You index video quite easily by parsing closed captioning. No speech recognition necessary.
Only for video clips with closed captioning, that were saved in a form that retained that captioning.
The flood of user-produced video clips that the digital photo flood tells me to expect certainly won't have captioning. The various university-tutorial lectures that various profs have put online don't have captioning. Similar video clips produced in-house for whatever purpose (training, demonstration, PR, etc) are unlikely to have captioning.
In summary, I do not expect this feature to be available for the majority of online video clips needing indexing.
Thats impressive for just 1 million, working in defense and knowing our contactors. 1 million dollars is bearly enough to get them to tell you how much it would cost for them to do the initial research to tell you if they can actually build what you want.
This is being done through CMU. $1 million funds a lab full of grad students, with a couple of chip spins per year, for several years.
I'm sure a few papers can be squeezed out of a speech-algorithm accelerator chip project.
- Voice controlled robots ("You missed a corner, vacuum cleaner")
- Data search by voice ("Find me a channel that plays Star Trek")
Kinda jumping ahead of yourself, aren't you? There are two steps to an operation like these, speech to text, and understanding the text you get out. Speech recognition gives you the first part, but you still have to be able to pull apart the sentence and figure out what it means.
While extracting full meaning is extremely difficult, extracting enough to get the job done when given context is much easier. Both of the tasks listed above are in the second category.
If a vacuum cleaner is listening, all it has to hear is its name (to confirm that the sentence is directed at it), and hear "corner". Depending on whether it's actively vacuuming or on standby, it would have to decide on its own whether you wanted it to re-vacuum the corners, or just sit in one.
If a search query program is listening, all it has to hear is "find" (to confirm that the sentence is directed at it), "channel", and "Star Trek". "channel" tells it to look at TV listings and radio channels. If its indexes of either pull up "Star Trek", it tells you that TV channels X, Y, and Z carry Star Trek, and prompts you for a follow-up query (e.g. "when does channel Z carry Star Trek", or "do any of the channels carry Star Trek between 6pm and 10pm?").
This kind of technology is already in use today (airline reservation systems are one of the more established applications). User-directed queries of databases and directories are one of the big emerging applications.
The only reason vacuum cleaner control isn't a big voice application yet, is that robot vacuum cleaner technology is still immature:) (though that's pretty close to being solved, with several interesting solution attempts already on the market).
In summary, these tasks are easier than you think, as long as the problem domain is properly constrained.
I work on product X and think of all the possibilities (list slightly feasible but most likely never going to happen features).
If this is really true what they're saying then people should put tons more money into product X!
Actually, use of speech recognition technology to index video clips for search engines _is_ both a very desirable technology, and something that can be done fairly easily (most professionally produced video, at least, takes great pains to have one speaker at a time and keep noise to a minimum). There's a fair bit of video content accessible via the web right now, and this will only increase (most new digital cameras can take video clips now - remember how quickly still pictures flooded the web when digicams first became available?).
Speech recognition technology has trouble when it's trying to sort out a noisy environment or a degraded communications channel, and has trouble holding useful open-ended conversations (as opposed to task-driven), but it's very capable in most other contexts. After all, the field has been under study for decades.
In summary, your mocking of the parent post is premature.
This begs the question of why they thought it was necessary to have the Hollywood stuntman ending (literally), instead of just designing a reentry vehicle with a low terminal velocity and just letting it crash into the ocean for a softer landing.
They thought that there was no way a touchdown - even with a parachute [best way of getting a low terminal velocity] - would leave the sample plates intact.
Who is going to verify their findings? What if this is all just some smoke and mirror news stories now so we all thing "yea they'll get something for the $260 million spent" only to never ever hear about it again.
Before the thing even entered the atmosphere we had JPLers saying ANY crash would destroy the experiments. Well we got 200+ mph into the earth, split open, dust everywhere, broken little bits but everything is going to be A OK.
There's a difference between "we can get useful scientific information from it" and "A-OK".
