Agreed, I have nothing but great stuff to say about my Brother printer. Cheap price, excellent usability, long lifetime including toner carts, really good drivers even for OS's that are newer than the printer, and a ton of advanced features that normally are only available for high end printers.
I work for one of the biggest faceless software corporations in the world. I actually spend a lot of time thinking about our customer needs and all the little details that affect the user experience. When we receive a bug report, often I am not allowed to talk directly to the customer, but I still think of them as a person and try find a solution that will make them happy and improve our product. Many of my coworkers have a similar attitude. The financial payoff for doing this is at best extremely indirect.
The fact is that humans often seek to do a good job regardless of the financial payoff. There's a whole theory about "Intrinsic Motivation" that explains this, see here:
Actually companies like Lodsys are doing a great job of raising awareness about the weaknesses of the current system. It's great to hear that they're going after big name companies with lobbying power, rather than hurting small businesses.
If the "thousands of dollars" are less than around $5000, and if Earthlink has a company office in your state, then you might be able to take this to small claims court. It doesn't require a lawyer, and the judge would typically be biased towards the individual versus the Company.
Really? Did you try Virgin America? Their planes are new, with stylish white plastic and black leather interior, with disco color lighting. They have a Linux-based entertainment system with free games and movies, seat-to-seat chat, and a shopping-cart style electronic ordering system for food/drinks. To celebrate the holiday, the internet was free for all of December. And the price is comparable to shitty airlines like Delta and US Air. Virgin America kicks ass.
Also, don't forget that DropBox is closed source, and you can't host your own server. So your private data is subject to one of those nasty "we can change the terms at any time" kind of agreements.
This isn't an interesting cipher, mathematically speaking, because the method is closed so it could be anything. All we have is some jumbled text and (presumably) a sensible answer that we're not privy to. [...] This is just a puzzle-book, and quite boring because it can actually just be gibberish and nobody would really care.
Err... so how was it possible to decode the other three sections then? Obviously it's not gibberish, it's intelligible English text encoded using familiar algorithms. And people have cared enough to invest the time to solve three of the four sections, including people from the CIA and NSA.
Being located on actual CIA grounds, this sculpture is also an artistic statement. It highlights just how much cryptography has changed since the old days when cracking codes was top secret and required expensive supercomputers.
But then, you wouldn't know anything about art, would you? You called the Kryptos sculpture a boring children's puzzle, versus your ideal of a "militarily-important" message encoded using AES. AES never would have been invented if everyone had your pragmatist view. Math is fueled by puzzles with no obligation for usefulness.
Mint.com is pretty great for connecting to whatever bank you have and it'll download your reports and also automatic categorization. I have almost 2 years of data in it, and they let you download it all CSV. It also has me in the habit of checking all of my accounts once a week, by just logging onto one website. Nice way to be on top of anything that might be fraudulent.
What about the privacy issues of a public web site that tracks a household's entire financial profile? Intuit's claim:
"We make money only when you do - We give you personalized ideas on how to save money by presenting the greatest savings from among thousands of financial products. If you decide to make a change that saves you some cash, we sometimes earn a small fee from the bank or company you switch to. You save a lot; we make a little."
And in the Privacy and Security Policy: "Simply put, we do not and will not sell or rent your personal information to anyone, for any reason, at any time."
But they DO seem to sell your information, as long as the data format can be construed as "anonymous" (before being combined with whatever other datasets the buyer might have):
"Intuit may make anonymous or aggregate personal information and disclose such data only in a non-personally identifiable manner to:
Advertisers and other third parties for their marketing and promotional purposes, such as the number of users who applied for a credit card or how many users clicked on a particular Intuit Offer;
Organizations approved by Intuit that conduct research into consumer spending; [...]
