Re:not the answer - you got that right!
on
Replacing SMTP?
·
· Score: 1
That said, Bayesian should be done on the server before the real client downloads it. Sure, it still gets to the server
There are a couple problems with that, 1) It takes a lot of processing power, which is mostly just going to waste on the client side. Performing Bayesian filtering for hundreds of people would require some pretty crazy hardware, I think. and 2) Having an individual Bayesian 'profile' on your computer is great, and you can filter things based on what you consider spam.
Re:not the answer - you got that right!
on
Replacing SMTP?
·
· Score: 1
Not to good if you have users who are very promiscuous in their choice of sending IP: cybercafe's, numerous dial-up ISPs, home DSLs and so on. The proposed workaround is to use subdomains with different server lists, falling back on an unrestricted list if required, but such use of subdomains in email addresses is not always desirable.
Well if they want to send mail from their customer@yourISP.com then obviously they should use your mail servers to do it. You can setup a webmail form for people using Cybercafes, and they should be able to setup their mail clients to use your server at home. If they want to send mail from their own domain then they can setup their own DNS server with RMX feilds.
I was getting all pissed about spam untill I got the baysian filter for Outlook. God that rocks. I still get spam, but now I can watch it all being deleted right before my eyes:P
The few things that do get through either have obvious subjects that I can just delete, or obviously designed to get around the filters.
I've been getting a few (like one or two) 'image' spams which contain a bunch of random text and an 'image' which is an advertisement. All we need to do is code Bayesian filters to comprehend the text embedded in images, and we should be on our way.
Seriously people, don't propose some draconian anti-Spam measures until you have a Bayesian filter installed. For most people the false-positive rate is better then if you delete Spam manually. That's right, you are more likely to falsely identify Spam then the filter.
I also think the whole sender-verification thing is good, where a person needs to reply to a letter to 'prove' they are a real person. (there's another term that's in vogue these days for that but I can't remember it)
I think Bayesian filters, sender-verification for high-scored msg, and reverse MX together will probably solve %99.99 of the spam problem. This is a lot better then charging for email (which is idiotic) and blocking off huge parts of the net in a sort of 'let god sort 'em out' attempt to keep people from spamming.(note that somethingawful just got there entire class B blocked by spews.org)
The trick isn't to try to stop people from spamming, but rather to prevent yourself from getting it.
Re:Check out Internet Mail 2000
on
Replacing SMTP?
·
· Score: 1
The "pay for email" approach would only work if it was possible to whitelist addresses who would then not have to pay. The mailing list problem then would not exist -- you simply require that anyone who signs up whitelists the mailing list address.
So why not just stick with the whitelist, and not worry about the 'pay for mail' part?
Sheesh, would it kill people to put a nice high-res color display on these things? A powerfull calculator with a high-res display would kick ass, and not just for gaming, it would be great for development too.
I have to say though, if it uses the old HP style syntax it's going to suck. TI calcs are a lot more intuitive.
It wasn't the MPAA, it was the networks IIRC. MPAA sued over VCRs in the 80s though. Heh. Jack Valenti said they were 'the boston strangler'. God what an idiot.
It's just incredibly asinine that companies which broadcast their content through to open air or pump it into our homes can even think to sue people who make PVRs that aren't to their liking. Go after someone who posts copied shows on their website or Kazza sure, but suing a company because their PVR has an eithernet port.
If *I* make it, I get to control it until my death. Period.
What you want is pretty much irrelevant. If copyrights were shortened there's nothing you can really do about it.
Personaly, I'd like to see a system where we have two kinds a of copyright, one that expires quickly, which prevents people from even sharing files, and another that lasts for decades, which prevents people from profiting of other people's work.
It's no big surprise to discover that most people who violate copyright laws aren't concerned about violating copyright laws. I'm more surprised by the other third - do they represent the traders of legal files (new Linux distros, freely tradable music etc.) or the truly stupid?
The only thing that's a crime with copyright law is creating anti-copy-protection devices. In the real world it would be like marketing a device that could open any lock.
Which would actually be totally legal (unless they passed a new law)
Seriously though, we live in a democracy, congress gets to set the limits it wants. If life + 90 years is 'reasonable' then so is a day. Copyright protection is a matter of practicality, not morality. If it's impractical in it's present state, then we should change it.