There's also a difference between dust and other contaminants deposited at low speed and what's mostly monatomic gas implanted at high speed (look up "ion implantation" in a semiconductor fabrication glossary for further discussion of this).
Processing will get a lot more difficult, as they have a bucketful of dust-contaminated shards instead of nice, organized, uncontaminated collection plates, but it's far from impossible.
Just not an option the mission team would have chosen if it could possibly be avoided.
All this sounds good and Interesting, and Becker did a tremendous ammount of development in this field. But I was just wondering, what about supercomputers like BlueGene/L which have very fast interconnects. Many supercomputers/distributed systems run MPI based programmes and such programmes need a high interprocess commmunication does anyone one know how good these are in a Bewoulf Cluster?
Anywhere from "terrible" to "almost not bad", depending on how much you're willing to pay for the interconnect network. The point of Beowulf-style clustering is low cost/node, allowing scientific computing to be done with commodity hardware (unheard-of at the time). While using something like Myrinet instead of Ethernet, and careful topology layout, can bring you to the "almost doesn't suck" stage, you'll still suffer heavily in communications-bound problems.
Fortunately, there are many interesting problems with low enough communications load to make commodity technology based clusters very, very useful.
Here's an obligatory link to Matrioshka Brains (a conceivable explanation for dark matter). If you haven't already seen this, you'll probably find it interesting.
While Matrioshka brains are indeed fascinating structures, they could only account for baryonic dark matter. The vast majority of dark matter is non-baryonic (in the astronomical sense, which differs somewhat from the particle physics sense, but in this case still means "not normal matter").
The best quality optics in the world won't get you past the diffraction limit, so the physical size of the CCD matters too. The diffraction limit imposes the condition that cramming ever more pixels into the same area eventually becomes fruitless: there's no more information to be had in that area. You want more information? Increase that area.
It's actually the size of the aperture, not the image plane, that imposes diffraction limits. On the image plane, the limit is imposed by the size of the beam waist you can get given the focal length of the lens and its aperture (which give the angle at which the beam converges, which gives the beam waist size for any given wavelength).
In the best possible case, you get a pixel size comparable to a wavelength of light (say one micron, for visible light). This requires a lens (or mirror!) with a diameter comparable to its focal length.
Typical pixel dimension is something like 5 microns. This is imposed by fabrication constraints (which change as process technologies get better) and by the fact that you have to have enough light falling on the pixel to produce a useful and low-noise signal (S:N improves as the square root of photon count, as long as circuit noise is low; more photons = less noise). There is a limit to how much light you can concentrate into a small region of a detector before damage occurs.
(For more information than you ever wanted to know about at least one type of image sensor, see our research group's web page.)
This problem can be solved easily if each ISP in the world is required to have a unique ID, and that no one can use SMTP without a unique ID issued by a trusted org.
They have unique IDs - IP addresses. You just have to trust that all servers relaying the message aren't tampering with the message. This is a problem with any ID scheme, short of strong cryptographic signatures (which have their own problems).
The best you get in practice is the ability to backtrack to the last known-trustworthy relay point. We can do this now without any change to the email system being needed.
ISP would then make sure that their customers have unique digital IDs (the customers themselves should not need to do anything).
The problem is that the ISP is _not_ _trustworthy_. This is a very large part of the current spamming problem: ISPs that don't terminate spamming and spam-related accounts.
Any proposed solution has to work despite significant numbers of untrustworthy nodes at all levels of the routing network.
You can suggest terminating troublesome ISPs, but we have the ability to do that already - the various blackhole lists are the limit of what we've had the will to implement so far. I don't see how a different infrastructure would change this.
With SMTP, it's easy to forge identities. That's something that could at least theoretically be solved cleanly with a new system.
Only if all mail servers are trustworthy. The problem is that they aren't, and won't be for the forseeable future.
And the crux of the matter; SMTP has no good mechanismism for dealing with "strangers". you either accept the email, or reject it.
