that is too expensive to defend against is just as effective as a valid one.
Once TomTom caves and removes the offending long name support, who else will stand up?
Don't confuse Fat with either Fat32 or Long file names. You can have a small partition large enough to hold your drivers on the dongle.
Many devices in the Win98 era came with a mini-cd with the drivers on it. I have 4 of 5 of these that all used the same drivers, and once you installed one it would work for many devices (cameras mostly in my case).
(And if "she don't want you installing shit" walk away. Its not your machine).
As for the Vista problem, this is why a signed module developed by and maintained by someone other than Joe Average Hacker is needed. Perhaps Red Hat or Novell or some such.
Look, we are not talking about Apple here, we are talking about Microsoft. Apple would just say NO and that's the end of it. (Ok, maybe a bit of a troll, but Apple is _very_ controlling).
Microsoft includes hundreds of drivers developed by third parties, and they would include an EXT driver too if the source was reliable and the inclusion would not encumber them with the GPL.
Note that they never sere alleging violation of FAT32 patents, merely long name support.
When you say "Nobody supports anything else" I presume you mean in small devices, cameras, phones, thumb drives, etc.
But use of other file systems on such devices would be no worse than distributing drivers (like was done for Win98) and using EXT2/3/4.
A high quality windows driver released free and clear for that pretty much flips the coin on Microsoft. There are several of these out there: ( http://www.fs-driver.org/http://ext2fsd.sourceforge.net/ ) but its not clear that they are robust enough for device vendors to ship millions of units with.
Why does anybody get alarmed when there's a patent suit between two PRODUCTIVE companies?
Because Microsoft's intent is to kill Linux with patents, by getting each vendor (TomTom, Red Hat, Novell, etc) to include patented crap into Linux. Death by a thousand paper cuts.
How successful they will be with this is anyone's guess.
At least TomTom removed the Long File Name support rather than licensing that, which would probably have brought an immediate law suit from FSF, as it would have been admission that long file names on FAT are patentable. They side stepped the issue. Nothing is solved.
Don't be so sure this spat is about patents mentioned. To do so is to focus on the hammer, instead of the nail they are trying to drive in Linux's coffin.
So that essentially says that Microsoft won, and TomTom got their butts kicked. Removing functionality does not sound like cross licensing, it sounds like knuckling under and making a show of trading patent licensing on some other meaningless issues to save face.
Does this also mean that Long File Names may be infringing in Samba?
Does this mean that TOMTOM is in trouble with FOSS, since they can't pass on these cross licensed features free and clear?
But all of those versions run like crap on a small machine.
Turning off Aero helps some, but the machine was underspeced for ANY version of Vista, and the manufacturer should have realized that.
By simply bundling in another 512meg of memory the manufacture could steal a march on their competition. Yet they chose to knuckle under to Microsoft and Intel.
> This saying was born from an era when government work was head-and-shoulders above par,
Government work above par? When has that EVER been the case?
Good enough for Government work ALWAYS refers to work done by low-bidders which was just barely good enough to pass inspection by un-caring government purchasing agents.
I'm saying that we need publicly available free systems upon which to build, so tat every project does not have to start from ground zero re-writing the world.
Nokia, Palm, Motorola, LG, all run out and develop a phone OS from scratch, then they do it all over again in two years.
The points made are perfectly valid, although the conclusions are somewhat misguided.
The reasons schools are not keeping up with the industry complexity is because the complexity is out of control.
We are losing the ability to build these things (complex buildings, software projects, networks,etc) with the entry level help that was usable in the past.
Its not a fault of the schools. Its a fault of the constant piling on of complexity while continuing to write/build everything from the ground up.
This is why projects like Linux and Android are so important.
Its going to be necessary to either standardize building blocks and automate large subprocesses, or stretch college to age 30.
I've hired CS grads. It takes a year to un-teach them so that the can be come useful enough to find and fix a simple bug.
Slashdot Mirror
So the dolt at the reception desk can lock out an entire company from registering at your site just because the all come from the same natted address?
User Friendly much?
CAPTCHA is not doomed, it it merely evolving.
Its easy to pick three soft targets and talk about strategies to defeat them, but in practice any automated defeat only works on really large volume sites (signing up for gmail for example) where the method is rather static.
Employing multiple Captcha types randomly makes this harder. Employing consecutive captcha pages of differing types where the user must remember things from one page to the next is harder still.
