That's like saying we can predict the weather because we have had big supercomputers crunching those numbers for decades.
Ok, so the analogy isn't perfect, but then again, what is?:-)
The point is, models are only as good as the data which is fed into them. Somewhere along the line, some human decided what factors should be put into the model and how they should be weighted. How accurate are those assumptions?
Personally, I'd rather not use the traditional technique to perfect those models.:-(
Although I don't disagree that there is a lot of hype when people start discussing nukes in general, I DO think it's a little idiotic to simply consider them 'just big bombs'.
The number of atmospheric tests which have been done to date pales in comparison to the number destructive tonnage available. Furthermore, those tests have been spread out over time, rather than having been conducted all at once.
Or does anyone really believe that a nuclear confrontation in today's world would involve one plane pitching one bomb, ala WWII?
NO ONE can do more than hypothesize about the effect of multiple, multi-megaton blasts (e.g. 'nuclear winter', etc.), and let's hope no one ever can.
Unfortunately, the more people who regard nuclear explosives as 'just big bombs', the greater the odds that we'll someday obtain empirical evidence.
One thing I did not see mentioned in this thread -- if you are a single-member LLC, you enjoy the protections inherent in being a corporation, yet you still have the option to file as an S-corp.
(I recently converted an S-Corp to an LLC, so I'm fairly current on LLC facts and figures.;)
And if you believe that, you deserve what you get.
Consider: Anyone who is withholding taxes on your behalf (which is *mandantory* and has been since WWII) is required to report that information to the IRS.
Voila! They have just 'filed' on your behalf, whether you do anything or not.
Actually, studies have long shown that the best time to submit your taxes to avoid audits is right on the cusp of the deadline. Your return has a larger chance of getting 'lost in the shuffle' this way.
On the other hand, there *is* a ~2% random audit process which can still result in your return being looked at in more detail - sometimes months or years after it is filed.
As for best tax filing tips -- and despite the piss poor attempts at DRM in this year's version -- skip the accountant and use TurboTax.
The other tax programs may be on a par with TT - I started using TT with its first release and have never looked back - but TT *is* a first-rate accountant - and one that never gets sleepy, overworked or makes mistakes.
Unless you're completely clueless about taxes, and can not understand the govt instructions (which are no worse than, say, the average dot-matrix printer user's manual from 1983;), the accountant is unlikely to have enough time to know your business well enough to 'find' more deductions, other than the 'obvious' ones.
If you have a corporate accountant on-staff / on-call, then absolutely let them handle the taxes. You DO trust them, right?
Finally, a comment regarding filing in general: If you are owed a refund, you do NOT have to file by the deadline. If you consistently refuse to file, after a few years the IRS will file on your behalf, using stock #s from your employer-submitted W2s. At that time you can file 1040-X forms showing the real numbers and claim your refund. (Refunds expire after 5 years!)
I don't recommend this practice, but I have used and have known others who have used it.
Jim, I read the cease-and-desist order, as well as the rant which has replaced the former site. And, well, all I can say is, don't you think you're overreacting???
As far as I can tell, PCI-SIG objects to the use of "PCI-SIG" and their logo, both of which seem to be reasonable objections on their part. Your site is NOT actually connected with theirs, after all.
So why not simply remove those two items and keep this list up? It almost sounds to me like you were tired of doing it, and this was just a convenient excuse.
As far as the letters "PCI" go, it is naive to believe that PCI-SIG can enforce the use or non-use of those letters in any way, shape or form. There are more than sufficient legal precedents which parallel how this mark has become a household word - and even if that were not the case, a simple addition of the appropriate R/C/SM/TM symbol, possibly with a footnote disclaimer, should more than cover the issue.
So all in all I must be missing something, as I just don't see what the big deal is all about.
Laying all jokes aside, as a friend of mine was wont to say, THEY CAN DO EXACTLY THAT.
Take a database of all captured internet traffic, do a search looking for the date/time and some key strings and determining the IP address from whence those "anonymous" posts originated is _trivial_.
This is BEING DONE TODAY. Are you a big enough fish to surface through the floatsam to the point where said search would be likely to be originated? Probably not. But we all know what they say about 'security through obscurity', right?
I don't mean to sound paranoid (heck, I'm even logged in;), but the fact is, that traffic IS being captured, those searches ARE possible, and not only is this 'legal', it's even being mandated!
Except, of course, for the fact that the FBI has both the legal authority and the ability to have monitored the http traffic as you posted, and are likely backtracking the IP address even as we speak. So, good luck with that one....
I don't know - anywhere? I have a quote in my inbox right now for an equivalent dual-CPU Intel box w/ 2G for less than I can buy the LX50 even with my standard Sun discount.