How does changing mail transport help with this? I can already bin mail into "known-good" and "potential spam" folders, but I still have to either read or not read whatever's in the "potential spam" folder. Kind of hard to avoid that decision.
A new system could provide for things like pennyblack, hashcash, challenge/response, or sender risks.
Any system that tries to force the sender to invest something valuable into the mail transaction requires a unified system of email administration across the planet. We don't have this, and aren't likely to any time soon. Most implementations I've heard about also require that mail servers be trustworthy, which we don't have, and won't have any time soon.
You also have the problem of any resource investment high enough to be a deterrent being high enough that users will switch to an alternate transport system, or just keep SMTP, rather than pay it. Zombies will also send mail just fine on the user's tab.
So, I have doubts about these proposals being usefully workable.
If each email is signed to be from a certified source (e.g. accountability is added) then total strangers can still email you.
Part of my point is that the signing sources won't be trustworthy, making signing valueless.
You can reject signatories that you don't trust, but you can reject mail servers you don't trust right now. In both cases, you end up throwing out legitimate mail with the spam.
A large part of the problem now is that there are ISPs who through apathy or intent harbour spammers. If we can't apply enough pressure to yank the accounts or de-peer these ISPs under the current system, what leads you to believe that yanking the accounts under a signing system or blacklisting these ISPs as signatories will be any easier?
That is _much_ better than today, were said spammer can blast millions of messages each day out - as fast as the network pipe allows and nothing ever happens to him.
I fail to see how signing changes this. We already have a record of where mail comes from - the header, as far along the chain as we trust the servers listed. Signing fails for the same reason headers fail - as soon as there's a step in the chain that we don't trust, we're sunk.
IMHO completely dropping email as we have it now is the only way against spam.
The problems is that any system with the features we demand of email has the faults of email.
The crux of it is - do you want someone you haven't heard of before to be able to email you?
If the answer is "yes", then you get spam.
If the answer is "no", you get something fundamentally different from email. You can also already implement this, by using a whitelist for both email addresses and originating mail servers (to filter forged friends' addresses).
Authenticating users and rubber-stamping their mail at mailservers doesn't help, because there are always untrustworthy mailservers run by ISPs who don't know enough or don't care enough to fix them. This is half of the source of the _current_ spamming problem. So, any decentralized email-like system is vulnerable to having spamming users and compromised mail servers exist. Compromised mail servers bring back forging, and you're pretty much back to square one. It gets a little harder to convincingly forge a sender address from a different mail server, but you can _already_ filter for that by using a server whitelist or using a DNS lookup (forward or reverse) for server lines in inbound mail.
Having a centralized mail server makes it harder to insert bogus traffic, but creates a huge bandwidth bottleneck, and concentrates power over mail in a way that's unlikely to be acceptable.
In just about any scheme, you can also get compromised user machines spewing mail from their own accounts with legitimate sign-in to any type of mail system at all.
In summary, the spam problem isn't going away under any system that serves the same purpose as email. You can also modify a standard email system to get most of the benefits of the different types of system that _would_ be more spam-resistant. So, there doesn't seem to be much point in proposing a system-wide overhaul.
You can't force learning. At best, you can present information that is quickly discarded. Why is, say, history required? I've *never* needed to have history. I don't know anyone that ever needed to know what year or on what boats Colombus sailed. Yet I imagine that this is on a test at least once for every American student.
You don't remember what year he sailed, but you remember that he _did_ sail. You also picked up quite a bit of context about the history of the US that you almost certainly use without even thinking about it, every time you have to identify what the source of an issue you read about in the paper is.
Without a knowledge of history, however superficial, there is no concept of where your culture came from and _why_.
You don't want to ask that of me. All the ones I didn't care about are the ones I don't use. What good is, say, English Literature (required at my school) for a network administrator?
It lets you know that there _is_ more to be learned about the topic, and gives you a few directions to go if you decide to look for it. You may not follow up on these leads, but others will.
Topics like history and literature are there to establish a minimal overview of topics, so that very fundamental concepts are known, and people know where to learn more if they want to or need to. It's very sobering to sit down and think about what kind of person would result if they had _no_ knowledge of just about any specific discipline students are exposed to in high school.