This kid is getting supporters in all the right places, and you'll notice that many of his detractors tend to be the type of people who were still arguing the Earth was flat back in the 1800's.
Can you give a specific example of someone, anyone, who believed the earth was flat in the 1800s, or are you just an idiot?
... if time is continuous and that there isn't a thing like single points in time (which effectively explain some things), why do you, human, believe that we could measure single points ? Could it be that computers functions even more identically to our brain that we suspected ?
Suspected by whom? Most people who know anything about CS theory consider all computational models (RAM machine (most computers), Neural network (brain), etc) to be basically equivalent.
You have the tortoise, you have Achilles, and you also have a rock which is between Achilles and the tortoise. In order to move 'infinitely close' to the tortus, Achilles needs to pass the rock, which is (say) 3 meters behind the tortoise. But doesn't the paradox also apply to Achilles and the rock? Doesn't it apply to all pairs of objects?
Of the paradox had any validity at all, then no motion whatsoever could ever happen. Obviously that's not the case.
At some point, Achilles is on the other side of the tortoise, whether or not he ever has the same position is irrelevant.
They guy's a collage dropout, and basically spewed a bunch of philosophical rubbish. He's either dense, or just exposing density everywhere else (i.e. trolling)
Not even the people on FARK.com bought into this crap (where it was posted a week ago). The paper is a bunch of crap and doesn't tell us anything either we don't already know, or is in any way usefull.
Obviously you're going to catch people doing things as obvious as that, but the Eiffel tower is still blow'd up.
But do you really think all insider traders are caught? Suppose Martha Stewart had gotten her info from someone at the FDA, who could have told her the tests were 'looking bad'. She could have sold early and been sneakier (for example, she could have deleted her own phone records rather then asking her secretary to do it). For only $40k, she probably didn't think it was that big of a deal.
Catching insider traders is hardly a solved problem.
Besides, if everyone who guesses right gets a visit from the FBI, then no one is going to play the damn game.
The deal is, AT&T had a 'standard' contract for licensing Unix, which IBM, Sequent, and other companies signed. Sun Microsystems didn't, they purchased their license outright, and thus don't have a contract (or something, not really important but sun can do whatever it wants.)
The standard contract said that AT&T has rights on any changes you make to your Unix code, IBM got a special amendment to the contract stating that they owned any changes they made to AIX. but Sequent did not, and Sequent, not IBM developed the tech that SCO is bitching about (NUMA, RCU, ETC). IBM, SCO and Sequent were all working together on project Monterey, which was supposed to be an 'industry standard' Unix, but IBM lost interest in Monterey, purchased Sequent and left SCO out to dry (now, this was the 'old' SCO, not Canopy/Caldera SCO).
Anyway, SCO thinks that it owns the contracts because it owns the original Unix IP (or at least the copyrights anyway, none of the patents)
In fact, the person who wrote the RCU code for Dynix (or whatever Sequent's *nix was called) for Sequent is the sameperson who wrote the RCU code in Linux for IBM.
but, and here's the thing, Sequent developed it's tech independent of Unix. RCU and that stuff would work on any modern OS's kernel. They published their papers and filed their patents before they implemented it in their own version of Unix.
So the issues are:
Did the contract rights really transfer to SCO?
Does the "naked" contract Sequent signed mean all their tech is owned by SCO now? Does it mean that the "naked" contract mean that any technology that you develop, and later put into a Unix clone automatically become the property of SCO?
Does IBM's special contract apply to stuff they acquire from other Unix vendors?
The answers to these questions aren't totally obvious, although to me they should mostly go to IBM's favor here. On the other hand, the claim that SCO owns all of Unix is ridiculous. The proper thing to do is tell everyone what files are infringing and then have them rewritten by the community. SCO seems to claim now that anyone who touches any Licensed Unix code, or any code by anyone who has touched any Licensed Unix code has sold their soul to them, which is just idiotic.
Slashdot Mirror
That said, Bayesian should be done on the server before the real client downloads it. Sure, it still gets to the server
There are a couple problems with that, 1) It takes a lot of processing power, which is mostly just going to waste on the client side. Performing Bayesian filtering for hundreds of people would require some pretty crazy hardware, I think. and 2) Having an individual Bayesian 'profile' on your computer is great, and you can filter things based on what you consider spam.
