Since spamming is all about getting the same text into lots of forums, simply have an anti-spam bot read random other forums. If they have messages which contain the same or nearly same text than what was posted on your forum, they're either spam or copypasta troll, and should be removed. Thing is there are a huge number of forums and a huge number of spambots so unless you have a very big sample or you know sites that are getting hit by the same spambots as you you are likely to miss a lot this way. You can't really use google to check because it doesn't update fast enough.
Alternatively or additionally, you could traverse any links and check if they contain any telltale signs of spam. Indeed you could and the spammer could take measures to hide those telltale signs.
I still think if you are a small site your best bet is to roll your own anti-spam measures and operate on the principle that it won't be worth it for the spammers to figure them out.
The best security I've seen on a sign-up form was "if you're a human, please leave this field blank". Bots tend to fill in all fields, so this already goes a long way towards filtering them out. It works for your small site sure, the problem is once you (or sites running your software if you write software for others to use) grow beyond a certain size the spammers will start targetting you/your software. Once they do that you need a much stronger system to stand a chance of keeping them at bay.
Something like KittenAuth [thepcspy.com] has been recommended, and still seems to be the best answer in my opinion. They don't give many details but presumablly there are a limited number of pictures in thier database each tagged with what animal it is. How hard would it be for a spammer to get hold of a significant proportion of the database?
To detect humans, wouldn't it be easier and less costly, and perhaps even more effective, to hold a large database of questions that are readable and solvable only by humans? I guess the question becomes how large is large. If you reuse tests too much then the spammers will just build their own database of soloutions.
Using a database of non computer created challenges is a good idea but there needs to be a system for keeping that database topped up. Recapatcha for example picks out words from old books that thier OCR software fails on and uses them to test your users.
Normally they give the user two words, one for which they know the answer already (that is at least two people have given the same answer for it) and one they don't. but if they see failures from an IP they switch to giving the user two words that they know the answer for already.
In general if you are relatively small I think a custom soloution is one of your best defenses against spammers. At the end of the day spamming is about getting as many people as possible to see your spam as possible for as little effort as possible. Investigating a contact form just to spam one small forum or a contact form for a few people at a company just isn't worth it.
* YUV->RGB color conversion (there isn't even a single right answer to this, it's subjective) There is a right answer if you know what definitions or RGB and YUV you are working to. Of course if you are displaying on a random monitor than it will never be right for everyone because of monitor variation so you probablly want to let the videophiles tweak this.
Audio language selection (how do you write code to tell if the guy is speaking in french as opposed to spanish?) You make sure the different "language" tracks in your test data are different enough that an automated system can recognise them.
does anyone have any idea how big PS3 games actually are? Sure the console may support 50 gigs on one disc but I would have thought with multiplatform games at least they would be about the same size on all the platforms they supported.
media companies don't want thier products (whether books, movies or games) to appear in a trickle. A trickle would mean retailers who manage to get stock are likely to charge high prices (which leaves customers with the impression the game is very expensive) and means spoilers and pirate copies are likely to be widely spread before most people get thier copy of the product.
So they decide on a release pricepoint, estimate the demand at that pricepoint and then organise a release so that the release day lines up with the marketing campaign and with the availibility of sufficiant product that supplies won't run out. Of course this all screws up if someone leaks the game first.
The problem is the kernel team aren't cooperating:(. They won't keep the -486 kernel slim enough and they won't sanction the creation of a seperate kernel flavor just for the floppy installer to use it.
With etch the kernel had grown to the point that they had to kick everything that wasn't absoloutly essential (including USB floppy support) from the boot floppy.
With lenny a couple of bad things happened. Firstly the UPX recompressor stopped working for current kernels. And then just as there was talk of fixing that the lenny kernel grew to the point that it just wasn't feasible to put it on a floppy at all.
The changelog presented in aptitude should be up to date as well. It's pulled from the package. That seems unlikely to me since it would require the package manager to download the whole package to show the changelog.
The shortened release cycle seems to be working pretty well. Afaict thier current strategy is to aim for 18 months and be happy with 24. They achieved that with etch and it seems likely they will achieve it with lenny.
Afaict cards will only work in OS-X if they are recognised by the EFI firmware and they will only be recognised by apples EFI firmware if they have the right firmware on the card. In other words you can't just go look up a hardware compatibility list and buy a card with the right chipset you have to buy a special apple card with the special firmware on it.