I'd be willing to bet they make a lot from these sales. "You save a lot; we make a little" indeed. Even if Intuit's current intentions are 100% honorable, let's not overlook the ubiquitous "we can change what you agreed to without your consent" clause in the Terms Of Service:
"Intuit may modify this Agreement from time to time. Any and all changes to this Agreement will be posted on the Mint.com site. In addition, the Agreement will always indicate the date it was last revised. You are deemed to accept and agree to be bound by any changes to the Agreement when you use the Service after those changes are posted."
This practice is where all the privacy trouble started with FaceBook. I have no idea how it's legally enforceable, but somehow it is, and in fact it's standard boilerplate for TOS contracts everywhere.:-)
Just out of curiosity -- what would happen if Google preemptively denied the UK access to its services? Maybe just for a month or two? If nobody in the UK could search using google.com? What would happen?
You assume that Google uses your private information in indirect, anonymous ways to improve advertising or predict general trends. But have you looked at Google's privacy policy?
"We restrict access to personal information to Google employees, contractors and agents who need to know that information in order to operate, develop or improve our services." http://www.google.com/intl/en/privacypolicy.html
That's the extent of the promise. They can use your data to improve their "services", which obviously include every possible industry and market. The exact potential of what Google can do is difficult to foresee. They have teams of people working 24/7 on new ways to exploit every last byte of data they collect.
Here are a few ideas I came up with:
- get insider trading tips for any market by searching people's private Gmail conversations or corporate Google Docs - detect DNS names that people are brainstorming, and then preemptively squat on these domains - search for discussions/documents relating to inventions, then preemptively patent the idea - search for evidence that will convict you of a crime (copying MP3's?), maybe under a police order in some jurisdiction - use private discussions to predict locations of possible "terrorist" attacks and sell this information to the military - predict when a limited item is going to be popular, then buy up those products and sell them at higher price - help insurance companies identify "high risk" customers - use your company's internal documents to directly compete with your company
My point is that privacy isn't just about you and your porno history. It's about a global economic market that's supposed to be a fair playing field, but is instead threatened by one company's growing monopoly on everyone else's data.
The solution isn't to sick the government on Google. The solution is to educate people about why they should protect their privacy, because apparently this concept has been forgotten in all the excitement of blogs and social nets.
Wikipedia provides full citations for the author/source of all uploaded photos. If a professional photographer wanted to increase his exposure (no pun intended), he could contribute to wikipedia under a free license. The upsides really dwarf the downsides.
Her: Outlook is so slow- the messages take forever to load!
Outlook was probably slow because you were loading against Google's IMAP server on the internet, rather than from an Exchange server in the same office.:-P
Me: Well, you don't get that with a web-based system, because it is much more efficient at getting to your messages faster than your single hard drive
A web-based system is LESS efficient because nothing is cached locally. I use Gmail for some accounts, and every time I jump to a new message ("conversation"), I have to wait several seconds while the page loads. I periodically have cases where I get a blank white page because the internet connection timed out. If the network goes offline for some reason, my e-mail is totally inaccessible. Outlook (or in my case Thunderbird) has NONE of these problems because all the messages are right their on your laptop's hard disk. You can read them and search them with no internet at all.
Her: Oh. Now, is there a way I can put the same message in multiple folders without making a duplcate? Me: Actually, with Gmail you can use labels to assign one message to multiple labels, making organization much easier
In other words, NO, Gmail does not have simple folders. It has a different system called labels. If you want to use Gmail, you need to learn how to use labels, and accept the claim that labels are better than folders.
This claim might be true, but it's interesting that to this day, simple folders are still the model used by file systems. When filing, you want things to have a single location in a nice hierarchical tree. Searching is useful, but it's not the same as filing.
Google's biggest challenge is not a technical one- it's a marketing one. Google has to convince everyone that they have a product that really is better.
No, Google's challenge is to actually be better. It's great for personal e-mail, but regular people rarely need to "manage" their personal e-mail, since it's mostly chatter that expires after 1 week. By contrast, business e-mails are documents that need to be searched and sorted, with deadline pressure from the boss.
You've shown that this woman was unable to refute your spirited technical arguments. But does that really mean she's clueless and incapable of deciding which product meets her needs?
