One problem is that it's difficult to decide whether it's someone who challenged as a matter of course or because SpamAssassin or whatnot flagged it as needing extra care. Perhaps the challenge mails containing the SA results would make this known. I can't see any issues with that... if a spammer actually gets and reads the challenge mail, the challenge/response system is broken anyway.
That said, I get better results with SpamAssassin (with Bayesian filter) and a procmail filter that sends SA-tagged mail to a special folder, where mutt automatically tags for deletion (so one key, after scanning senders and subject lines, deletes the spam).
He's not talking about SpamAssassin. He's talking about the idiotic whitelisting schemes (ala TMDA), that assume that all mail not whitelisted is spam and do not deliver it. In short, they always have a "0%" false positive rate, because they silently throw away all mail that's positive (and often blacklist the sender of that mail, so there's no way to tell them that their software dumped mail).
As a matter of policy, I do not respond to whitelisting requests because the sender of the whitelisting request has already accused, with zero basis in fact, of being a spammer, and beyond that, there is zero ground for a civil relationship and thus no reason whatsoever for me to communicate with them.
In this case, you would kill all the legitimate mailing lists, newsletters and other solicited mass-mailings.
See, the fault in your system is that you don't check at all whether or not you want that particular kind of mail. So you just bash everyone on the head when they come through the door, arguing that there have been too many rude guests lately.
That's what Microsoft's proposal does, too. So considering that this accomplishes the same exact thing, with the same exact problems, doing it in an easier to implement manner is a better idea.
I already have my mailserver sleep for 60 seconds on any SMTP error.
If it's on the mail server, this is effective, but how is the server going to know that a given address should be whitelisted? And on what criteria would it whitelist (and the server determine that an address has been whitelisted)?
If it's on the client side, well, you've cut down on the number of spams the user sees. You haven't cut down on the storage and transmission costs for the ISP.
The ideal solution is simple: remove common carrier as a defense in civil spam cases (along with copyright infringement, but repeal the DMCA also).
Any network that transmits spam (or material infringing on copyright) is liable for the spam/infringement. The damaged party is then able to sue up the chain.
You receive a spam. You sue an ISP with operations in the US. You win the judgement.
At this point, the ISP passes the liability onto their customers, either by suing the customers (or peer, if we're getting international) who originated the spam, or by simply incorporating the insurance premiums to protect against spam-suits into their pricing (much like how insurance companies can immediately turn around and sue someone for causing them to pay out a claim).
If the first approach is taken, then non-spamming customers are not affected. Spammers themselves end up paying the costs.
If the second is taken, ISPs that issue pink contracts are going to end up having to charge more to the non-spammers (who are effectively subsidizing the spammers). Naturally, the non-spammers would leave as better and better deals became available, thus forcing the costs to be born more by the smaller fry in the spam industry (who would thus drop out).
Imagine having Comcast say to their customers, "You're running an open relay on your end of the cable connection. Fix it or we're bumping you to a $1,000 a month plan." Or "You're running an unpatched Windows system that's spewing spam; fix it within 24 hours or we're either permanently barring you from service or putting you on a $1,000/month plan."
The same is true of international peering. Say that C&W is routing packets from Korea to the US. They start getting sued heavily for transporting spam. They respond by either going after the Korean ISPs that are harboring spammers or by simply declaring that, since business in Korea is so fraught with liability, that they're quintupling their bandwidth rates for Korea.
What I've heard is MySQL's main point in favor (at least on the MyISAM table) is the performance, especially on a command sequence that largely consists of SELECTs, with comparatively few write ops. It is probably reasonable to suspect that at least some of that performance advantage over (as near as I can tell) every other SQL (or near-SQL) implementation comes from not implementing the extraneous stuff.
Thus, if you have an application where the queries are going to mostly be SELECTs (and the application effectively does its own constraint checking and approximation of foreign keys), MySQL with MyISAM is almost certainly the best choice, as you don't need to use the extra features that others provide, and the performance increase is useful.
Does this mean that MySQL/MyISAM is ideal for any other type of application? No. Is it even suitable for any other application? Absolutely not, and anyone who says otherwise is probably insane.
But there's a large application space where the constraints and query sets are well-suited for MySQL/MyISAM (web CMS type things being but one example of this... basically anything that just needs a datastore that's reasonably relational where an SQL-like interface is useful).
Grandparent is a DB-fascist who doesn't think that programmers/admins are able to decide for themselves whether those features are needed in their particular application and thus should be required to use them in all cases.
AFAIC, MySQL's support of a diverse range of table types with differing feature profiles is one of the greatest features MySQL has.
There is an overhead to transactions and not every application requires transactions (for instance, a web board). Thus, having the flexibility to use transactions for some tables and not for others is actually a good thing. Use the right tool for the job.
Now if only InnoDB supported all the features of MyISAM...
