However, the stats in The Fine Article indicate that a statistically significant proportion of web users are in fact avoiding ads.
I'd have been interested to see the raw figures, for example relating page impressions to ad impressions per browser family (i.e. what ratio of FF-ers actually AdBlock). All we know is that they click less, but is it simply because they click less, or because they block more. It's almost certainly both, but in what proportions?
I know what you're trying to say, but the thing is that if filtering doesn't take place on the client then it can either be circumvented, or cannot be customised sufficiently fine-grainedly enough.
But you're right, it's better to filter uncontrovertable indesirable mails as early as possible though, so blacklisting/RBHL/virus-scanning should be done prior to the client ever receiving the mail.
Why do I remember seeing exactly this technology (as in non-contact vinyl reading) demonstrated on BBC's/Tomorrow's World/ back in the 1980s? We saw it actually demonstrated live, it wasn't just a theoretical idea. IIRC they played a Cliff Richard album, and IIRC they also, with great humour, scratched the fuck out of it for a second test, which the reader passed admirably.
That was nigh on 20 years ago. It appears that the wheel has been reinvented...
What a complete waste of space he/she was. I beleive there were about a dozen TLA mickey-take nicks at one time. Unlike the rest of them, I still use this sappy nick! I don't mind too much, as it dates me and shows, to those that recognise it, my general attitude to waste-of-space trolls.
"That's like saying that because I would like to fly, I'll just imagine/think myself into having wings... a few generations down the road..."
Not at all.
You can plan evolution. You put yourself, and thousands of like-minded individuals, in an environment which provides a concrete advantage to having a particular trait.
If the central asian tribes of 10000 years ago had _wanted_ to be shorter and stumpier, then they could have chosen to migrate further north, and then procreate sufficiently such that the ill-suited ones dying would not be terminal for the tribe.
And who the buggery bollocks mentioned "a few generations"? Resistance to disease can become a prediminant trait in 20 generations, but that's a population evolution, a change in the balance of the genetic make-up of the species as a whole, and involves little or no genetic evolution. Genetic evolution on the whole takes many more generations just to become significant.
If you can contrive an environment where having wing-like upper limbs becomes a concrete survival/breeding advantage for tens of thousands of generations, then yes, you will see evolution in that direction occur.
Either that or wait a few years for the US to plunge the world into nuclear war, and watch your genes mutate like billy-o.
THL
Re:Reminds me of EYES
on
Fish with Limbs
·
· Score: 5, Interesting
"...multiple seperate occasions..."
Not just that, but _differently_ in each of the major eye types. There's no "animal eye", there are at least 3. (I forget exactly how they're categorised, but they're roughly vertibrate, cephalopod, and creepy-crawly.)
In particular, our retinas are 'back to front'. It's an flawed design, and that's why we have blind spots -- it's where the nerves leave the inside of the eyeball. If our eyes came about through _design_ then it was _crap_ design.
God-freaks can take the soft cheese out of their ears now.
"Yes, just like most of the bigger Linux distro makers out there. Commercial Linux distros are bad things, right?"
Like Lindows?
As a Debian user and contributor, I can say yes to that one quite easily. Also, as someone who tried to find whether a cerain package that I contribute to for Debian, whose upstream no longer exists, is also carried by Red Hat and SuSE et al. I can say with all honesty that those large commercial distributions are an utter abomination. It was nigh on impossible to find out what was in their distributions, despite the fact that the only product they sell is those very packages.
Slammer will not compress, it's too high entropy. (Although it can be tweaked to make it slightly smaller and the PRNG could be simplified too at no loss of workingness (as it doesn't work properly anyway!)).
However, don't confuse a UDP packet with an Ethernet MTU. Most of the time it's not being transported around framed as an ethernet packet, so that's a red herring. Slammer works as a UDP packet, and therefore was limitted by UDP's 64KB, not Ethernet's varying MTU size.
But UDP packets can be massive (relative to 376 bytes anyway). Even if you restrict it to typical ethernet MTU size (1500 bytes) so that you can guarantee no fragmentation while on ether, that's still room for 376 useful bytes and 1.1kB of.PNG thumbnail of the author's girlfriend or whatever. And when you consider most of the time is not spent going over ether, it's (1.5k, that is) a silly restriction anyway.
Don't worry - the reason you don't follow what he's saying is because he's karma-whoring and trying to spout something theat looks vaguely mathematical. e.g.: """ Ofcourse, if there is indeed a usufull pattern, it may help to find the primes---that are required for factorization---faster, """
Which is patent nonsense. Finding non-trivial primes is in no way required for factorisation.