Sorry. You are wrong. You are viewing a child that has already been tainted by home life and more importantly the school system.
I keep seeing this statement repeated, but it never sounds plausible.
You are making the implicit assumption that kids _are_ interested in _everything_. Work with a few kids. Preferably, work with a _lot_ of kids. Kids are, by and large, self-centered little munchkins who optimize their lives for personal pleasure (arguably the same is true for most adults). To most kids, math means nothing, and is stuff their teacher has to make them study. Ditto literature in high school. Ditto history. The ones who aren't interested in math for its own sake won't care about applying math to personal finances until years after they'd have to have learned it, and the ones who aren't interested in literature or history for its own sake are unlikely to understand why having it as cultural context is important for understanding the world at large.
Kids left to their own devices play. Occasionally, they'll study a handful of subjects that interest them until they get bored. Very occasionally, they won't get bored, and from there you get the computer hobbyists and artists and literature hobbyists and so forth. These are _rare_. Most kids just want to have mostly-non-educational fun.
Visit a laboratory school at any university and see whats up. Children WANT to know.
Bollocks. The kids in those schools are there (and stay there) because they _wanted_ to learn. You have a very selective pre-filter. It's the _rest_ that are the problem.
You have to package it into something they will enjoy.
Package trig in a way that all high-school students will enjoy and be motivated to learn more about. Go ahead.
I keep seeing things like _this_ proposed to. The fact remains that you can't _make_ the students care. The best you can do is encourage the ones who do, and force the rest to study enough to pass their tests. Of the latter, some may retain something, and most will at least remember the subject exists.
If you've worked with little kids, one of the first things you notice is that almost every single one of them really, really wants to learn.
But somehow, during about K-4th grade, most of the kids in the US educational system seem to have that crushed out of them.
I work with kids anywhere from 4 years old to 15 on a regular basis. Kids are curious - yes. Kids want to learn what's important for them to learn - no. They want to learn about what they think is cool.
Think back to your high school days. How many of the courses did you take that you actually cared about? Given the option, would you have been in that school, or been outside playing or at home playing computer games? How many of these courses that you didn't care about then, are you glad you took now?
The whole premise behind the school system is that there are things kids Need To Know, and they're going to learn them whether they care about them at the time or not. Every time I hear someone suggest that kids should only learn what they're interested in I shake my head. It's only _after_ you need it that you realize what you needed to know, and very few kids have "planning for the future" as a priority at all.
In summary, your observations are adequately explained by kids not being interested in complex subjects they don't care about, not by their desire to learn being "crushed" by some oppressive authority.
Re:"affecting literally millions of people."
on
Make Money Fast
·
· Score: 1
I have to ask - Do you live in a crack house or something? [or have landlords that fly under the government's radar] Where in the world do landlords accept cash as a payment for rent?
Anywhere in this city, as far as I can tell. You see cheques too, but cash is popular for two reasons:
Cash doesn't bounce.
Sub-letting.
It's pretty common for friends (at least in my age bracket) to split rent on an apartment (fact of life; rent is high, students and former students paying off student loans are poor). Scraping together rent in the form of cash is a lot easier than dealing with cheques in situations like this.
If you have steady income and enough of a cash cushion in your bank account to rent your own place and pay by cheque, good for you. There are a very large number of people who struggle to make ends meet who don't.
No matter what the something-for-nothing kiddies here will say, every time a song is illegally copied, the downloader is able to hear a work they would have otherwise paid for. Barring a random play on radio, you personally have to pay to listen to most copyrighted works.
The CDs I didn't buy as a result of hearing on the radio or in a movie theatre, I bought because I heard them played by friends and liked them.
Very, very rarely do I buy a CD without knowing that I like what's on it beforehand. The money spent on a lemon could be so much better spent elsewhere - it's not worth the risk. This, combined with the ease of listening beforehand from legitimate sources (do none of you have friends with CD collections?) makes me skeptical of your argument.