However making a financial-type decision based on the assumption that current tax advantages would persist indefinitely is a recipe for disaster. The tax code can change at any time and you effectively have no recourse in that case.
Consider the case of the real estate limited partnerships and the Tax Reform Act of 1986.
Nor do I feel sorry for people who lost their money by buying every share of Pets.com they could find in 1999. Or those who bought junk bonds in the 80s.
Yes, you're feeling horrible and you've lost everything. But those weren't tears in your eyes when you agreed to put your money there; they were dollar signs.
The way to become rich is not through greed; greed is counter-productive.
Bulls and bears make money, but pigs become bacon.
Mandrake, in many ways is the most like Debian of the commercial distros.
First of all is the commitment to an entirely free software distribution (with non-free but gratis software relegated to PLF, and non-free not necessarily gratis software available in a separate (aka Club/Commercial) repository). This is the main social analogue.
Technically, there is much that is similar. Mandrake uses the Debian menu system (since Mandrake is, along with Debian, a major distro that is largely wm/de agnostic). I believe the alternatives system is shared with Debian. Toss in urpmi, which is largely analogous to apt-get (with a slightly different feature set) and you might get the impression that, if Debian were to create an rpm-based distribution emphasizing ease of use, it would be very similar to Mandrake.
The other thing that I'd like to add is that this sort of cheap labor is the way to build a middle class which ultimately lays the economic groundwork for education and thus innovation.
See for instance Taiwan and South Korea, which were havens of sweatshop labor decades ago but used that as a means to ignite their economies. Meanwhile India rejected sweatshops in favor of education and other initiatives (which are advocated by most of the left in the West to this day as the means to bring prosperity to the Third World) and found themselves making minimal economic progress. Only in recent years have they embraced being a low-cost source of labor.
The simple fact is that an economy needs to convert its resources into a more liquid form in order to advance economically. Some Third World nations have oil and mineral reserves, which allow them to bypass the cheap labor step. Others, having only a pool of labor, have to convert that labor into cash and do so efficiently.
Of course, an argument can be made that it's far better to be in the latter camp than to be in the former; mineral/oil reserves tend to be expensive to exploit (whereas labor is cheap) and thus only the upper classes/governments of those nations are able to exploit them and there's not much of a trickle down effect. Thus you end up with massively unstable situations where a core group of extremely rich, generally hereditary control everything while you have a large undercurrent of underclass who derive virtually no benefit from the oil/mineral exploitation and nothing in between (cf. Saudi Arabia, Texas of not too long ago (arguably today), Venezuela, etc.). Contrast this with nations/regions not blessed with exportable natural resources that have turned to having cheap labor in their economies (cf. the United Kingdom, New England, the Netherlands, Japan, Korea, etc.).
Looking at that list, it is somewhat interesting to note that the latter group are far more advanced in every field of human endeavor than the former group. It is also interesting to note that all the latter are either part of the Anglosphere by definition (New England [aka the Pinnacle of Western Civilization, with Boston as the true Jerusalem] and the United Kingdom) or have generally tended to be "honorary members of the Anglosphere". Further contemplation on those issues would be required.
Having now read the article, I'm struck by the stupidity.
The main argument against sports games is advanced by the sporting goods manufacturing industry. Of course kids are going to play less physical games when they can do it in a video game (whether that in and of itself is a bad thing is another matter).
However, that doesn't necessarily mean that they're not going to watch games on TV. I haven't played baseball in years, and I'll still watch it if it's a good game (especially if the Red Sox are playing). I haven't played football (touch or tackle) in years, and I am a massive fan. I've never played hockey (except for a few games of street hockey years ago) and I'll watch a hockey game.
Ratings across the board are declining (thanks to more channels and electronic entertainment in general). Sports, however, is actually declining more slowly than news and entertainment programming.
For instance, American Football is a sport which is known for tactical depth and rule density.
In addition to the NFL, I'm a huge fan of NFL Europe (arguably a bigger fan of NFLE than the NFL). On online fora (web boards and mailing lists), I've encountered large numbers of people (mostly under 20) in the UK who became fans of the sport thanks to playing Madden. The game allowed them to get into the rules and understand the strategies, which is a critical part of becoming a fan of the sport.
I pity those of you in Australia who were out drinking, come home to read Slashdot (since your geeks, there's no way you're bringing a chick home), and see the same story repeated four times...
Slashdot Mirror
One problem is that it's difficult to decide whether it's someone who challenged as a matter of course or because SpamAssassin or whatnot flagged it as needing extra care. Perhaps the challenge mails containing the SA results would make this known. I can't see any issues with that... if a spammer actually gets and reads the challenge mail, the challenge/response system is broken anyway.
That said, I get better results with SpamAssassin (with Bayesian filter) and a procmail filter that sends SA-tagged mail to a special folder, where mutt automatically tags for deletion (so one key, after scanning senders and subject lines, deletes the spam).