I think it's more about letting another company handle your company's email. There is so much critical information about a company in their email, why would they trust it to any external company, even if it is Google.
Thousands of companies leave their mail on other companies servers when they use Hosted Exchange. The issues usually boils down to whether or not a company wants to admin their own Exchange servers in-house.
You're comparing apples and oranges here. With hosted Exchange, you're entrusting your data to a medium-sized company that specializes in hosting Exchange. They charge a fee because that's really their business plan. With Google Apps, you're entrusting your data to a massive leviathan that aims to eventually be a competitor for every business in every industry, and who specializes in mining the hell out of everyone else's data. Google doesn't charge a fee because your data is way more valuable to them than the actual cost of hosting it.
Sure, Google has a privacy policy. But what good is a promise to only use your data to "improve our services" and "develop new services", when those "services" are completely unbounded? Google is constantly trying to invent new services, and inevitably its services will turn into a conflict of interest.
Google might be appropriate for individuals who don't see any value in data privacy. But it's not appropriate for a business.
It sounds like PrivateEye is the $19.95 edition for consumers using a simple web cam. Whereas Chameleon is the "high end" version using a special "Gazetracker" hardware device that probably has a much better reaction time. There's no price listing for Chameleon, i.e. it's intended for someone spending taxpayer's money rather than their own.
It took me awhile to figure out that this "PROTECTING DATA IN USE" image is actually an interactive Flash applet. What you do is hover the mouse over "Oculus in Action", and then wait until a blue/red oval slides across the screen. After the oval disappears, you can use the mouse to bring it back on the screen and move it around. The text inside the oval is readable, everything outside is scrambled.
The obscured text is pretty strange actually. On the "This is what the attacker sees" preview tab, the letters in each word are shuffled and easily deciphered. But on the "Oculus in Action" tab, they substitute random words of equal length, apparently sampled from a corpus of gay Satanic rites:
States --> Yapper Government --> Satanology degree --> faerie classifies --> spermatova determine --> doohinkus classifying --> luciferidae
(No joke, these are real excerpts!)
From moving the oval around and trying to read what's inside, it's pretty apparent that reaction time is extremely important to the usability of this product. Since there's no downloadable demo program (and a whole lot of marketing patter), I'm guessing that PrivateEye is way too sluggish to be practical. Chameleon might be usable, but you probably have to pay full price to find that out as well heheh.
I decided to act that way especially after Kaspersky products which are always said to be ''too heavy'' ended up saving a 512MB RAM having Celeron like low end CPU. It turns out, the ''people'' had problem with it, not us.
Kaspersky tends to be underrepresented in anti-virus discussions, maybe because they don't market as heavily. But IMO it's totally worth the price tag. I finally shelled out for Kaspersky AntiVirus (not the full firewall thing) in December of last year, when two virus infections caused enough downtime to impact my consulting hours. An Adobe PDF vulnerability was enabling my PC to be infected from simply browsing web pages with Firefox, even with AVG Internet Security fully enabled.
I tried products like Symantec and McAfee, but they're very "noisy" GUI's (in terms of advertising their presence), and it's difficult to temporarily disable them. I need this feature because I use driver debuggers and other programmer tools that conflict with antivirus services. This was a major factor in my decision to use Kaspersky, which is a very no-nonsense app with an "off" switch that works.
As far as detection rates, I browse pages and run files from a lot of (ahem) untrusted sources, and Kaspersky catches at least one real virus for me every month. No misses so far. In addition to actual threats, Kaspersky also detects potential vulnerabilities such as outdated Java or Flash DLL's, which is pretty cool. So if you can afford for-pay protection, definitely give it a try.
On my Windows machine I use Kaspersky which performs better but it was a bit of a pain to install and required that I remove Spy-bot which is a load of rubbish.
