Huh? There are tons of people who start swearing & cursing out loud in places like McDonald's. The only reason they'd get kicked out is if they appeared to pose a threat to the safety of people around them, or started interfering with the business itself depending on how long it went on. If they're going on for a while, even if they never said anything profane or offensive, they still should be asked to leave, because it is the disruptiveness, not offensiveness, that is the problem.
Posting a picture of a common minor vulgarity is not disruptive to anything.
He claims there are death threats. He's shown himself to say anything regardless of truth. The people bugging him are a bunch of riled-up webcomic readers doing it for the lulz, probably using real contact info.
I don't believe he's received actual death threats, but is playing up a sympathy card.
I've heard that Apple is pursuing the 3840x2160 "4K" resolution as a more accepted standard. The old 3840x2400 displays were 22" at >200dpi, the current 3840x2160 monitors used for cinema development and other high-end stuff is 40"+ at "normal" dot pitches. 32"-37" seems appropriate for this resolution.
Also IIRC, the 30" cinema display was 2560x1600, the 27" one was 2560x1440. The shift was coincident with the manufacturer's rejection of the 16:10 form factor in favor of the short-screen 16:9 to match TV aspect ratios, not consumer rejection of the size.
Jan 1 = Sunday, 30 days Feb 1 = Tuesday, 30 days Mar 1 = Thursday, 31 days
Apr 1 = Sunday, 30 days May 1 = Tuesday, 30 days Jun 1 = Thursday, 31 days...
Then every 5-6 years, there's a leap *week* at the end of the year after December called Xtr, so Xtr 1, 2015 through Xtr 7, 2015 would exist as valid dates (in whatever order your country uses).
Seeing as you're just piling boxes around, why not just pile harddrives around the computer? As long as your motherboard has enough SATA ports, you don't *need* to have a fancy case with physical slots for all the HDs. Consumer-grade motherboards with 6-8 SATA slots with a few more eSATA are not uncommon.
Then again, if you only have 3TB, buying some cheaper 1-2TB drives, integrating them into your mix and eventually moving your data over, should allow you to power down a lot of your equipment especially if you're mostly using it for storage, not processing.
What we need to do is not "integrate" tech into a classroom setting, but replace the classroom setting with tech. Then we can use the classroom for group work, question/answer or discussion sessions, and demonstrations. Learning a lot of the core materials solely through reading, watching videos, and automated quizzing can get you a long way through the tedium normally found in the classroom; then you can use the face-time of the school facilities for better purposes.
But just throwing shiny toys in kids' hands is a recipe for disaster.
I agree, a lot of the corporate training stuff is not that helpful; it's often too condensed into a continual barrage of nearly-opaque info. Having online courses like this is a great balance between such problems and those of a university setting. (I took the Stanford AI and ML classes)
Of course it's theory; CS is literally the science of computation. It's not about engineering.
Granted, getting good engineering out of software devs reveals a gaping education/training hole that needs to be addressed, but that is not Computer Science.
University education is about personal enrichment and beyond-the-industry research, not job training. The current western world remains confused about this, with employers who think a degree means the applicant automatically has good on-the-job capabilities (especially in computing), and students thinking a piece of very expensive paper is a magic life betterment token (correlation is not causation).
This experience is fine when you're student-age, but things change once you're a working professional adult who seeks to increase their own understanding for things they're already doing for real. Having well-presented information that's tangibly useful, instead of having to perform lots of self-learning to have it sink in, is exactly what fits the bill.
Many educators seek to create some sort of "community of education" around a student lifestyle, instead of offering direct quality knowledge to experienced customers.
The Stanford AI course had lots of technical problems that they were unprepared for - ambiguous English phrasing, uneven level of practice versus test, missing technical explanations, and so on.
But the Machine Learning class was almost perfect. Incredibly clear, precise, complete and human-friendly explanations; gave a thorough treatment of real-world caveats, testing, debugging tips; wasn't so ridiculously focused on locking in grades; and all around a way better experience than the AI class (I took both). I have nothing but respect for Prof. Ng and his staff.
There really was no excuse for the problems the AI class had in its teaching style, and those problems really didn't have anything to do with the technology.
Ah, but here we're talking about "shadow accounts", so there are no posts from Steve to associate with.
Why not? Steve might post on twitter, or wherever. He probably uses the same email address for various accounts, or reveals his full name somewhere in various other public online identities he uses. The computer doesn't *care* how spurious or inconsequential links between data are, it can aggregate them.
Also, if you think that Facebook only has internal Facebook data to work with, you're being incredibly naive. They have "value" in getting as much information about everybody from everywhere they possibly can. There have been comments posted here on Slashdot about 2 people, each without Facebook accounts, that get Facebook invites from each other. They both ended up thinking "Well, if *they* finally broke down and got an account there, maybe I should too", and only found out that the invites were unsolicited afterwards. That was all formulated from information neither has ever submitted to Facebook.