Sometimes it is possible to reflash cards if you can find a firmware dump from a mac card with the exact same hardware as the PC card you are dealing with but doing so almost certainly voids any warranty on the card and probbablly has a nontrivial risk of bricking the card. IIRC it also requires a PC to do the reflashing.
I always thought that it was simply the sound pressure level that caused hearing loss I think the point is that with outside noise sealed off you can enjoy the music at much lower volume because you aren't trying to drown out the ambiant noise.
Unfortunately it means those who DO want thier music to replace ambient noise end up listening to it far too loud. Either because they are ignorant of the existance of ones that do block outside noise or because they find them unreasonablly expensive.
It seems what is needed is
1: clear education that there are two types of earbuds, ones designed to block outside noise and ones designed not to and that drowning out outside noise by cranking up the volume is a very bad idea. 2: earbuds that block outside noise widely availible at a reasonable price.
but I also feel that I have to point out that they're close to being as upgradeable as the PC in each product subcategory. True, the thing is apple doesn't make anything directly comparable to normal PCs so those who want to run OS-X "legally" are pushed into buying a weird system.
The mac mini is far less upgradable than a normal low end PC. The imac is far less upgradeable than a normal midrange PC.
As for the mac pro that is highly upgradable but it is very expensive initaly and ram and graphics upgrades require special parts (the ram seems comparable in price to server ram, the graphics cards are far more expensive than equivilent PC cards and there is much less choice).
Afaict the convetional way to do hexadecimal on a 7 segment display is to use capital A,lower case b, capital C, lowercase d , capital E and capital F . Theese along with the numbers can be represeted in a reasonale fasion on a 7 segment display without ambiguity.
Are you sure, I was under the impression that in IEEE standard floating point it was infinity and whether it is positive or negative infinity depends on whether the zero is postive or negative (indeed IIRC dividing like this is the reccomended way to test if a zero is positive or negative since positive zero and negative zero compare equal).
Computers are discrete systems, and either need to use approximations for using calculus or need to use something else. IMO the point of knowing calculus for programming is not to get the computer to perform calculus (though that may be usefull sometimes) but to work things out before you write the code.
For example some curve fitting algorithms need to know the derivitive of the function you are fitting with respect to each of it's parameters. Sure you could approximate those derivitives by numerical methods but your code is likely to be far more efficiant if you can work them out analytically in advance (particularlly if some of them turn out to be constants)
Calculus is also usefull for approximations either because you need more speed or because you are porting to a system (for example I had to approximate the exponential when porting said curve fitting code from C to verilog)
Of course if your applications aren't doing any number crunching or someone else has already worked everything out for you it's not so important.
Afaict there are two types of user for theese new "netbooks" and the closely related EEEBox
One is people who want a simple system for web email and maybe some light word processing. Theese people are happy with the cheaper linux models.
Then there are people who want an ultraportable laptop (or in the EEEBox's case a miniture desktop) but couldn't previously afford one. Yeah the specs on an EEE aren't great but it's perfectly capable of running older games like starcraft (blizzard generously modified this to be able to run without CD recently), zoo tycoon (with a CD crack), the windows ports of duke nukem 3D, . It is also perfectly capable of running older versions of MS office (My experiance was with 2000 but 2003 would probablly also be find and even 2007 would probablly run though maybe not at acceptable speed). The screen resoloution was an issue with the 700 series but the 900 mostly fixes that (occasionally one has to go into scaled down or side scroll mode but only for a handfull of applications).
Those in the latter group will want the OS they are familiar with. For most people that means windows XP:/
Also given that few ordinary consumer PCs are still sold with XP and even fewer will be after MS stops letting OEMs ship systems pre-downgraded I suspect theese "netbooks" and the EEEBox are picking up a load of people who's primary criteria for thier new machine is "NOT VISTA".
if you're going to build a Linux laptop, why not pick hardware known to be compatible? Because for ASUS putting together a set of linux drivers that will make the hardware work is a minor cost. When you are buying in the quantities ASUS is hardware manufacturers *WILL* cooperate.
Afaict thier primary aim was to make a PC that was both small and cheap. I dunno why they shipped with linux first (they provided instructions and drivers for setting up windows from the start) but my suspiscion is they did it primerally to get better terms out of microsoft (which they have succeeded in doing)
Slashdot Mirror
and i think server 2K3 and XP proffesional x64 edition are NT 5.2 though i'm not positive on that.