As the sonoluminescence experiements and the fusors show, extreme temperatures are NOT a prereq for fusion, just for ultra-high energy fusion
Bad examples, I'm afraid:). Both of these are conventional high-temperature plasma fusion (just with unusual confinement schemes). Neither come anywhere close to breakeven (or significant power production period), instead being used as convenient laboratory sources for x-rays and neutrons.
P&F and other cold fusion advocates claimed that they _did_ produce substantial amounts of fusion energy without a plausible mechanism (density of deuterium in the electrodes was still far too low for tunnelling to produce a measurable rate of fusion).
Re:"affecting literally millions of people."
on
Make Money Fast
·
· Score: 1
How many $100 or even $50 bills do you see on a daily basis coming out of people's wallets? (or even flashed from within one?)
I see them on a monthly basis - that's how most people I know pay rent.
I didn't think banks gave _out_ the 1000-dollar bills. Live and learn:).
Canada has become so hooked on Interac that cash is almost a nono..
Not in Toronto, that I've seen. Cash is just rarely used for _big_ things, as carrying that much around is a big hit if it gets lost or stolen (whereas I wouldn't be too upset if I was mugged for $20-$30). Cash for buying lunch, and credit or debit for buying anything more substantial.
They'd cry "But where are all the neutrons", or "You'd be dead by now if that much excess heat were actually being produced." What most of these so-called entrenched scientists failed to realize was, this was something entirely new.
What you appear to be overlooking is the fact that they _should_ be dead by now if nuclear processes were generating the measured heat. I have a far easier time believing that electrochemical energy storage was happening with an input that wasn't being adequately measured, or that some other mode of chemical energy production was occurring, than believing that a century of nuclear physics has to be outright repealed (as opposed to just extended). Fusion produces excited nuclei. Excited nuclei shed energy in ways that are a) very measurable and b) unhealthy for bystanders. The bystanders are alive, and the neutrons and gamma rays are missing. That rules out fusion.
Slashdot Mirror
You index video quite easily by parsing closed captioning. No speech recognition necessary.
Only for video clips with closed captioning, that were saved in a form that retained that captioning.
The flood of user-produced video clips that the digital photo flood tells me to expect certainly won't have captioning. The various university-tutorial lectures that various profs have put online don't have captioning. Similar video clips produced in-house for whatever purpose (training, demonstration, PR, etc) are unlikely to have captioning.
In summary, I do not expect this feature to be available for the majority of online video clips needing indexing.
Thats impressive for just 1 million, working in defense and knowing our contactors. 1 million dollars is bearly enough to get them to tell you how much it would cost for them to do the initial research to tell you if they can actually build what you want.
This is being done through CMU. $1 million funds a lab full of grad students, with a couple of chip spins per year, for several years.
I'm sure a few papers can be squeezed out of a speech-algorithm accelerator chip project.
- Voice controlled robots ("You missed a corner, vacuum cleaner")
:) (though that's pretty close to being solved, with several interesting solution attempts already on the market).
- Data search by voice ("Find me a channel that plays Star Trek")
Kinda jumping ahead of yourself, aren't you? There are two steps to an operation like these, speech to text, and understanding the text you get out. Speech recognition gives you the first part, but you still have to be able to pull apart the sentence and figure out what it means.
While extracting full meaning is extremely difficult, extracting enough to get the job done when given context is much easier. Both of the tasks listed above are in the second category.
If a vacuum cleaner is listening, all it has to hear is its name (to confirm that the sentence is directed at it), and hear "corner". Depending on whether it's actively vacuuming or on standby, it would have to decide on its own whether you wanted it to re-vacuum the corners, or just sit in one.
If a search query program is listening, all it has to hear is "find" (to confirm that the sentence is directed at it), "channel", and "Star Trek". "channel" tells it to look at TV listings and radio channels. If its indexes of either pull up "Star Trek", it tells you that TV channels X, Y, and Z carry Star Trek, and prompts you for a follow-up query (e.g. "when does channel Z carry Star Trek", or "do any of the channels carry Star Trek between 6pm and 10pm?").
