Sure the computer knows if a fruity aftertaste is desired -- it's called a configuration parameter. And while the humans know the desired results, they may not have the observational ability to guide it there.
A basic machine learning approach would go as follows: * prepare several batches in slightly different ways * measure everything imaginable at various stages of the growing and vinting process * when the wine is ready, ask experts to rate values like "fruity aftertaste" and "body strength" * throw all the numbers together and look for patterns * next time around, make adjustments based on observed patterns * evaluate the next batch and feed it back into the program
My mother in law likes to send cash in the mail (wrapped in tinfoil, no less). We were wondering out loud if we could get her to stop. She seems incapable of using a check, though she must have an account. I wondered aloud if we should explain PayPal to her, but then I wondered if she'd fall for those spams. Since PayPal spam is probably more frequent than people stealing USPS mail, we resigned ourselves to receiving cash (unrequested) for birthdays and holidays.
I installed Cyberduck on my mom's Mac for FTP access to her website. I like the application -- it's got an easy drag-n-drop interface, it supports SFTP, it's free and open source. But the best feature is that its' dock icon is a big rubber duck. So when my mom calls and says "How do I edit my web page again?" I can say "Okay... now drag that file from your desktop to the duck." She has made the mental link that the duck will take the files in its bill and fly to the web server, where everyone can see it. She can understand the less metaphoric version, but it's the colorful explanation that makes her remember what to do.
In the developer cubicles, when we want to get someone's attention we often throw something -- a pen cap, a soft frisbee, a packet of parmesan cheese -- at the person so they'll take off their headphones and look around. We call it an "Instant Message."
One of the gifts at our office White Elephant party yesterday was about 30 small koosh balls. That ought to keep things interesting.
My university graduated more CS majors after the dot com crash than before it. Partly this was due to general growth, partly it was due to a major decrease in demand for programmers without a degree, party it was due to laid off tech professionals going back to school. But I didn't meet anyone who dropped out of the CS department because prospective salaries were lower. There were only a few girls in my freshman year CS classes and there was about the same percentage during my senior year. (If I recall, there were 7-10 girls in my 100+ person senior projects class, which everyone had to take.)
There are broadly two reasons to pursue a computer science degree. The first is because the pay is good. The second is because the student loves problem solving, abstract and applied math, and making computers do cool stuff.
Geeks usually don't care about social stigmata, negative stereotypes, or lack of hygene associated with a subject field as long as they get to play with stuff they think is cool. The important question is this: Of the girl geeks (and there are many), why aren't there very many girl *computer* geeks?
(One could an analogous question: Of the boy geeks (and there are many), why aren't there vary many boy *knitting* geeks? I don't know if the answers are similar.)
I'm not sure if browsers have gotten smarter about this, but if you specify pixel sizes on images, the browser can leave the apropriate amount of space open so the page text doesn't jump wildly while modem users are reading it. Back in the day, some browsers wouldn't even render the page until they knew how big each image was going to be.
And while I've never surfed the web on a cell phone, I'll bet my webpages look just fine in them. They looked pretty good in lynx last I checked. And most of them probably work well in an aural browser. That's because they're almost all text. Coincidentally, there's a very high content-to-bandwidth correlation.
Now Mac-using graphic designers, confronted with sites that don't work in IE, can claim that if Microsoft wanted their browser to work they would provide it to the people who produce web content.:-)
If the sunset provisions are demonstrably a boon to national security, the DoJ is free to write a bill with less hasty and better reviewed language and have Congress authorize the powers again.
Well I just searched for Comfortably Numb (not an obscure song by any stretch) and didn't find any "at the top results," but I did get the same hits to lyrics and Pink Floyd that I did before.
Searches for some less well-known artists (The Moldy Peaches, Momus, Robert Johnson) didn't find any special results, but the first hit was what I was looking for. A search for Electric Ladyland finds lots of links to buy the album, but no special Jimi Hendrix links.
But at least searching for U2 gives me a picture of the band and not a spy plane.