IIRC the actual version numbers reported by the API are (i've left out the obvious ones)
95: 4.0
98: 4.1
ME: 4.2
2K: NT 5.0 (NT is a seperate flag reported along with the version number
XP: NT 5.1
vista: NT 6
Since spamming is all about getting the same text into lots of forums, simply have an anti-spam bot read random other forums. If they have messages which contain the same or nearly same text than what was posted on your forum, they're either spam or copypasta troll, and should be removed.
Thing is there are a huge number of forums and a huge number of spambots so unless you have a very big sample or you know sites that are getting hit by the same spambots as you you are likely to miss a lot this way. You can't really use google to check because it doesn't update fast enough.
Alternatively or additionally, you could traverse any links and check if they contain any telltale signs of spam.
Indeed you could and the spammer could take measures to hide those telltale signs.
I still think if you are a small site your best bet is to roll your own anti-spam measures and operate on the principle that it won't be worth it for the spammers to figure them out.
The best security I've seen on a sign-up form was "if you're a human, please leave this field blank". Bots tend to fill in all fields, so this already goes a long way towards filtering them out.
It works for your small site sure, the problem is once you (or sites running your software if you write software for others to use) grow beyond a certain size the spammers will start targetting you/your software. Once they do that you need a much stronger system to stand a chance of keeping them at bay.
Something like KittenAuth [thepcspy.com] has been recommended, and still seems to be the best answer in my opinion.
They don't give many details but presumablly there are a limited number of pictures in thier database each tagged with what animal it is. How hard would it be for a spammer to get hold of a significant proportion of the database?
To detect humans, wouldn't it be easier and less costly, and perhaps even more effective, to hold a large database of questions that are readable and solvable only by humans?
I guess the question becomes how large is large. If you reuse tests too much then the spammers will just build their own database of soloutions.
Using a database of non computer created challenges is a good idea but there needs to be a system for keeping that database topped up. Recapatcha for example picks out words from old books that thier OCR software fails on and uses them to test your users.
Normally they give the user two words, one for which they know the answer already (that is at least two people have given the same answer for it) and one they don't. but if they see failures from an IP they switch to giving the user two words that they know the answer for already.
In general if you are relatively small I think a custom soloution is one of your best defenses against spammers. At the end of the day spamming is about getting as many people as possible to see your spam as possible for as little effort as possible. Investigating a contact form just to spam one small forum or a contact form for a few people at a company just isn't worth it.
* YUV->RGB color conversion (there isn't even a single right answer to this, it's subjective)
There is a right answer if you know what definitions or RGB and YUV you are working to. Of course if you are displaying on a random monitor than it will never be right for everyone because of monitor variation so you probablly want to let the videophiles tweak this.
Audio language selection (how do you write code to tell if the guy is speaking in french as opposed to spanish?)
You make sure the different "language" tracks in your test data are different enough that an automated system can recognise them.
does anyone have any idea how big PS3 games actually are? Sure the console may support 50 gigs on one disc but I would have thought with multiplatform games at least they would be about the same size on all the platforms they supported.
media companies don't want thier products (whether books, movies or games) to appear in a trickle. A trickle would mean retailers who manage to get stock are likely to charge high prices (which leaves customers with the impression the game is very expensive) and means spoilers and pirate copies are likely to be widely spread before most people get thier copy of the product.
So they decide on a release pricepoint, estimate the demand at that pricepoint and then organise a release so that the release day lines up with the marketing campaign and with the availibility of sufficiant product that supplies won't run out. Of course this all screws up if someone leaks the game first.
The problem is the kernel team aren't cooperating :(. They won't keep the -486 kernel slim enough and they won't sanction the creation of a seperate kernel flavor just for the floppy installer to use it.
With etch the kernel had grown to the point that they had to kick everything that wasn't absoloutly essential (including USB floppy support) from the boot floppy.
With lenny a couple of bad things happened. Firstly the UPX recompressor stopped working for current kernels. And then just as there was talk of fixing that the lenny kernel grew to the point that it just wasn't feasible to put it on a floppy at all.
It doesn't say where it downloads the changelog from though............
The changelog presented in aptitude should be up to date as well. It's pulled from the package.
That seems unlikely to me since it would require the package manager to download the whole package to show the changelog.
The shortened release cycle seems to be working pretty well. Afaict thier current strategy is to aim for 18 months and be happy with 24. They achieved that with etch and it seems likely they will achieve it with lenny.
Afaict cards will only work in OS-X if they are recognised by the EFI firmware and they will only be recognised by apples EFI firmware if they have the right firmware on the card. In other words you can't just go look up a hardware compatibility list and buy a card with the right chipset you have to buy a special apple card with the special firmware on it.