I was referring more to those who view whitelisting schemes as the be-all and end-all of spam-fighting and who deploy TMDA for all incoming mail.
He's not talking about SpamAssassin. He's talking about the idiotic whitelisting schemes (ala TMDA), that assume that all mail not whitelisted is spam and do not deliver it. In short, they always have a "0%" false positive rate, because they silently throw away all mail that's positive (and often blacklist the sender of that mail, so there's no way to tell them that their software dumped mail).
As a matter of policy, I do not respond to whitelisting requests because the sender of the whitelisting request has already accused, with zero basis in fact, of being a spammer, and beyond that, there is zero ground for a civil relationship and thus no reason whatsoever for me to communicate with them.
That's what Microsoft's proposal does, too. So considering that this accomplishes the same exact thing, with the same exact problems, doing it in an easier to implement manner is a better idea.
I already have my mailserver sleep for 60 seconds on any SMTP error.
And again, where does the whitelisting occur?
If it's on the mail server, this is effective, but how is the server going to know that a given address should be whitelisted? And on what criteria would it whitelist (and the server determine that an address has been whitelisted)?
If it's on the client side, well, you've cut down on the number of spams the user sees. You haven't cut down on the storage and transmission costs for the ISP.
The ideal solution is simple: remove common carrier as a defense in civil spam cases (along with copyright infringement, but repeal the DMCA also).
Any network that transmits spam (or material infringing on copyright) is liable for the spam/infringement. The damaged party is then able to sue up the chain.
You receive a spam. You sue an ISP with operations in the US. You win the judgement.
At this point, the ISP passes the liability onto their customers, either by suing the customers (or peer, if we're getting international) who originated the spam, or by simply incorporating the insurance premiums to protect against spam-suits into their pricing (much like how insurance companies can immediately turn around and sue someone for causing them to pay out a claim).
If the first approach is taken, then non-spamming customers are not affected. Spammers themselves end up paying the costs.
If the second is taken, ISPs that issue pink contracts are going to end up having to charge more to the non-spammers (who are effectively subsidizing the spammers). Naturally, the non-spammers would leave as better and better deals became available, thus forcing the costs to be born more by the smaller fry in the spam industry (who would thus drop out).
Imagine having Comcast say to their customers, "You're running an open relay on your end of the cable connection. Fix it or we're bumping you to a $1,000 a month plan." Or "You're running an unpatched Windows system that's spewing spam; fix it within 24 hours or we're either permanently barring you from service or putting you on a $1,000/month plan."
The same is true of international peering. Say that C&W is routing packets from Korea to the US. They start getting sued heavily for transporting spam. They respond by either going after the Korean ISPs that are harboring spammers or by simply declaring that, since business in Korea is so fraught with liability, that they're quintupling their bandwidth rates for Korea.
What I've heard is MySQL's main point in favor (at least on the MyISAM table) is the performance, especially on a command sequence that largely consists of SELECTs, with comparatively few write ops. It is probably reasonable to suspect that at least some of that performance advantage over (as near as I can tell) every other SQL (or near-SQL) implementation comes from not implementing the extraneous stuff.
Thus, if you have an application where the queries are going to mostly be SELECTs (and the application effectively does its own constraint checking and approximation of foreign keys), MySQL with MyISAM is almost certainly the best choice, as you don't need to use the extra features that others provide, and the performance increase is useful.
Does this mean that MySQL/MyISAM is ideal for any other type of application? No. Is it even suitable for any other application? Absolutely not, and anyone who says otherwise is probably insane.
But there's a large application space where the constraints and query sets are well-suited for MySQL/MyISAM (web CMS type things being but one example of this... basically anything that just needs a datastore that's reasonably relational where an SQL-like interface is useful).
Grandparent is a DB-fascist who doesn't think that programmers/admins are able to decide for themselves whether those features are needed in their particular application and thus should be required to use them in all cases.
AFAIC, MySQL's support of a diverse range of table types with differing feature profiles is one of the greatest features MySQL has.
There is an overhead to transactions and not every application requires transactions (for instance, a web board). Thus, having the flexibility to use transactions for some tables and not for others is actually a good thing. Use the right tool for the job.
Now if only InnoDB supported all the features of MyISAM...
However making a financial-type decision based on the assumption that current tax advantages would persist indefinitely is a recipe for disaster. The tax code can change at any time and you effectively have no recourse in that case.
Consider the case of the real estate limited partnerships and the Tax Reform Act of 1986.
Exactly.
I don't feel sorry at all for the guy.
Nor do I feel sorry for people who lost their money by buying every share of Pets.com they could find in 1999. Or those who bought junk bonds in the 80s.
Yes, you're feeling horrible and you've lost everything. But those weren't tears in your eyes when you agreed to put your money there; they were dollar signs.