The goal is therefore to break your program up into appropriate parts so that almost everything is trivial. You can therefore comment the parts that are non-trivial and not worry about the rest.
The ideal method is a few lines long and has a descriptive name and parameters. The test for this quality is to try to write a comment for it -- if you find you're just repeating the code, you've done a good job.
public Collection getUnripeFruits() {
Collection result = new LinkedList();
for (Iterator iter = fruits.iterator(); iter.hasNext();) {
Fruit fruit = (Fruit) iter.next();
if (!fruit.isRipe() && !fruit.isRotten()) {
result.add(fruit);
}
}
return result; }
You're mistaking disinterested for dispassionate and uninterested. The ideal situation for a researcher is to be very passionate about the work, but not be passionate about a particular outcome. Disinterested means you don't have a vested interest in a specific result. Uninterested means the matter doesn't hold your attention. They are very different.
A scientist should be emotional about the experiment. Ego will be wrapped up -- good science will boost the ego, bad science will crush it. But if you're doing research with the goal of showing a specific result (rather than testing a hypothesis) you run the danger of tainting the results.
My job involves integrating with OCR, and we've looked at quite a few options. Though there are some bargains, you get what you pay for.
The big players are Abbyy and Scansoft. Both have extensive feature lists, from handy GUIs to form/document layout to Asian language support. They also come with a hefty price tag. Their Windows support is best, but they have software for others. Single user applications are reasonably priced in the two-digit figure range. However, we decided not to integrate with either of them in part because of the price tag for high volume server-side processing. If you only have a hundred or two forms to do at a time, a workstation solution may be your best bet.
We chose to integrate with Transym, a cheap but pretty good engine. It does a good job at what it tries to do, which is recognize standard printed text. We then take that text and extract meaningful data, like dates and names, from the output text + position information. Pretty much every other cheap/free package we looked at had pretty lousy performance on our straight-forward documents (primarily typed paragraphs).
ICR (recognizing handwriting) and IMR (mark recognition) is another bag. There are very few players in this arena. They work best when the domain is well-defined (the U.S. Postal Service, for instance, does pretty well at recognizing zip codes). If you're trying to recognize dates and check boxes, the form definition software that Abbyy and Scansoft provide probably fits your needs best.
Finally, you need to consider how reliable you want your process to be and how much quality control you want. Even the best OCR engine makes errors, and ICR is quite a bit behind that. You can't blindly trust OCR output unless you're willing to deal with incomplete data. If you're going to have a human verify the computer's work for only a few fields, you may not be gaining significant efficiency.
(I don't claim to have evaluated every potential option. There may be software we missed, software we didn't evaluate because it didn't meet our integration needs, and software that's come to light after we did our search.)
Clearly the time I would spend putting on a tie and combing my hair in the morning could be better spent writing code. Unfortunately, it usually gets spent making disapproving comments on Bugzilla.
Actual line from our standup meetings: "We'll have customers in the office tomorrow, so be sure to wear a clean shirt."
I don't know that I'd describe our current electricity distribution system is "perfectly functional." There are more cars than people in the U.S. And while a change to electric cars wouldn't happen overnight, having millions of cars charging every night would require a major increase in production capacity, probably coupled with an increase in electric bills for non-electric car users (due to the law of supply and demand).
IPv6 fans ought to read D.J. Bernstein's excellent article on the subject. In short, the main problem is that the two protocols aren't easily interoperable, so investment in IPv6 infrastructure is without short-term return.
Slashdot Mirror
Sure the computer knows if a fruity aftertaste is desired -- it's called a configuration parameter. And while the humans know the desired results, they may not have the observational ability to guide it there.