And Sun upgrades are so economical, too. Witness the difference between the 'medium' and 'large' configurations -- you lose 36G of hard disk and gain 1G of memory -- for only a thousand dollars.
If you can't beat those prices, perhaps you'd allow me to start supplying some of your parts?
You're joking, right? The only reason Scott thinks he can sell it to the bosses is because those self-same geekoids have already implemented it anyway - something that those same bosses have finally started to notice. It's easy to justify Intel vs. Sun, especially in today's marketplace. Support on a 8-way Intel box is $5k (24x7 4-hour), while the equivalent Sun is:
With numbers like this, it's easy to make people pay attention. (Admittedly a V880 or something would probably be cheaper than the E4500 --- but we don't own any V hardware so I have no comparative quotes there.)
Even Sun's Linux box is no bargain; I laughed when my VAR gave me a quote.
Hello?!? Earth to afidel??? a) Who cares what Dell ships, and b) we bought our first 8-way boxes over a year ago. Blade servers are seeing increased n-way performance across multiple blades, and SGI has announced (w/ Jan availability) a 64-way Itanium box which (they say) will blow to doors off an E15k.
The world is fast catching up with the best Sun has to offer, for a HECK of a lot cheaper, and with lower service contracts to boot. (And let's not even talk about the Ultra II and its designed-for-ECC-failure modules.)
This can do nothing positive except possibly generate some short term income for Bitkeeper, while at the same time has enormous negative - and long-term - consequences for same.
Forcing open source projects which are evolved enough to bootstrap onto their own code typically causes a rapid bug-fix cycle (), which is never a bad thing. For those that aren't evolved enough, having to spend unnecessary work (or money) will simply provide motivation to increase those efforts so that said bootstrapping can happen. And closed source efforts are either already licensed or can simply ignore the new dictum (so long as those political realities do not yield internal squealers, anyway).
I, for one, would be interested to see the hard numbers in license acquisition and renewals for BK for say a year in either direction of this moment - but ultimately I think they're pretty predictable.
Perhaps BK will 'merge' with MickeySoft, and disappear like so many before it....
Perhaps as a DBA, but certainly not as a System Adminsitrator or System Designer. The biggest advantage to having a single file space from Oracle's point of view are:
a) backups are simpler, and b) space management can be more granular, and c) both raw and cooked file systems can use the same APIs / formats / etc.
The number of pointers open for a given number of tables is equivalent regardless of where those pointers point. And the downside to having more granular space management is that the overhead and responsibility for doing it is added into the application -- meaning even more pointers.
There are actually advantages to separating databases out into separate table spaces, in order to distribute them across different busses, spindles, arrays, etc. (Or didn't you make it to class the day they learned that in DBA school?) This is obviously easier and more flexible with multiple cooked pieces.
All of these differences, in practical terms? Anyone running Oracle who isn't running raw partitions (and I include my current employer) is just stupid -- if you can afford the overhead of cooked partitions, why are you using Oracle in the first place? And if you are using raw partitions, how do you know what's under the covers????
Sure, there are some theoretical numbers that could get dragged out and paraded around here - big deal. Until there is an apples-to-apples comparison possible, this argument is so specious as to be ludicrous.
I think ya'll are missing the point. I block IM software (as does my company) NOT because of the arguable waste of time, but because of the rampant privacy issues and exploits in the IM world (and because it's just another way to get SPAM, or at best, people asking if they can be allowed to send you SPAM). If you want to IM, pick up the phreaking phone. Otherwise, what's wrong with e-mail? Or private / secure messaging facilities if you absolutely have to have that sort of thing?
I also had to laugh at the reports of employers who would fire people for 'excessive encrypted traffic'.More and more of what goes out of our company firewalls (and even dedicated links) is encrypted, and EVERYTHING that goes into my network (save DNS, [E]SMTP, NTP and reponses to outgoing requests), likewise. With SSH and port redirection so readily available, why would anyone NOT encrypt their personal traffic???
Please, if you are going to pick nits, at least try not to create further offenses during your diatribe. "In which one did you find 'potatos'", if you please!
Let's see; a _potentially_ harmful problem was identified -- and fixed -- thereby improving the overall Linux world, if only by some minute amount.
The equivalent in the Windows world? Oh, yeah; we know about that. Wait for the next jumbo service patch and pray.
Remind me again - Windows 2k shipped with _how many_ known bugs???
Slashdot Mirror
That's like saying we can predict the weather because we have had big supercomputers crunching those numbers for decades.
:-)
:-(
Ok, so the analogy isn't perfect, but then again, what is?