The way to become rich is not through greed; greed is counter-productive.
Bulls and bears make money, but pigs become bacon.
Mandrake, in many ways is the most like Debian of the commercial distros.
First of all is the commitment to an entirely free software distribution (with non-free but gratis software relegated to PLF, and non-free not necessarily gratis software available in a separate (aka Club/Commercial) repository). This is the main social analogue.
Technically, there is much that is similar. Mandrake uses the Debian menu system (since Mandrake is, along with Debian, a major distro that is largely wm/de agnostic). I believe the alternatives system is shared with Debian. Toss in urpmi, which is largely analogous to apt-get (with a slightly different feature set) and you might get the impression that, if Debian were to create an rpm-based distribution emphasizing ease of use, it would be very similar to Mandrake.
MySQL does have transactions, and has had them for quite some time. Stored procedures are due in a future version.
Damn it you beat me to it.
The other thing that I'd like to add is that this sort of cheap labor is the way to build a middle class which ultimately lays the economic groundwork for education and thus innovation.
See for instance Taiwan and South Korea, which were havens of sweatshop labor decades ago but used that as a means to ignite their economies. Meanwhile India rejected sweatshops in favor of education and other initiatives (which are advocated by most of the left in the West to this day as the means to bring prosperity to the Third World) and found themselves making minimal economic progress. Only in recent years have they embraced being a low-cost source of labor.
The simple fact is that an economy needs to convert its resources into a more liquid form in order to advance economically. Some Third World nations have oil and mineral reserves, which allow them to bypass the cheap labor step. Others, having only a pool of labor, have to convert that labor into cash and do so efficiently.
Of course, an argument can be made that it's far better to be in the latter camp than to be in the former; mineral/oil reserves tend to be expensive to exploit (whereas labor is cheap) and thus only the upper classes/governments of those nations are able to exploit them and there's not much of a trickle down effect. Thus you end up with massively unstable situations where a core group of extremely rich, generally hereditary control everything while you have a large undercurrent of underclass who derive virtually no benefit from the oil/mineral exploitation and nothing in between (cf. Saudi Arabia, Texas of not too long ago (arguably today), Venezuela, etc.). Contrast this with nations/regions not blessed with exportable natural resources that have turned to having cheap labor in their economies (cf. the United Kingdom, New England, the Netherlands, Japan, Korea, etc.).
Looking at that list, it is somewhat interesting to note that the latter group are far more advanced in every field of human endeavor than the former group. It is also interesting to note that all the latter are either part of the Anglosphere by definition (New England [aka the Pinnacle of Western Civilization, with Boston as the true Jerusalem] and the United Kingdom) or have generally tended to be "honorary members of the Anglosphere". Further contemplation on those issues would be required.
How far HuSi has come in a few short months... it's now said in the same breath as K5 and MeFi).
Having now read the article, I'm struck by the stupidity.
The main argument against sports games is advanced by the sporting goods manufacturing industry. Of course kids are going to play less physical games when they can do it in a video game (whether that in and of itself is a bad thing is another matter).
However, that doesn't necessarily mean that they're not going to watch games on TV. I haven't played baseball in years, and I'll still watch it if it's a good game (especially if the Red Sox are playing). I haven't played football (touch or tackle) in years, and I am a massive fan. I've never played hockey (except for a few games of street hockey years ago) and I'll watch a hockey game.
Ratings across the board are declining (thanks to more channels and electronic entertainment in general). Sports, however, is actually declining more slowly than news and entertainment programming.
For instance, American Football is a sport which is known for tactical depth and rule density.
In addition to the NFL, I'm a huge fan of NFL Europe (arguably a bigger fan of NFLE than the NFL). On online fora (web boards and mailing lists), I've encountered large numbers of people (mostly under 20) in the UK who became fans of the sport thanks to playing Madden. The game allowed them to get into the rules and understand the strategies, which is a critical part of becoming a fan of the sport.
...except the controller would have to revolve and have forty buttons.
It would quickly be copied in a much less tasteful manner by the Mike Portnoy edition.
Mandrake is not based on Red Hat, and hasn't been for sometime (apart from the occasional patch from RawHide being merged into a package in Cooker).
You're finally getting that... I've had that functionality with XMMS (and some sawfish keybindings) for eons.
In return for which, you get an ugly UI and a program that takes over your collection and reorganizes it according to its own whims.
iTunes, the media player for dolts on welfare.
Winamp, the media player for lazy fucks on welfare.
XMMS, the media player for smart people.
Nice to see that you're still around at one of the sites I frequent... ;o)
That would depend on whether Aunt Madge Taco could do it...
I pity those of you in Australia who were out drinking, come home to read Slashdot (since your geeks, there's no way you're bringing a chick home), and see the same story repeated four times...
Wow... an on-topic FP!