A basic machine learning approach would go as follows:
* prepare several batches in slightly different ways
* measure everything imaginable at various stages of the growing and vinting process
* when the wine is ready, ask experts to rate values like "fruity aftertaste" and "body strength"
* throw all the numbers together and look for patterns
* next time around, make adjustments based on observed patterns
* evaluate the next batch and feed it back into the program
My mother in law likes to send cash in the mail (wrapped in tinfoil, no less). We were wondering out loud if we could get her to stop. She seems incapable of using a check, though she must have an account. I wondered aloud if we should explain PayPal to her, but then I wondered if she'd fall for those spams. Since PayPal spam is probably more frequent than people stealing USPS mail, we resigned ourselves to receiving cash (unrequested) for birthdays and holidays.
I installed Cyberduck on my mom's Mac for FTP access to her website. I like the application -- it's got an easy drag-n-drop interface, it supports SFTP, it's free and open source. But the best feature is that its' dock icon is a big rubber duck. So when my mom calls and says "How do I edit my web page again?" I can say "Okay... now drag that file from your desktop to the duck." She has made the mental link that the duck will take the files in its bill and fly to the web server, where everyone can see it. She can understand the less metaphoric version, but it's the colorful explanation that makes her remember what to do.
In the developer cubicles, when we want to get someone's attention we often throw something -- a pen cap, a soft frisbee, a packet of parmesan cheese -- at the person so they'll take off their headphones and look around. We call it an "Instant Message."
One of the gifts at our office White Elephant party yesterday was about 30 small koosh balls. That ought to keep things interesting.
Just like in The Gold Rush, many Indians are forced to eat their own shoes to survive.
My university graduated more CS majors after the dot com crash than before it. Partly this was due to general growth, partly it was due to a major decrease in demand for programmers without a degree, party it was due to laid off tech professionals going back to school. But I didn't meet anyone who dropped out of the CS department because prospective salaries were lower. There were only a few girls in my freshman year CS classes and there was about the same percentage during my senior year. (If I recall, there were 7-10 girls in my 100+ person senior projects class, which everyone had to take.)
There are broadly two reasons to pursue a computer science degree. The first is because the pay is good. The second is because the student loves problem solving, abstract and applied math, and making computers do cool stuff.
Geeks usually don't care about social stigmata, negative stereotypes, or lack of hygene associated with a subject field as long as they get to play with stuff they think is cool. The important question is this: Of the girl geeks (and there are many), why aren't there very many girl *computer* geeks?
(One could an analogous question: Of the boy geeks (and there are many), why aren't there vary many boy *knitting* geeks? I don't know if the answers are similar.)
This is only fitting, given that RSS is a name for four or five very different XML structures.
I'm not sure if browsers have gotten smarter about this, but if you specify pixel sizes on images, the browser can leave the apropriate amount of space open so the page text doesn't jump wildly while modem users are reading it. Back in the day, some browsers wouldn't even render the page until they knew how big each image was going to be.
And while I've never surfed the web on a cell phone, I'll bet my webpages look just fine in them. They looked pretty good in lynx last I checked. And most of them probably work well in an aural browser. That's because they're almost all text. Coincidentally, there's a very high content-to-bandwidth correlation.
Now Mac-using graphic designers, confronted with sites that don't work in IE, can claim that if Microsoft wanted their browser to work they would provide it to the people who produce web content. :-)
If the sunset provisions are demonstrably a boon to national security, the DoJ is free to write a bill with less hasty and better reviewed language and have Congress authorize the powers again.
Old programmers don't die, they just get garbage collected.
Well I just searched for Comfortably Numb (not an obscure song by any stretch) and didn't find any "at the top results," but I did get the same hits to lyrics and Pink Floyd that I did before.
Searches for some less well-known artists (The Moldy Peaches, Momus, Robert Johnson) didn't find any special results, but the first hit was what I was looking for. A search for Electric Ladyland finds lots of links to buy the album, but no special Jimi Hendrix links.
But at least searching for U2 gives me a picture of the band and not a spy plane.
The goal is therefore to break your program up into appropriate parts so that almost everything is trivial. You can therefore comment the parts that are non-trivial and not worry about the rest.