Right. So if I post "Steve still owes me that $50, don't think he's ever going to pay it back" on facebook, what's the data mining heuristic by which one of the people named Steve gets his credit rating adjusted?
If you mentioned in a prior post that you went out to dinner with some friends and it was terrible because the food was bad and Steve forgot his wallet, and Steve posts about having a good time with his friends at dinner that night, then an association can be made that the one who's not paying you back is the one who has links to you who posted that he went out to dinner that same night. Sure, it's a stretch, but the more links like these there are, the stronger the mined beliefs become. Much of this sort of association can be done with simple vocabulary filters yielding very low-probability results, but again you get a ton of these little improbabilities adding up and it becomes a probable result.
Sorry, it just doesn't work.
This is exactly how it works. Any single data point has very little information. Tons of little data points in aggregate reveals new information that cannot be inferred from looking at the points on their own. Comparing time stamps between separate information sources also reveals connections that are not expressed verbally.
Facebook has "shadow accounts" of people it's datamined, but have not actually opened a facebook account with. So even if you don't have an account, your friends & family talking about you, or any data it harvests that contain references to you, are stored there.
I would suspect that Facebook's sharing of information with legal authorities includes shadow accounts, and if banks or credit reporting agencies strike a deal with Facebook to get at "indicator" information about people, they'd be able to view shadow accounts as well.
The whole original point of patents was to document inventions so that they'd get out into the public, in exchange for a temporary legal limitation on who can use it. Without them, all sorts of inventions were used internally as trade secret, and as people passed on or changed careers, innovation was lost.
The PTO is granting way too many junk patents (I haven't RTFA, so I don't know how garbage this one is), but if fewer patents were granted, or terms were shortened based on some metric of innovation or uniqueness, things would be better off. Problem is, the government believes that the number of patents granted per year is a productivity & innovation metric; in reality with such quantity granted it's just legal strong-arming and innovating stifling.
Slashdot Mirror
Huh? There are tons of people who start swearing & cursing out loud in places like McDonald's. The only reason they'd get kicked out is if they appeared to pose a threat to the safety of people around them, or started interfering with the business itself depending on how long it went on. If they're going on for a while, even if they never said anything profane or offensive, they still should be asked to leave, because it is the disruptiveness, not offensiveness, that is the problem.
Posting a picture of a common minor vulgarity is not disruptive to anything.
He claims there are death threats. He's shown himself to say anything regardless of truth. The people bugging him are a bunch of riled-up webcomic readers doing it for the lulz, probably using real contact info.
I don't believe he's received actual death threats, but is playing up a sympathy card.
There is a strong selection bias here. Your friend spends a lot of time with people where it works, and very little time on people who just hang up.
You do know that all 3 current gen game consoles are driven by PPC CPUs, right?
I've heard that Apple is pursuing the 3840x2160 "4K" resolution as a more accepted standard. The old 3840x2400 displays were 22" at >200dpi, the current 3840x2160 monitors used for cinema development and other high-end stuff is 40"+ at "normal" dot pitches. 32"-37" seems appropriate for this resolution.
Also IIRC, the 30" cinema display was 2560x1600, the 27" one was 2560x1440. The shift was coincident with the manufacturer's rejection of the 16:10 form factor in favor of the short-screen 16:9 to match TV aspect ratios, not consumer rejection of the size.
And this is why the "Pirate Party" needs to change their name. I suggest the "Open Party".
Jan 1 = Sunday, 30 days
Feb 1 = Tuesday, 30 days
Mar 1 = Thursday, 31 days
Apr 1 = Sunday, 30 days ...
May 1 = Tuesday, 30 days
Jun 1 = Thursday, 31 days
Then every 5-6 years, there's a leap *week* at the end of the year after December called Xtr, so Xtr 1, 2015 through Xtr 7, 2015 would exist as valid dates (in whatever order your country uses).
Seeing as you're just piling boxes around, why not just pile harddrives around the computer? As long as your motherboard has enough SATA ports, you don't *need* to have a fancy case with physical slots for all the HDs. Consumer-grade motherboards with 6-8 SATA slots with a few more eSATA are not uncommon.
Then again, if you only have 3TB, buying some cheaper 1-2TB drives, integrating them into your mix and eventually moving your data over, should allow you to power down a lot of your equipment especially if you're mostly using it for storage, not processing.
What we need to do is not "integrate" tech into a classroom setting, but replace the classroom setting with tech. Then we can use the classroom for group work, question/answer or discussion sessions, and demonstrations. Learning a lot of the core materials solely through reading, watching videos, and automated quizzing can get you a long way through the tedium normally found in the classroom; then you can use the face-time of the school facilities for better purposes.
But just throwing shiny toys in kids' hands is a recipe for disaster.
It's sure been a long time since I've used/seen $TITLE in online conversation.