I disagree -- SpyBot is not a passive scanner. It hooks into the operating system in fairly complex ways, similar to an anti-virus program (or actual virus). You cannot expect such programs to coexist without eventually interfering with each other. I suppose Kaspersky and Safer Networking could collaborate to ensure compatibility (e.g. by providing documentation and guarantees regarding the ways they interface with the OS), but this is fairly unrealistic for two competitors.
If Microsoft provided a standardized API interface for virus scanners, the problem would be much simpler. But is that even possible? These tools defend against a very wide range of inventive attacks.
When used correctly and consistently every single time, condoms are about 98% preventive against pregnancy. However, the effectiveness rate for first-year condom users is about 86%, as only an estimated 3% of these users use condoms correctly and consistently during that time. After that milestone, the prevention rate increases, and with typical consistent use the pregnancy rate is 2-4 out of 100 women per year.
Applying basic probability, if ("ideally") the chance of pregnancy is 2%, then the chance of NOT getting pregnant over 10 years is (100%-2%)^10 = 81.7%. In other words, 1 out of 5 dudes is gonna paying child support after 10 years.
More realistically, if the failure rate is 4%, then that's (100%-4%)^10 = 66.4% chance of NOT getting pregnant. In other words, 1 out of THREE dudes will be paying child support after 10 years. Condoms suck!
And statistics is counterintuitive. If contraceptive rates were quoted per-decade rather than per-year, contraception research would actually get the funding it deserves.
Interesting point... except that a lifetime cost of $648,000 is worth way less than $400,000 up front. Give me $400,000 today, and I'll happily pay you $1800/month for the next 30 years. Just tell me where to send the checks!;-)
I tried to read Daniel Solove's PDF, but fell asleep around page 15. His citations are mostly from anonymous internet blog posters, and the essay has a rambling nature that wanders down every side street before (presumably) arriving at some kind of conclusion. That's a great style for a college PhD thesis, but I'm not a panel of professors obligated to read every page he writes. To be persuasive, he needs to convince me that his essay is worth my time. Instead of 3 pages outlining the document structure and 25 pages slowly building to a point, it should have been 3 pages summarizing the thesis and 25 pages of supplementary minutia.
If you've been told Unix stores your password encrypted somewhere, someone was glossing over the details to the point of making false statements. People can't reverse the process of decrypting your password because your password isn't stored there to begin with.
So you're saying that if your password was loosely related to a dictionary word, and if a hacker gained root access to that server, then you would be completely unconcerned about having used the same password on other servers? You would sleep well at night, with complete confidence that the text spit out by "John the Ripper" is going to be some totally unrelated text string?
In your reply, please include the IP address of your server.;-)
How can you compare Java code-completion to C++ code-completion? Java has a relatively simple grammar with no macros. C++ is a very complex language that mixes together #include's, inline assembly, old-school C syntax, compile-time execution via templates,.NET extensions, auto-generated ATL/COM wrappers, etc.
This filter is nice because it doesn't require you to remember all the values that you want to average together
Why would you need to remember all the values? As long as you remember the number of values and their total you're fine.
Not true unless you're talking about a very small number of samples. With integer or fixed point variables, the total will eventually overflow. With floating point variables, the precision errors will steadily compound and can corrupt the result.
A better solution is to store an array of readings over a time window (e.g. the past 5 seconds), and repeatedly recalculate the array average. However, this requires memory and CPU resources that often aren't available for embedded systems. People try to optimize the average by keeping a running total (i.e. adding the newest sample and subtracting the oldest sample), but this reintroduces the problem of steadily accumulating floating point errors. There are more robust algorithms (e.g. from Knuth or Welford) that address this problem.
One last point is that a windowed average is only mathematically valid for a signal that doesn't have too much noise. Otherwise you're looking at a lowpass filtering problem, and a windowed average is essentially the worst of "finite impulse response" filters.
Slashdot Mirror
Minor rearrangements of existing facts can be considered to be innovative new facts.
Consider this cool paper from Aragones et al:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=643545
Agreed, I have nothing but great stuff to say about my Brother printer. Cheap price, excellent usability, long lifetime including toner carts, really good drivers even for OS's that are newer than the printer, and a ton of advanced features that normally are only available for high end printers.