Not to good if you have users who are very promiscuous in their choice of sending IP: cybercafe's, numerous dial-up ISPs, home DSLs and so on. The proposed workaround is to use subdomains with different server lists, falling back on an unrestricted list if required, but such use of subdomains in email addresses is not always desirable.
Well if they want to send mail from their customer@yourISP.com then obviously they should use your mail servers to do it. You can setup a webmail form for people using Cybercafes, and they should be able to setup their mail clients to use your server at home. If they want to send mail from their own domain then they can setup their own DNS server with RMX feilds.
I was getting all pissed about spam untill I got the baysian filter for Outlook. God that rocks. I still get spam, but now I can watch it all being deleted right before my eyes :P
The few things that do get through either have obvious subjects that I can just delete, or obviously designed to get around the filters.
I've been getting a few (like one or two) 'image' spams which contain a bunch of random text and an 'image' which is an advertisement. All we need to do is code Bayesian filters to comprehend the text embedded in images, and we should be on our way.
Seriously people, don't propose some draconian anti-Spam measures until you have a Bayesian filter installed. For most people the false-positive rate is better then if you delete Spam manually. That's right, you are more likely to falsely identify Spam then the filter.
I also think the whole sender-verification thing is good, where a person needs to reply to a letter to 'prove' they are a real person. (there's another term that's in vogue these days for that but I can't remember it)
I think Bayesian filters, sender-verification for high-scored msg, and reverse MX together will probably solve %99.99 of the spam problem. This is a lot better then charging for email (which is idiotic) and blocking off huge parts of the net in a sort of 'let god sort 'em out' attempt to keep people from spamming.(note that somethingawful just got there entire class B blocked by spews.org)
The trick isn't to try to stop people from spamming, but rather to prevent yourself from getting it.
The "pay for email" approach would only work if it was possible to whitelist addresses who would then not have to pay. The mailing list problem then would not exist -- you simply require that anyone who signs up whitelists the mailing list address.
So why not just stick with the whitelist, and not worry about the 'pay for mail' part?
I mean seriously what's the deal? It seems like they're mostly just speculating on the outcome of this trial thing..
They have enough pure cash on hand to buy SCO outright.
Sheesh, would it kill people to put a nice high-res color display on these things? A powerfull calculator with a high-res display would kick ass, and not just for gaming, it would be great for development too.
I have to say though, if it uses the old HP style syntax it's going to suck. TI calcs are a lot more intuitive.
It wasn't the MPAA, it was the networks IIRC. MPAA sued over VCRs in the 80s though. Heh. Jack Valenti said they were 'the boston strangler'. God what an idiot.
It's just incredibly asinine that companies which broadcast their content through to open air or pump it into our homes can even think to sue people who make PVRs that aren't to their liking. Go after someone who posts copied shows on their website or Kazza sure, but suing a company because their PVR has an eithernet port.
It's just assinine.
If *I* make it, I get to control it until my death. Period.
What you want is pretty much irrelevant. If copyrights were shortened there's nothing you can really do about it.
Personaly, I'd like to see a system where we have two kinds a of copyright, one that expires quickly, which prevents people from even sharing files, and another that lasts for decades, which prevents people from profiting of other people's work.
It's no big surprise to discover that most people who violate copyright laws aren't concerned about violating copyright laws. I'm more surprised by the other third - do they represent the traders of legal files (new Linux distros, freely tradable music etc.) or the truly stupid?
I don't know, Are only the stupid immoral?
The only thing that's a crime with copyright law is creating anti-copy-protection devices. In the real world it would be like marketing a device that could open any lock.
Which would actually be totally legal (unless they passed a new law)
Bring on the revolution!
Seriously though, we live in a democracy, congress gets to set the limits it wants. If life + 90 years is 'reasonable' then so is a day. Copyright protection is a matter of practicality, not morality. If it's impractical in it's present state, then we should change it.
Note to RIAA: we will dance on your grave.
Can't you just do a proof by contradiction?