The point is, models are only as good as the data which is fed into them. Somewhere along the line, some human decided what factors should be put into the model and how they should be weighted. How accurate are those assumptions?
Personally, I'd rather not use the traditional technique to perfect those models.
Although I don't disagree that there is a lot of hype when people start discussing nukes in general, I DO think it's a little idiotic to simply consider them 'just big bombs'.
The number of atmospheric tests which have been done to date pales in comparison to the number destructive tonnage available. Furthermore, those tests have been spread out over time, rather than having been conducted all at once.
Or does anyone really believe that a nuclear confrontation in today's world would involve one plane pitching one bomb, ala WWII?
NO ONE can do more than hypothesize about the effect of multiple, multi-megaton blasts (e.g. 'nuclear winter', etc.), and let's hope no one ever can.
Unfortunately, the more people who regard nuclear explosives as 'just big bombs', the greater the odds that we'll someday obtain empirical evidence.
One thing I did not see mentioned in this thread -- if you are a single-member LLC, you enjoy the protections inherent in being a corporation, yet you still have the option to file as an S-corp.
;)
(I recently converted an S-Corp to an LLC, so I'm fairly current on LLC facts and figures.
And if you believe that, you deserve what you get.
Consider: Anyone who is withholding taxes on your behalf (which is *mandantory* and has been since WWII) is required to report that information to the IRS.
Voila! They have just 'filed' on your behalf, whether you do anything or not.
Actually, studies have long shown that the best time to submit your taxes to avoid audits is right on the cusp of the deadline. Your return has a larger chance of getting 'lost in the shuffle' this way.
;), the accountant is unlikely to have enough time to know your business well enough to 'find' more deductions, other than the 'obvious' ones.
On the other hand, there *is* a ~2% random audit process which can still result in your return being looked at in more detail - sometimes months or years after it is filed.
As for best tax filing tips -- and despite the piss poor attempts at DRM in this year's version -- skip the accountant and use TurboTax.
The other tax programs may be on a par with TT - I started using TT with its first release and have never looked back - but TT *is* a first-rate accountant - and one that never gets sleepy, overworked or makes mistakes.
Unless you're completely clueless about taxes, and can not understand the govt instructions (which are no worse than, say, the average dot-matrix printer user's manual from 1983
If you have a corporate accountant on-staff / on-call, then absolutely let them handle the taxes. You DO trust them, right?
Finally, a comment regarding filing in general: If you are owed a refund, you do NOT have to file by the deadline. If you consistently refuse to file, after a few years the IRS will file on your behalf, using stock #s from your employer-submitted W2s. At that time you can file 1040-X forms showing the real numbers and claim your refund. (Refunds expire after 5 years!)
I don't recommend this practice, but I have used and have known others who have used it.
How about the ethics of this person teaching a class for which he is admittedly not qualified? Or the ethics of using /. to compile a course syllabus?
(The _wisdom_ of the latter is beyond the scope of this comment!)
Jim, I read the cease-and-desist order, as well as the rant which has replaced the former site. And, well, all I can say is, don't you think you're overreacting???
As far as I can tell, PCI-SIG objects to the use of "PCI-SIG" and their logo, both of which seem to be reasonable objections on their part. Your site is NOT actually connected with theirs, after all.
So why not simply remove those two items and keep this list up? It almost sounds to me like you were tired of doing it, and this was just a convenient excuse.
As far as the letters "PCI" go, it is naive to believe that PCI-SIG can enforce the use or non-use of those letters in any way, shape or form. There are more than sufficient legal precedents which parallel how this mark has become a household word - and even if that were not the case, a simple addition of the appropriate R/C/SM/TM symbol, possibly with a footnote disclaimer, should more than cover the issue.
So all in all I must be missing something, as I just don't see what the big deal is all about.
Laying all jokes aside, as a friend of mine was wont to say, THEY CAN DO EXACTLY THAT.
;), but the fact is, that traffic IS being captured, those searches ARE possible, and not only is this 'legal', it's even being mandated!
Take a database of all captured internet traffic, do a search looking for the date/time and some key strings and determining the IP address from whence those "anonymous" posts originated is _trivial_.
This is BEING DONE TODAY. Are you a big enough fish to surface through the floatsam to the point where said search would be likely to be originated? Probably not. But we all know what they say about 'security through obscurity', right?
I don't mean to sound paranoid (heck, I'm even logged in
Except, of course, for the fact that the FBI has both the legal authority and the ability to have monitored the http traffic as you posted, and are likely backtracking the IP address even as we speak. So, good luck with that one....
I don't know - anywhere? I have a quote in my inbox right now for an equivalent dual-CPU Intel box w/ 2G for less than I can buy the LX50 even with my standard Sun discount.