Filtering snow out of images is fairly easy given the advancements in image processing and character recognition available today.
Recognizing word puzzles is quite a different thing. The "mental effort" captcha's magic is not actually in the effort required, but merely in the fact that it is difficult to write bots to answer the questions that appear in verbal form. Especially when the verbal form has a twist.
I've seen some fairly unique examples:
"Multiply the number of fingers on your right hand by the number of fingers on your left hand and enter the number of noses you have below."
"Add the number of feet in a mile to the number of inches in a foot and enter the gender of your birth mother below."
"Please select the Wrong answers from the following list instead of the Correct answers:" (followed by some question and answers to check.
These are not hard individually, but when there is a large number of them rotated thru a page only visual inspection by humans can reliably get them right.
Yes, you can pay some 3 cent an hour person to read and record them all, but new ones can be generated just as quickly.
Look, even if what you say is true, you are still asked to believe you can get 160Mbits on copper, AND you can get it for $20 bucks.
I doubt there really exists such a box at that price. I doubt you could get a tech into the dwelling to hook it up for that price.
Someone is glossing over a subscription funded subsidy here, and the comparison is just bogus.
>AT&T (and the rest) have either been way over-charging for their voice service for a very long time, or that they are way undercharging (taking a loss) on their data service.
I don't see that math when I look at my bill.
We have multiple lines of service with ATT. Each additional after the first costs 9 bucks. However we pay a 30 dollar data plan charge PER LINE for smartphone usage. (iPhone, Blackberry, etc).
So I pay much more for the data plan than I do for the minutes.
> Sue the Judge
Good luck with that.
Remember folks that this is a government funded and sponsored organization.
http://en.wikipedia.org/wiki/CSIRO
The research is largely paid for by the Australian tax payer, and the ownership of the fruits of their labor should also rest there, rather than beinb plowed back into building an ivory tower that suffers from out of control growth in the face of shrinking budgets,
In the US we are a little suspect of agencies like this, (although we do have some).
We tend to prefer the NASA, DARPA, TVA model of publicly funded organizations, where thousands of ideas are released to the general public and find their way quickly onto the market.
There is no evidence that CO2 is warming the planet.
http://www.hawaiireporter.com/story.aspx?1ad63198-0a1f-4b5b-8fb8-96df07d70d41
> On top of that, it would mandate the destruction of any wrongly obtained information discovered
Why not toss out any court case based on such wrongly discovered info?
Why not criminal prosecutions of those issuing the letters from which information was wrongly discovered?
Why not remove the muzzeling of anyone issued such letters? After all if they were improperly issued letters in the first place any inducement to STFU about such a letter must also be wrong.
Tune her radio to the BBC. They paid the fee already, and as a tax payer, she is part owner of the BBC like every other citizen. Let the PRS sort it out with the BBC.
It can not be that you have to pay a fee to transit the material by radio, and also a fee to convert the radio to sound waves without it ALSO requiring a fee to convert the sound waves to ear-pressure changes.
Does the PRS require listeners to be licensed?
You have a problem with freedom of choice?
> It wouldn't affect the average home user, but think about the implications for I.T. Staff.
Said Staff would be rejoicing because proprietary company documents would stop walking out the door on every disgruntled employee's thumb drive.
I'm sure that corporate IT is not by far the largest thumb drive consumer, let alone cameras, SD cards, Cell phones, etc.
You can always format the thumb drive again with what ever file system you want.
Besides, you make a mountain out of a mole hill. Microsoft would distribute the driver for you, just like they do for hundreds of other third party drivers.
>The settlement neither implies that Microsoft patents are valid nor that TomTom's products were or are infringing.
These guys are whistling past the grave yard.
They weren't party to the suit, just putting their oar in, too little too late.
As far as TomTom was concerned the patents were valid. Cry all you want about cost of a defense, the point remains that a patent that is too expensive to defend against is just as effective as a valid one.
Once TomTom caves and removes the offending long name support, who else will stand up?
Don't confuse Fat with either Fat32 or Long file names. You can have a small partition large enough to hold your drivers on the dongle.
Many devices in the Win98 era came with a mini-cd with the drivers on it. I have 4 of 5 of these that all used the same drivers, and once you installed one it would work for many devices (cameras mostly in my case).