It's not required for trial-division (they're all trivial), pollard rho, pollard P-1 (all trivial), williams P+1 (all trivial), ECM (all trivial), shanks' squfof, the Quadratic sieve (all trivial), or the Number Field Sieve (all trivial). Have I missed any? Oh yes I have - it's also not required for Euler's algoritm, Legendre's algoritm, Fermat's algorithm, or Lehman's algorithm. Nor for the Continued Fraction algorithm. There are some more esoteric ones (Bach's cyclotomic technique, gaussian rings, hyperelliptic curves etc) too, and they don't require it either.
All techniques apart from trial-division simply have the factors drop out using a GCD of the composite number and the _difference between two elements_ in the ring of residues. The difference in the techniques is how you generate the two elements that you are taking the difference of. You don't actually generate the prime factor that will drop out at all.
Horizon was so good the independents immediately countered it with "Equinox", IIRC. Exactly the same formula, just with 2 advert breaks in the middle. Fortunately thy used to air on different nights so that you could see both.
Slashdot Mirror
100% correct.
However, the stats in The Fine Article indicate that a statistically
significant proportion of web users are in fact avoiding ads.
I'd have been interested to see the raw figures, for example relating
page impressions to ad impressions per browser family (i.e. what ratio
of FF-ers actually AdBlock). All we know is that they click less, but
is it simply because they click less, or because they block more. It's
almost certainly both, but in what proportions?
THL.
I know what you're trying to say, but the thing is that if filtering
doesn't take place on the client then it can either be circumvented, or
cannot be customised sufficiently fine-grainedly enough.
But you're right, it's better to filter uncontrovertable indesirable
mails as early as possible though, so blacklisting/RBHL/virus-scanning
should be done prior to the client ever receiving the mail.
THL.
"come up with", eh?
/Tomorrow's World/ back in the 1980s?
Why do I remember seeing exactly this technology (as in non-contact vinyl
reading) demonstrated on BBC's
We saw it actually demonstrated live, it wasn't just a theoretical idea.
IIRC they played a Cliff Richard album, and IIRC they also, with great
humour, scratched the fuck out of it for a second test, which the reader
passed admirably.
That was nigh on 20 years ago. It appears that the wheel has been
reinvented...
THL.
Yes. Well remembered!
What a complete waste of space he/she was. I beleive there were about a dozen TLA mickey-take nicks at one time. Unlike the rest of them, I still use this sappy nick! I don't mind too much, as it dates me and shows, to those that recognise it, my general attitude to waste-of-space trolls.
TLH
bash-2.05b$ perl -pe 's/(\S+)\s*\(at\)\s*(\S+)\s*\(dot\)\s*(\S+)/$1\@$2 .$3/gi'
foo (at) bar (dot) com
foo@bar.com
All trivial regexps.
THL.
Is that guy for real?
"blue shifted light" - wrong!
"essentially X-ray vision" - wrong!
(Section "Darkly through a lens", first paragraph.)
THL.
"That's like saying that because I would like to fly, I'll just imagine/think myself into having wings... a few generations down the road ..."
Not at all.
You can plan evolution. You put yourself, and thousands of like-minded individuals, in an environment which provides a concrete advantage to having a particular trait.
If the central asian tribes of 10000 years ago had _wanted_ to be shorter and stumpier, then they could have chosen to migrate further north, and then procreate sufficiently such that the ill-suited ones dying would not be terminal for the tribe.
And who the buggery bollocks mentioned "a few generations"? Resistance to disease can become a prediminant trait in 20 generations, but that's a population evolution, a change in the balance of the genetic make-up of the species as a whole, and involves little or no genetic evolution. Genetic evolution on the whole takes many more generations just to become significant.
If you can contrive an environment where having wing-like upper limbs becomes a concrete survival/breeding advantage for tens of thousands of generations, then yes, you will see evolution in that direction occur.
Either that or wait a few years for the US to plunge the world into nuclear war, and watch your genes mutate like billy-o.
THL
"...multiple seperate occasions..."
Not just that, but _differently_ in each of the major eye types. There's no "animal eye", there are at least 3. (I forget exactly how they're categorised, but they're roughly vertibrate, cephalopod, and creepy-crawly.)
In particular, our retinas are 'back to front'. It's an flawed design, and that's why we have blind spots -- it's where the nerves leave the inside of the eyeball. If our eyes came about through _design_ then it was _crap_ design.
God-freaks can take the soft cheese out of their ears now.
THL.
"doesn't threatening lawsuits on false grounds count as extortion?"
You're probably thinking of barratry.
http://en.wikipedia.org/wiki/Barratry
Repeatedly bringing or threatenting lawsuits without merit. Exact details vary from jurisdiction to jurisdiction.
THL.
"There's really no difference between those equally idiotic statements."
Wrong.
When was the last time you saw a mail client fuck up your braces?
When was the last time you saw an editor decide to insert a different character than a brace when you wanted to insert one?
When was the last time a web-browser rendered a brace in such a way that the mouse couldn't drag/select it?
Never, never, and never.
However, all the above can and do fuck up white-space.