Microsoft Research has been working on this for many years and has at least two such projects:
http://en.wikipedia.org/wiki/Singularity_(operating_system)
http://en.wikipedia.org/wiki/Midori_(operating_system)
The code is ahead-of-time compiled using the Bartok compiler:
http://en.wikipedia.org/wiki/Bartok_(compiler)
I work for one of the biggest faceless software corporations in the world. I actually spend a lot of time thinking about our customer needs and all the little details that affect the user experience. When we receive a bug report, often I am not allowed to talk directly to the customer, but I still think of them as a person and try find a solution that will make them happy and improve our product. Many of my coworkers have a similar attitude. The financial payoff for doing this is at best extremely indirect.
The fact is that humans often seek to do a good job regardless of the financial payoff. There's a whole theory about "Intrinsic Motivation" that explains this, see here:
http://www.youtube.com/watch?v=u6XAPnuFjJc
Actually companies like Lodsys are doing a great job of raising awareness about the weaknesses of the current system. It's great to hear that they're going after big name companies with lobbying power, rather than hurting small businesses.
Here's some strong evidence to the contrary:
http://reverseblade.blogspot.com/2009/02/c-versus-c-versus-java-performance.html
If the "thousands of dollars" are less than around $5000, and if Earthlink has a company office in your state, then you might be able to take this to small claims court. It doesn't require a lawyer, and the judge would typically be biased towards the individual versus the Company.
Really? Did you try Virgin America? Their planes are new, with stylish white plastic and black leather interior, with disco color lighting. They have a Linux-based entertainment system with free games and movies, seat-to-seat chat, and a shopping-cart style electronic ordering system for food/drinks. To celebrate the holiday, the internet was free for all of December. And the price is comparable to shitty airlines like Delta and US Air. Virgin America kicks ass.
Also, don't forget that DropBox is closed source, and you can't host your own server. So your private data is subject to one of those nasty "we can change the terms at any time" kind of agreements.
Compare with SparkleShare...
Err... so how was it possible to decode the other three sections then? Obviously it's not gibberish, it's intelligible English text encoded using familiar algorithms. And people have cared enough to invest the time to solve three of the four sections, including people from the CIA and NSA.
Being located on actual CIA grounds, this sculpture is also an artistic statement. It highlights just how much cryptography has changed since the old days when cracking codes was top secret and required expensive supercomputers.
But then, you wouldn't know anything about art, would you? You called the Kryptos sculpture a boring children's puzzle, versus your ideal of a "militarily-important" message encoded using AES. AES never would have been invented if everyone had your pragmatist view. Math is fueled by puzzles with no obligation for usefulness.
What about the privacy issues of a public web site that tracks a household's entire financial profile? Intuit's claim:
"We make money only when you do - We give you personalized ideas on how to save money by presenting the greatest savings from among thousands of financial products. If you decide to make a change that saves you some cash, we sometimes earn a small fee from the bank or company you switch to. You save a lot; we make a little."
And in the Privacy and Security Policy: "Simply put, we do not and will not sell or rent your personal information to anyone, for any reason, at any time."
But they DO seem to sell your information, as long as the data format can be construed as "anonymous" (before being combined with whatever other datasets the buyer might have):
"Intuit may make anonymous or aggregate personal information and disclose such data only in a non-personally identifiable manner to:
I'd be willing to bet they make a lot from these sales. "You save a lot; we make a little" indeed. Even if Intuit's current intentions are 100% honorable, let's not overlook the ubiquitous "we can change what you agreed to without your consent" clause in the Terms Of Service:
"Intuit may modify this Agreement from time to time. Any and all changes to this Agreement will be posted on the Mint.com site. In addition, the Agreement will always indicate the date it was last revised. You are deemed to accept and agree to be bound by any changes to the Agreement when you use the Service after those changes are posted."