(And if "she don't want you installing shit" walk away. Its not your machine).
As for the Vista problem, this is why a signed module developed by and maintained by someone other than Joe Average Hacker is needed. Perhaps Red Hat or Novell or some such.
Look, we are not talking about Apple here, we are talking about Microsoft. Apple would just say NO and that's the end of it. (Ok, maybe a bit of a troll, but Apple is _very_ controlling).
Microsoft includes hundreds of drivers developed by third parties, and they would include an EXT driver too if the source was reliable and the inclusion would not encumber them with the GPL.
I was thinking about SMBFS portion (which is not technically Samba).
Note that they never sere alleging violation of FAT32 patents, merely long name support.
When you say "Nobody supports anything else" I presume you mean in small devices, cameras, phones, thumb drives, etc.
But use of other file systems on such devices would be no worse than distributing drivers (like was done for Win98) and using EXT2/3/4.
A high quality windows driver released free and clear for that pretty much flips the coin on Microsoft. There are several of these out there: ( http://www.fs-driver.org/ http://ext2fsd.sourceforge.net/ ) but its not clear that they are robust enough for device vendors to ship millions of units with.
Why does anybody get alarmed when there's a patent suit between two PRODUCTIVE companies?
Because Microsoft's intent is to kill Linux with patents, by getting each vendor (TomTom, Red Hat, Novell, etc) to include patented crap into Linux. Death by a thousand paper cuts.
How successful they will be with this is anyone's guess.
At least TomTom removed the Long File Name support rather than licensing that, which would probably have brought an immediate law suit from FSF, as it would have been admission that long file names on FAT are patentable. They side stepped the issue. Nothing is solved.
Don't be so sure this spat is about patents mentioned. To do so is to focus on the hammer, instead of the nail they are trying to drive in Linux's coffin.
Read some of th links at the top of this page: http://linux.slashdot.org/article.pl?sid=09/03/05/1624221
So that essentially says that Microsoft won, and TomTom got their butts kicked. Removing functionality does not sound like cross licensing, it sounds like knuckling under and making a show of trading patent licensing on some other meaningless issues to save face.
Does this also mean that Long File Names may be infringing in Samba?
Does this mean that TOMTOM is in trouble with FOSS, since they can't pass on these cross licensed features free and clear?
But all of those versions run like crap on a small machine.
Turning off Aero helps some, but the machine was underspeced for ANY version of Vista, and the manufacturer should have realized that.
By simply bundling in another 512meg of memory the manufacture could steal a march on their competition. Yet they chose to knuckle under to Microsoft and Intel.
Where have you been for the last decade.
There is an International Space Station orbiting the earth. You might have heard of this when you weren't trolling here on SlashDot.
If not, or you have been off-planet for the last many years, here is a link for you:
http://quest.nasa.gov/events/cotf/webquest/index.html
> This saying was born from an era when government work was head-and-shoulders above par,
Government work above par?
When has that EVER been the case?
Good enough for Government work ALWAYS refers to work done by low-bidders which was just barely good enough to pass inspection by un-caring government purchasing agents.
It has never signified excellence.
Oh, yes, lets tweak this patent just a tad and see if we can extend it for another 20 years.
So if a production one costs 10 dollars, 3 time 10 is $30,
Then, because its less, we have to subtract his costs of $30 from the production cost of $10, it costs him minus 20 dollars to build each one?
You mean it was 1/3 the cost of a production unit.
There is no such thing as "3 times less" of anything.
I'm saying that we need publicly available free systems upon which to build, so tat every project does not have to start from ground zero re-writing the world.
Nokia, Palm, Motorola, LG, all run out and develop a phone OS from scratch, then they do it all over again in two years.
How did this get modded troll?
The points made are perfectly valid, although the conclusions are somewhat misguided.
The reasons schools are not keeping up with the industry complexity is because the complexity is out of control.
We are losing the ability to build these things (complex buildings, software projects, networks,etc) with the entry level help that was usable in the past.
Its not a fault of the schools. Its a fault of the constant piling on of complexity while continuing to write/build everything from the ground up.
This is why projects like Linux and Android are so important.
Its going to be necessary to either standardize building blocks and automate large subprocesses, or stretch college to age 30.
I've hired CS grads. It takes a year to un-teach them so that the can be come useful enough to find and fix a simple bug.
How could it be GPL AND Patented?