This kid is getting supporters in all the right places, and you'll notice that many of his detractors tend to be the type of people who were still arguing the Earth was flat back in the 1800's.
Can you give a specific example of someone, anyone, who believed the earth was flat in the 1800s, or are you just an idiot?
... if time is continuous and that there isn't a thing like single points in time (which effectively explain some things), why do you, human, believe that we could measure single points ? Could it be that computers functions even more identically to our brain that we suspected ?
Suspected by whom? Most people who know anything about CS theory consider all computational models (RAM machine (most computers), Neural network (brain), etc) to be basically equivalent.
You have the tortoise, you have Achilles, and you also have a rock which is between Achilles and the tortoise. In order to move 'infinitely close' to the tortus, Achilles needs to pass the rock, which is (say) 3 meters behind the tortoise. But doesn't the paradox also apply to Achilles and the rock? Doesn't it apply to all pairs of objects?
Of the paradox had any validity at all, then no motion whatsoever could ever happen. Obviously that's not the case.
At some point, Achilles is on the other side of the tortoise, whether or not he ever has the same position is irrelevant.
They guy's a collage dropout, and basically spewed a bunch of philosophical rubbish. He's either dense, or just exposing density everywhere else (i.e. trolling)
Not even the people on FARK.com bought into this crap (where it was posted a week ago). The paper is a bunch of crap and doesn't tell us anything either we don't already know, or is in any way usefull.
You mean they're offshoring game development now too!? When will the horrors stop!!
Oh, wait.
Obviously you're going to catch people doing things as obvious as that, but the Eiffel tower is still blow'd up.
But do you really think all insider traders are caught? Suppose Martha Stewart had gotten her info from someone at the FDA, who could have told her the tests were 'looking bad'. She could have sold early and been sneakier (for example, she could have deleted her own phone records rather then asking her secretary to do it). For only $40k, she probably didn't think it was that big of a deal.
Catching insider traders is hardly a solved problem.
Besides, if everyone who guesses right gets a visit from the FBI, then no one is going to play the damn game.
Hacked is the proper term, unless they were breaking a copy-protection scheme.
The deal is, AT&T had a 'standard' contract for licensing Unix, which IBM, Sequent, and other companies signed. Sun Microsystems didn't, they purchased their license outright, and thus don't have a contract (or something, not really important but sun can do whatever it wants.)
The standard contract said that AT&T has rights on any changes you make to your Unix code, IBM got a special amendment to the contract stating that they owned any changes they made to AIX. but Sequent did not, and Sequent, not IBM developed the tech that SCO is bitching about (NUMA, RCU, ETC). IBM, SCO and Sequent were all working together on project Monterey, which was supposed to be an 'industry standard' Unix, but IBM lost interest in Monterey, purchased Sequent and left SCO out to dry (now, this was the 'old' SCO, not Canopy/Caldera SCO).
Anyway, SCO thinks that it owns the contracts because it owns the original Unix IP (or at least the copyrights anyway, none of the patents)
In fact, the person who wrote the RCU code for Dynix (or whatever Sequent's *nix was called) for Sequent is the sameperson who wrote the RCU code in Linux for IBM.
but, and here's the thing, Sequent developed it's tech independent of Unix. RCU and that stuff would work on any modern OS's kernel. They published their papers and filed their patents before they implemented it in their own version of Unix.
So the issues are:
Did the contract rights really transfer to SCO?
Does the "naked" contract Sequent signed mean all their tech is owned by SCO now? Does it mean that the "naked" contract mean that any technology that you develop, and later put into a Unix clone automatically become the property of SCO?
Does IBM's special contract apply to stuff they acquire from other Unix vendors?
The answers to these questions aren't totally obvious, although to me they should mostly go to IBM's favor here. On the other hand, the claim that SCO owns all of Unix is ridiculous. The proper thing to do is tell everyone what files are infringing and then have them rewritten by the community. SCO seems to claim now that anyone who touches any Licensed Unix code, or any code by anyone who has touched any Licensed Unix code has sold their soul to them, which is just idiotic.
There were no reports of fraud or irregularities related to the new machines.
But how would you ever know? And anyway, if an election offical was corrupt, they could give out fake cards.
1) bet on assassination of president
2) assassinate president
3) profit!