And Sun upgrades are so economical, too. Witness the difference between the 'medium' and 'large' configurations -- you lose 36G of hard disk and gain 1G of memory -- for only a thousand dollars.
If you can't beat those prices, perhaps you'd allow me to start supplying some of your parts?
You're joking, right? The only reason Scott thinks he can sell it to the bosses is because those self-same geekoids have already implemented it anyway - something that those same bosses have finally started to notice. It's easy to justify Intel vs. Sun, especially in today's marketplace. Support on a 8-way Intel box is $5k (24x7 4-hour), while the equivalent Sun is:
E4500 base: $11,779.20
4 @ CPU/mem cards: $15,436.80
With numbers like this, it's easy to make people pay attention. (Admittedly a V880 or something would probably be cheaper than the E4500 --- but we don't own any V hardware so I have no comparative quotes there.)
Even Sun's Linux box is no bargain; I laughed when my VAR gave me a quote.
Hello?!? Earth to afidel??? a) Who cares what Dell ships, and b) we bought our first 8-way boxes over a year ago. Blade servers are seeing increased n-way performance across multiple blades, and SGI has announced (w/ Jan availability) a 64-way Itanium box which (they say) will blow to doors off an E15k.
The world is fast catching up with the best Sun has to offer, for a HECK of a lot cheaper, and with lower service contracts to boot. (And let's not even talk about the Ultra II and its designed-for-ECC-failure modules.)
This can do nothing positive except possibly generate some short term income for Bitkeeper, while at the same time has enormous negative - and long-term - consequences for same.
Forcing open source projects which are evolved enough to bootstrap onto their own code typically causes a rapid bug-fix cycle (), which is never a bad thing. For those that aren't evolved enough, having to spend unnecessary work (or money) will simply provide motivation to increase those efforts so that said bootstrapping can happen. And closed source efforts are either already licensed or can simply ignore the new dictum (so long as those political realities do not yield internal squealers, anyway).
I, for one, would be interested to see the hard numbers in license acquisition and renewals for BK for say a year in either direction of this moment - but ultimately I think they're pretty predictable.
Perhaps BK will 'merge' with MickeySoft, and disappear like so many before it....
Perhaps as a DBA, but certainly not as a System Adminsitrator or System Designer. The biggest advantage to having a single file space from Oracle's point of view are:
a) backups are simpler, and
b) space management can be more granular, and
c) both raw and cooked file systems can use the same APIs / formats / etc.
The number of pointers open for a given number of tables is equivalent regardless of where those pointers point. And the downside to having more granular space management is that the overhead and responsibility for doing it is added into the application -- meaning even more pointers.
There are actually advantages to separating databases out into separate table spaces, in order to distribute them across different busses, spindles, arrays, etc. (Or didn't you make it to class the day they learned that in DBA school?) This is obviously easier and more flexible with multiple cooked pieces.
All of these differences, in practical terms? Anyone running Oracle who isn't running raw partitions (and I include my current employer) is just stupid -- if you can afford the overhead of cooked partitions, why are you using Oracle in the first place? And if you are using raw partitions, how do you know what's under the covers????
Sure, there are some theoretical numbers that could get dragged out and paraded around here - big deal. Until there is an apples-to-apples comparison possible, this argument is so specious as to be ludicrous.
I think ya'll are missing the point. I block IM software (as does my company) NOT because of the arguable waste of time, but because of the rampant privacy issues and exploits in the IM world (and because it's just another way to get SPAM, or at best, people asking if they can be allowed to send you SPAM). If you want to IM, pick up the phreaking phone. Otherwise, what's wrong with e-mail? Or private / secure messaging facilities if you absolutely have to have that sort of thing?
I also had to laugh at the reports of employers who would fire people for 'excessive encrypted traffic'.More and more of what goes out of our company firewalls (and even dedicated links) is encrypted, and EVERYTHING that goes into my network (save DNS, [E]SMTP, NTP and reponses to outgoing requests), likewise. With SSH and port redirection so readily available, why would anyone NOT encrypt their personal traffic???
Weather.com is almost entirely Linux based, and what little isn't either belongs to The Weather Channel (and is Solaris) or is going away....
> Which one did you find 'potatos' in?
Please, if you are going to pick nits, at least try not to create further offenses during your diatribe. "In which one did you find 'potatos'", if you please!
Let's see; a _potentially_ harmful problem was identified -- and fixed -- thereby improving the overall Linux world, if only by some minute amount. The equivalent in the Windows world? Oh, yeah; we know about that. Wait for the next jumbo service patch and pray. Remind me again - Windows 2k shipped with _how many_ known bugs???