THL.
Prolog
Not improved - just as bad:
..."
"Like Python, Prothon uses indentation
Oh joy.
THL.
"Yes, just like most of the bigger Linux distro makers out there. Commercial Linux distros are bad things, right?"
Like Lindows?
As a Debian user and contributor, I can say yes to that one quite easily. Also, as someone who tried to find whether a cerain package that I contribute to for Debian, whose upstream no longer exists, is also carried by Red Hat and SuSE et al. I can say with all honesty that those large commercial distributions are an utter abomination. It was nigh on impossible to find out what was in their distributions, despite the fact that the only product they sell is those very packages.
THL.
Slammer will not compress, it's too high entropy. (Although it can be tweaked to make it slightly smaller and the PRNG could be simplified too at no loss of workingness (as it doesn't work properly anyway!)).
However, don't confuse a UDP packet with an Ethernet MTU. Most of the time it's not being transported around framed as an ethernet packet, so that's a red herring. Slammer works as a UDP packet, and therefore was limitted by UDP's 64KB, not Ethernet's varying MTU size.
THL
Can't you replace
d4:b8 01 01 01 01 mov $0x1010101,%eax
d9:31 c9 xor %ecx,%ecx
db:b1 18 mov $0x18,%cl
dd:50 push %eax
de:e2 fd loop 0xdd
e0:
(Oooh look - objdump's got a bug, it says:
de:e2 fd loop 0xdd
e1:35 01 01 01 05 xor $0x5010101,%eax
e5:50 push %eax
It seems to think that the 1st instruction is 3 bytes, and the 2nd instruction is 4 bytes!
)
With something like
d4:31 c9 xor %ecx,%ecx
d6:b1 18 mov $0x18,%cl
d8:68 01 01 01 01 push $0x1010101
dd:e2 f9 loop 0xd8
df:
to save 1 byte?
OK, OK, it's pretty tight code.
THL
But you don't deny that MS took money off lots of people to sell them the _broken_ version in the first place, do you?
They are _not_ without blame, they wrote the bugs and sold them with a "if this software is crap, tough shit" EULA.
The logical conclusion, if you agree that they aren't without blame, is:
_yes_ they should be blamed.
THL
But UDP packets can be massive (relative to 376 bytes anyway). Even if you restrict it to typical ethernet MTU size (1500 bytes) so that you can guarantee no fragmentation while on ether, that's still room for 376 useful bytes and 1.1kB of .PNG thumbnail of the author's girlfriend or whatever.
And when you consider most of the time is not spent going over ether, it's (1.5k, that is) a silly restriction anyway.
THL
Fatuous argument. Nuke the server. No machine can survive a direct nuking therefore no software can be secure.
You've got to specify your threat modal before you judge security. You can't just bring in arbitrary threats post-facto.
Phil
Until someone configures their linux system to recognise .exe's in mail, and automatically launch Wine for them.
Don't laugh, it _will_ happen.
THL
"""
This is what Palladium is all about. Executable code is signed, and it can only run if you choose to trust the publisher.
"""
How do you know if you trust them or not?
THL
Don't worry - the reason you don't follow what he's saying is because he's
karma-whoring and trying to spout something theat looks vaguely mathematical.
e.g.:
"""
Ofcourse, if there is indeed a usufull pattern, it may help to
find the primes---that are required for factorization---faster,
"""
Which is patent nonsense. Finding non-trivial primes is in no way required
for factorisation.
It's not required for trial-division (they're all trivial), pollard rho,
pollard P-1 (all trivial), williams P+1 (all trivial), ECM (all trivial),
shanks' squfof, the Quadratic sieve (all trivial), or the Number Field Sieve
(all trivial). Have I missed any? Oh yes I have - it's also not required for
Euler's algoritm, Legendre's algoritm, Fermat's algorithm, or Lehman's
algorithm. Nor for the Continued Fraction algorithm. There are some more
esoteric ones (Bach's cyclotomic technique, gaussian rings, hyperelliptic
curves etc) too, and they don't require it either.
All techniques apart from trial-division simply have the factors drop out
using a GCD of the composite number and the _difference between two elements_
in the ring of residues. The difference in the techniques is how you
generate the two elements that you are taking the difference of. You don't
actually generate the prime factor that will drop out at all.
THL.
Erk, that's not so hot. I've not seen it for over 5 years, probably. I've not even had a telly for 3 - and things like this are the reason why.
THL.
So, what MS wants us to do is open connections to their site (and sites hosted (hah, typed 'hosed' initially) on IIS), and leave them open?
Surely we can oblige, if that's how they think it's done.
THL.
Lesley Judd.
I've just cum.
THL.
Horizon was so good the independents immediately countered it with "Equinox", IIRC. Exactly the same formula, just with 2 advert breaks in the middle. Fortunately thy used to air on different nights so that you could see both.
THL.