The ideal method is a few lines long and has a descriptive name and parameters. The test for this quality is to try to write a comment for it -- if you find you're just repeating the code, you've done a good job.
public Collection getUnripeFruits() {
Collection result = new LinkedList();
for (Iterator iter = fruits.iterator(); iter.hasNext();) {
Fruit fruit = (Fruit) iter.next();
if (!fruit.isRipe() && !fruit.isRotten()) {
result.add(fruit);
}
}
return result;
}
No comment.
All we need now is a Tesla-style tower to deliver free wireless power to New Orleans!
I'm betting the domains x.y and x.x will be highly sought after in the lucrative gay dating site market.
You're mistaking disinterested for dispassionate and uninterested. The ideal situation for a researcher is to be very passionate about the work, but not be passionate about a particular outcome. Disinterested means you don't have a vested interest in a specific result. Uninterested means the matter doesn't hold your attention. They are very different.
A scientist should be emotional about the experiment. Ego will be wrapped up -- good science will boost the ego, bad science will crush it. But if you're doing research with the goal of showing a specific result (rather than testing a hypothesis) you run the danger of tainting the results.
My job involves integrating with OCR, and we've looked at quite a few options. Though there are some bargains, you get what you pay for.
The big players are Abbyy and Scansoft. Both have extensive feature lists, from handy GUIs to form/document layout to Asian language support. They also come with a hefty price tag. Their Windows support is best, but they have software for others. Single user applications are reasonably priced in the two-digit figure range. However, we decided not to integrate with either of them in part because of the price tag for high volume server-side processing. If you only have a hundred or two forms to do at a time, a workstation solution may be your best bet.
We chose to integrate with Transym, a cheap but pretty good engine. It does a good job at what it tries to do, which is recognize standard printed text. We then take that text and extract meaningful data, like dates and names, from the output text + position information. Pretty much every other cheap/free package we looked at had pretty lousy performance on our straight-forward documents (primarily typed paragraphs).
ICR (recognizing handwriting) and IMR (mark recognition) is another bag. There are very few players in this arena. They work best when the domain is well-defined (the U.S. Postal Service, for instance, does pretty well at recognizing zip codes). If you're trying to recognize dates and check boxes, the form definition software that Abbyy and Scansoft provide probably fits your needs best.
Finally, you need to consider how reliable you want your process to be and how much quality control you want. Even the best OCR engine makes errors, and ICR is quite a bit behind that. You can't blindly trust OCR output unless you're willing to deal with incomplete data. If you're going to have a human verify the computer's work for only a few fields, you may not be gaining significant efficiency.
(I don't claim to have evaluated every potential option. There may be software we missed, software we didn't evaluate because it didn't meet our integration needs, and software that's come to light after we did our search.)
Nah, he'll just pull the plug.
I never thought I'd hear "Edison cylinder" and "state of the art" in the same sentence.
Clearly the time I would spend putting on a tie and combing my hair in the morning could be better spent writing code. Unfortunately, it usually gets spent making disapproving comments on Bugzilla.
Actual line from our standup meetings: "We'll have customers in the office tomorrow, so be sure to wear a clean shirt."
Firms using windows are banging their head against a wall.
MacOS, BSD, Unix... what's the difference?
I don't know that I'd describe our current electricity distribution system is "perfectly functional." There are more cars than people in the U.S. And while a change to electric cars wouldn't happen overnight, having millions of cars charging every night would require a major increase in production capacity, probably coupled with an increase in electric bills for non-electric car users (due to the law of supply and demand).
IPv6 fans ought to read D.J. Bernstein's excellent article on the subject. In short, the main problem is that the two protocols aren't easily interoperable, so investment in IPv6 infrastructure is without short-term return.
So God implemented the eye several times.
So He created some species and then decided they weren't very good. He refactored Homo erectus into Homo habilus!
So God was lazy and screweed around for the first days of the project, then got under a deadline rush and had to work on the weekends.
I see nothing to discount my theory that God is a programmer. Maybe He even reads Slashdot.
And I can't blame Him for some of the problems of the world. DNA is a tough API, and the best debugging tool is a toad.