Sometimes it is possible to reflash cards if you can find a firmware dump from a mac card with the exact same hardware as the PC card you are dealing with but doing so almost certainly voids any warranty on the card and probbablly has a nontrivial risk of bricking the card. IIRC it also requires a PC to do the reflashing.
I always thought that it was simply the sound pressure level that caused hearing loss
I think the point is that with outside noise sealed off you can enjoy the music at much lower volume because you aren't trying to drown out the ambiant noise.
Unfortunately it means those who DO want thier music to replace ambient noise end up listening to it far too loud. Either because they are ignorant of the existance of ones that do block outside noise or because they find them unreasonablly expensive.
It seems what is needed is
1: clear education that there are two types of earbuds, ones designed to block outside noise and ones designed not to and that drowning out outside noise by cranking up the volume is a very bad idea.
2: earbuds that block outside noise widely availible at a reasonable price.
but I also feel that I have to point out that they're close to being as upgradeable as the PC in each product subcategory.
True, the thing is apple doesn't make anything directly comparable to normal PCs so those who want to run OS-X "legally" are pushed into buying a weird system.
The mac mini is far less upgradable than a normal low end PC. The imac is far less upgradeable than a normal midrange PC.
As for the mac pro that is highly upgradable but it is very expensive initaly and ram and graphics upgrades require special parts (the ram seems comparable in price to server ram, the graphics cards are far more expensive than equivilent PC cards and there is much less choice).
What course are you doing?
Afaict the convetional way to do hexadecimal on a 7 segment display is to use capital A,lower case b, capital C, lowercase d , capital E and capital F . Theese along with the numbers can be represeted in a reasonale fasion on a 7 segment display without ambiguity.
Are you sure, I was under the impression that in IEEE standard floating point it was infinity and whether it is positive or negative infinity depends on whether the zero is postive or negative (indeed IIRC dividing like this is the reccomended way to test if a zero is positive or negative since positive zero and negative zero compare equal).
Computers are discrete systems, and either need to use approximations for using calculus or need to use something else.
IMO the point of knowing calculus for programming is not to get the computer to perform calculus (though that may be usefull sometimes) but to work things out before you write the code.
For example some curve fitting algorithms need to know the derivitive of the function you are fitting with respect to each of it's parameters. Sure you could approximate those derivitives by numerical methods but your code is likely to be far more efficiant if you can work them out analytically in advance (particularlly if some of them turn out to be constants)
Calculus is also usefull for approximations either because you need more speed or because you are porting to a system (for example I had to approximate the exponential when porting said curve fitting code from C to verilog)
Of course if your applications aren't doing any number crunching or someone else has already worked everything out for you it's not so important.
The EEEBox looks pretty similar to the EEEPC 1000H to me. The differences seem to be
* more flexible memory card reader
* more ports
* no battery
* no screen
Storage, processor, chipset and ram all seem to be the same.
Afaict there are two types of user for theese new "netbooks" and the closely related EEEBox
One is people who want a simple system for web email and maybe some light word processing. Theese people are happy with the cheaper linux models.
Then there are people who want an ultraportable laptop (or in the EEEBox's case a miniture desktop) but couldn't previously afford one. Yeah the specs on an EEE aren't great but it's perfectly capable of running older games like starcraft (blizzard generously modified this to be able to run without CD recently), zoo tycoon (with a CD crack), the windows ports of duke nukem 3D, . It is also perfectly capable of running older versions of MS office (My experiance was with 2000 but 2003 would probablly also be find and even 2007 would probablly run though maybe not at acceptable speed). The screen resoloution was an issue with the 700 series but the 900 mostly fixes that (occasionally one has to go into scaled down or side scroll mode but only for a handfull of applications).
Those in the latter group will want the OS they are familiar with. For most people that means windows XP :/
Also given that few ordinary consumer PCs are still sold with XP and even fewer will be after MS stops letting OEMs ship systems pre-downgraded I suspect theese "netbooks" and the EEEBox are picking up a load of people who's primary criteria for thier new machine is "NOT VISTA".
if you're going to build a Linux laptop, why not pick hardware known to be compatible?
Because for ASUS putting together a set of linux drivers that will make the hardware work is a minor cost. When you are buying in the quantities ASUS is hardware manufacturers *WILL* cooperate.
Afaict thier primary aim was to make a PC that was both small and cheap. I dunno why they shipped with linux first (they provided instructions and drivers for setting up windows from the start) but my suspiscion is they did it primerally to get better terms out of microsoft (which they have succeeded in doing)