This practice is where all the privacy trouble started with FaceBook. I have no idea how it's legally enforceable, but somehow it is, and in fact it's standard boilerplate for TOS contracts everywhere. :-)
Just out of curiosity -- what would happen if Google preemptively denied the UK access to its services? Maybe just for a month or two? If nobody in the UK could search using google.com? What would happen?
You assume that Google uses your private information in indirect, anonymous ways to improve advertising or predict general trends. But have you looked at Google's privacy policy?
"We restrict access to personal information to Google employees, contractors and agents who need to know that information in order to operate, develop or improve our services."
http://www.google.com/intl/en/privacypolicy.html
That's the extent of the promise. They can use your data to improve their "services", which obviously include every possible industry and market. The exact potential of what Google can do is difficult to foresee. They have teams of people working 24/7 on new ways to exploit every last byte of data they collect.
Here are a few ideas I came up with:
- get insider trading tips for any market by searching people's private Gmail conversations or corporate Google Docs
- detect DNS names that people are brainstorming, and then preemptively squat on these domains
- search for discussions/documents relating to inventions, then preemptively patent the idea
- search for evidence that will convict you of a crime (copying MP3's?), maybe under a police order in some jurisdiction
- use private discussions to predict locations of possible "terrorist" attacks and sell this information to the military
- predict when a limited item is going to be popular, then buy up those products and sell them at higher price
- help insurance companies identify "high risk" customers
- use your company's internal documents to directly compete with your company
My point is that privacy isn't just about you and your porno history. It's about a global economic market that's supposed to be a fair playing field, but is instead threatened by one company's growing monopoly on everyone else's data.
The solution isn't to sick the government on Google. The solution is to educate people about why they should protect their privacy, because apparently this concept has been forgotten in all the excitement of blogs and social nets.
Wikipedia provides full citations for the author/source of all uploaded photos. If a professional photographer wanted to increase his exposure (no pun intended), he could contribute to wikipedia under a free license. The upsides really dwarf the downsides.
-Gonz
Her: Outlook is so slow- the messages take forever to load!
Outlook was probably slow because you were loading against Google's IMAP server on the internet, rather than from an Exchange server in the same office. :-P
Me: Well, you don't get that with a web-based system, because it is much more efficient at getting to your messages faster than your single hard drive
A web-based system is LESS efficient because nothing is cached locally. I use Gmail for some accounts, and every time I jump to a new message ("conversation"), I have to wait several seconds while the page loads. I periodically have cases where I get a blank white page because the internet connection timed out. If the network goes offline for some reason, my e-mail is totally inaccessible. Outlook (or in my case Thunderbird) has NONE of these problems because all the messages are right their on your laptop's hard disk. You can read them and search them with no internet at all.
Her: Oh. Now, is there a way I can put the same message in multiple folders without making a duplcate?
Me: Actually, with Gmail you can use labels to assign one message to multiple labels, making organization much easier
In other words, NO, Gmail does not have simple folders. It has a different system called labels. If you want to use Gmail, you need to learn how to use labels, and accept the claim that labels are better than folders.
This claim might be true, but it's interesting that to this day, simple folders are still the model used by file systems. When filing, you want things to have a single location in a nice hierarchical tree. Searching is useful, but it's not the same as filing.
Google's biggest challenge is not a technical one- it's a marketing one. Google has to convince everyone that they have a product that really is better.
No, Google's challenge is to actually be better. It's great for personal e-mail, but regular people rarely need to "manage" their personal e-mail, since it's mostly chatter that expires after 1 week. By contrast, business e-mails are documents that need to be searched and sorted, with deadline pressure from the boss.
You've shown that this woman was unable to refute your spirited technical arguments. But does that really mean she's clueless and incapable of deciding which product meets her needs?
-Gonz
I think it's more about letting another company handle your company's email. There is so much critical information about a company in their email, why would they trust it to any external company, even if it is Google.
Thousands of companies leave their mail on other companies servers when they use Hosted Exchange. The issues usually boils down to whether or not a company wants to admin their own Exchange servers in-house.