I agree, a lot of the corporate training stuff is not that helpful; it's often too condensed into a continual barrage of nearly-opaque info. Having online courses like this is a great balance between such problems and those of a university setting. (I took the Stanford AI and ML classes)
Of course it's theory; CS is literally the science of computation. It's not about engineering.
Granted, getting good engineering out of software devs reveals a gaping education/training hole that needs to be addressed, but that is not Computer Science.
University education is about personal enrichment and beyond-the-industry research, not job training. The current western world remains confused about this, with employers who think a degree means the applicant automatically has good on-the-job capabilities (especially in computing), and students thinking a piece of very expensive paper is a magic life betterment token (correlation is not causation).
This experience is fine when you're student-age, but things change once you're a working professional adult who seeks to increase their own understanding for things they're already doing for real. Having well-presented information that's tangibly useful, instead of having to perform lots of self-learning to have it sink in, is exactly what fits the bill.
Many educators seek to create some sort of "community of education" around a student lifestyle, instead of offering direct quality knowledge to experienced customers.
The Stanford AI course had lots of technical problems that they were unprepared for - ambiguous English phrasing, uneven level of practice versus test, missing technical explanations, and so on.
But the Machine Learning class was almost perfect. Incredibly clear, precise, complete and human-friendly explanations; gave a thorough treatment of real-world caveats, testing, debugging tips; wasn't so ridiculously focused on locking in grades; and all around a way better experience than the AI class (I took both). I have nothing but respect for Prof. Ng and his staff.
There really was no excuse for the problems the AI class had in its teaching style, and those problems really didn't have anything to do with the technology.
The move "The Core" has ruined me for any article mentioning planetary cores. :-(
Like this:
[3000 pages of obtuse, meaningless, but impressive-looking technical mumbo jumbo] Therefore, there is no other way.
Just curious, what would they burn in an application like this to power its boilers? Oil? Diesel? Coal? (Lead-lined cadmium? Child laborers?)
Ah, but here we're talking about "shadow accounts", so there are no posts from Steve to associate with.
Why not? Steve might post on twitter, or wherever. He probably uses the same email address for various accounts, or reveals his full name somewhere in various other public online identities he uses. The computer doesn't *care* how spurious or inconsequential links between data are, it can aggregate them.
Also, if you think that Facebook only has internal Facebook data to work with, you're being incredibly naive. They have "value" in getting as much information about everybody from everywhere they possibly can. There have been comments posted here on Slashdot about 2 people, each without Facebook accounts, that get Facebook invites from each other. They both ended up thinking "Well, if *they* finally broke down and got an account there, maybe I should too", and only found out that the invites were unsolicited afterwards. That was all formulated from information neither has ever submitted to Facebook.
Right. So if I post "Steve still owes me that $50, don't think he's ever going to pay it back" on facebook, what's the data mining heuristic by which one of the people named Steve gets his credit rating adjusted?
If you mentioned in a prior post that you went out to dinner with some friends and it was terrible because the food was bad and Steve forgot his wallet, and Steve posts about having a good time with his friends at dinner that night, then an association can be made that the one who's not paying you back is the one who has links to you who posted that he went out to dinner that same night. Sure, it's a stretch, but the more links like these there are, the stronger the mined beliefs become. Much of this sort of association can be done with simple vocabulary filters yielding very low-probability results, but again you get a ton of these little improbabilities adding up and it becomes a probable result.
Sorry, it just doesn't work.
This is exactly how it works. Any single data point has very little information. Tons of little data points in aggregate reveals new information that cannot be inferred from looking at the points on their own. Comparing time stamps between separate information sources also reveals connections that are not expressed verbally.
I know 3 people named Steve Wright. How does Facebook know which one I'm referring to at any particular time.
That is *exactly* the problem that data mining techniques solve (or at least give reasonable guesses).
Facebook has "shadow accounts" of people it's datamined, but have not actually opened a facebook account with. So even if you don't have an account, your friends & family talking about you, or any data it harvests that contain references to you, are stored there.
I would suspect that Facebook's sharing of information with legal authorities includes shadow accounts, and if banks or credit reporting agencies strike a deal with Facebook to get at "indicator" information about people, they'd be able to view shadow accounts as well.
(sorry for the rant)
Sorry? If I hadn't already posted, I'd mod you up.
The whole original point of patents was to document inventions so that they'd get out into the public, in exchange for a temporary legal limitation on who can use it. Without them, all sorts of inventions were used internally as trade secret, and as people passed on or changed careers, innovation was lost.
The PTO is granting way too many junk patents (I haven't RTFA, so I don't know how garbage this one is), but if fewer patents were granted, or terms were shortened based on some metric of innovation or uniqueness, things would be better off. Problem is, the government believes that the number of patents granted per year is a productivity & innovation metric; in reality with such quantity granted it's just legal strong-arming and innovating stifling.
*crash*