You're comparing apples and oranges here. With hosted Exchange, you're entrusting your data to a medium-sized company that specializes in hosting Exchange. They charge a fee because that's really their business plan. With Google Apps, you're entrusting your data to a massive leviathan that aims to eventually be a competitor for every business in every industry, and who specializes in mining the hell out of everyone else's data. Google doesn't charge a fee because your data is way more valuable to them than the actual cost of hosting it.
Sure, Google has a privacy policy. But what good is a promise to only use your data to "improve our services" and "develop new services", when those "services" are completely unbounded? Google is constantly trying to invent new services, and inevitably its services will turn into a conflict of interest.
Google might be appropriate for individuals who don't see any value in data privacy. But it's not appropriate for a business.
-Gonz
The web site doesn't clearly explain the difference between "Chameleon" versus "PrivateEye". I found the answers in this PDF:
http://oculislabs.com/Oculis_Whitepaper_1.pdf
It sounds like PrivateEye is the $19.95 edition for consumers using a simple web cam. Whereas Chameleon is the "high end" version using a special "Gazetracker" hardware device that probably has a much better reaction time. There's no price listing for Chameleon, i.e. it's intended for someone spending taxpayer's money rather than their own.
The demo is on the Chameleon page:
http://oculislabs.com/Products/ChameleonP.htm
It took me awhile to figure out that this "PROTECTING DATA IN USE" image is actually an interactive Flash applet. What you do is hover the mouse over "Oculus in Action", and then wait until a blue/red oval slides across the screen. After the oval disappears, you can use the mouse to bring it back on the screen and move it around. The text inside the oval is readable, everything outside is scrambled.
The obscured text is pretty strange actually. On the "This is what the attacker sees" preview tab, the letters in each word are shuffled and easily deciphered. But on the "Oculus in Action" tab, they substitute random words of equal length, apparently sampled from a corpus of gay Satanic rites:
States --> Yapper
Government --> Satanology
degree --> faerie
classifies --> spermatova
determine --> doohinkus
classifying --> luciferidae
(No joke, these are real excerpts!)
From moving the oval around and trying to read what's inside, it's pretty apparent that reaction time is extremely important to the usability of this product. Since there's no downloadable demo program (and a whole lot of marketing patter), I'm guessing that PrivateEye is way too sluggish to be practical. Chameleon might be usable, but you probably have to pay full price to find that out as well heheh.
-Gonz
I decided to act that way especially after Kaspersky products which are always said to be ''too heavy'' ended up saving a 512MB RAM having Celeron like low end CPU. It turns out, the ''people'' had problem with it, not us.
Kaspersky tends to be underrepresented in anti-virus discussions, maybe because they don't market as heavily. But IMO it's totally worth the price tag. I finally shelled out for Kaspersky AntiVirus (not the full firewall thing) in December of last year, when two virus infections caused enough downtime to impact my consulting hours. An Adobe PDF vulnerability was enabling my PC to be infected from simply browsing web pages with Firefox, even with AVG Internet Security fully enabled.
I tried products like Symantec and McAfee, but they're very "noisy" GUI's (in terms of advertising their presence), and it's difficult to temporarily disable them. I need this feature because I use driver debuggers and other programmer tools that conflict with antivirus services. This was a major factor in my decision to use Kaspersky, which is a very no-nonsense app with an "off" switch that works.
As far as detection rates, I browse pages and run files from a lot of (ahem) untrusted sources, and Kaspersky catches at least one real virus for me every month. No misses so far. In addition to actual threats, Kaspersky also detects potential vulnerabilities such as outdated Java or Flash DLL's, which is pretty cool. So if you can afford for-pay protection, definitely give it a try.
-Gonz
On my Windows machine I use Kaspersky which performs better but it was a bit of a pain to install and required that I remove Spy-bot which is a load of rubbish.
I disagree -- SpyBot is not a passive scanner. It hooks into the operating system in fairly complex ways, similar to an anti-virus program (or actual virus). You cannot expect such programs to coexist without eventually interfering with each other. I suppose Kaspersky and Safer Networking could collaborate to ensure compatibility (e.g. by providing documentation and guarantees regarding the ways they interface with the OS), but this is fairly unrealistic for two competitors.
If Microsoft provided a standardized API interface for virus scanners, the problem would be much simpler. But is that even possible? These tools defend against a very wide range of inventive attacks.
-Gonz
Despite that they're a good trade when weighed against the possibility of 18 years of child support, or your penis turning green and falling off.
Not really. This site has some statistics:
When used correctly and consistently every single time, condoms are about 98% preventive against pregnancy. However, the effectiveness rate for first-year condom users is about 86%, as only an estimated 3% of these users use condoms correctly and consistently during that time. After that milestone, the prevention rate increases, and with typical consistent use the pregnancy rate is 2-4 out of 100 women per year.
Applying basic probability, if ("ideally") the chance of pregnancy is 2%, then the chance of NOT getting pregnant over 10 years is (100%-2%)^10 = 81.7%. In other words, 1 out of 5 dudes is gonna paying child support after 10 years.
More realistically, if the failure rate is 4%, then that's (100%-4%)^10 = 66.4% chance of NOT getting pregnant. In other words, 1 out of THREE dudes will be paying child support after 10 years. Condoms suck!
And statistics is counterintuitive. If contraceptive rates were quoted per-decade rather than per-year, contraception research would actually get the funding it deserves.
-Gonz
Interesting point... except that a lifetime cost of $648,000 is worth way less than $400,000 up front. Give me $400,000 today, and I'll happily pay you $1800/month for the next 30 years. Just tell me where to send the checks! ;-)
-Gonz
I tried to read Daniel Solove's PDF, but fell asleep around page 15. His citations are mostly from anonymous internet blog posters, and the essay has a rambling nature that wanders down every side street before (presumably) arriving at some kind of conclusion. That's a great style for a college PhD thesis, but I'm not a panel of professors obligated to read every page he writes. To be persuasive, he needs to convince me that his essay is worth my time. Instead of 3 pages outlining the document structure and 25 pages slowly building to a point, it should have been 3 pages summarizing the thesis and 25 pages of supplementary minutia.
-Gonz
If you've been told Unix stores your password encrypted somewhere, someone was glossing over the details to the point of making false statements. People can't reverse the process of decrypting your password because your password isn't stored there to begin with.
So you're saying that if your password was loosely related to a dictionary word, and if a hacker gained root access to that server, then you would be completely unconcerned about having used the same password on other servers? You would sleep well at night, with complete confidence that the text spit out by "John the Ripper" is going to be some totally unrelated text string?
In your reply, please include the IP address of your server. ;-)
-Gonz
How can you compare Java code-completion to C++ code-completion? Java has a relatively simple grammar with no macros. C++ is a very complex language that mixes together #include's, inline assembly, old-school C syntax, compile-time execution via templates, .NET extensions, auto-generated ATL/COM wrappers, etc.
-Gonz
This filter is nice because it doesn't require you to remember all the values that you want to average together
Why would you need to remember all the values? As long as you remember the number of values and their total you're fine.
Not true unless you're talking about a very small number of samples. With integer or fixed point variables, the total will eventually overflow. With floating point variables, the precision errors will steadily compound and can corrupt the result.
A better solution is to store an array of readings over a time window (e.g. the past 5 seconds), and repeatedly recalculate the array average. However, this requires memory and CPU resources that often aren't available for embedded systems. People try to optimize the average by keeping a running total (i.e. adding the newest sample and subtracting the oldest sample), but this reintroduces the problem of steadily accumulating floating point errors. There are more robust algorithms (e.g. from Knuth or Welford) that address this problem.
One last point is that a windowed average is only mathematically valid for a signal that doesn't have too much noise. Otherwise you're looking at a lowpass filtering problem, and a windowed average is essentially the worst of "finite impulse response" filters.
-Gonz