This kind of technology is already in use today (airline reservation systems are one of the more established applications). User-directed queries of databases and directories are one of the big emerging applications.
The only reason vacuum cleaner control isn't a big voice application yet, is that robot vacuum cleaner technology is still immature
In summary, these tasks are easier than you think, as long as the problem domain is properly constrained.
I work on product X and think of all the possibilities (list slightly feasible but most likely never going to happen features).
If this is really true what they're saying then people should put tons more money into product X!
Actually, use of speech recognition technology to index video clips for search engines _is_ both a very desirable technology, and something that can be done fairly easily (most professionally produced video, at least, takes great pains to have one speaker at a time and keep noise to a minimum). There's a fair bit of video content accessible via the web right now, and this will only increase (most new digital cameras can take video clips now - remember how quickly still pictures flooded the web when digicams first became available?).
Speech recognition technology has trouble when it's trying to sort out a noisy environment or a degraded communications channel, and has trouble holding useful open-ended conversations (as opposed to task-driven), but it's very capable in most other contexts. After all, the field has been under study for decades.
In summary, your mocking of the parent post is premature.
This begs the question of why they thought it was necessary to have the Hollywood stuntman ending (literally), instead of just designing a reentry vehicle with a low terminal velocity and just letting it crash into the ocean for a softer landing.
They thought that there was no way a touchdown - even with a parachute [best way of getting a low terminal velocity] - would leave the sample plates intact.
They were right.
Who is going to verify their findings? What if this is all just some smoke and mirror news stories now so we all thing "yea they'll get something for the $260 million spent" only to never ever hear about it again.
Before the thing even entered the atmosphere we had JPLers saying ANY crash would destroy the experiments. Well we got 200+ mph into the earth, split open, dust everywhere, broken little bits but everything is going to be A OK.
There's a difference between "we can get useful scientific information from it" and "A-OK".
There's also a difference between dust and other contaminants deposited at low speed and what's mostly monatomic gas implanted at high speed (look up "ion implantation" in a semiconductor fabrication glossary for further discussion of this).
Processing will get a lot more difficult, as they have a bucketful of dust-contaminated shards instead of nice, organized, uncontaminated collection plates, but it's far from impossible.
Just not an option the mission team would have chosen if it could possibly be avoided.
All this sounds good and Interesting, and Becker did a tremendous ammount of development in this field. But I was just wondering, what about supercomputers like BlueGene/L which have very fast interconnects. Many supercomputers/distributed systems run MPI based programmes and such programmes need a high interprocess commmunication does anyone one know how good these are in a Bewoulf Cluster?
Anywhere from "terrible" to "almost not bad", depending on how much you're willing to pay for the interconnect network. The point of Beowulf-style clustering is low cost/node, allowing scientific computing to be done with commodity hardware (unheard-of at the time). While using something like Myrinet instead of Ethernet, and careful topology layout, can bring you to the "almost doesn't suck" stage, you'll still suffer heavily in communications-bound problems.
Fortunately, there are many interesting problems with low enough communications load to make commodity technology based clusters very, very useful.
Here's an obligatory link to Matrioshka Brains (a conceivable explanation for dark matter). If you haven't already seen this, you'll probably find it interesting.
While Matrioshka brains are indeed fascinating structures, they could only account for baryonic dark matter. The vast majority of dark matter is non-baryonic (in the astronomical sense, which differs somewhat from the particle physics sense, but in this case still means "not normal matter").
And since this gives a minimum useful pixel size, the only way to get more (useful) pixels is to have a bigger physical CCD. Just like was said.
For reasons completely other than what he said. Hence, the reply.
The best quality optics in the world won't get you past the diffraction limit, so the physical size of the CCD matters too. The diffraction limit imposes the condition that cramming ever more pixels into the same area eventually becomes fruitless: there's no more information to be had in that area. You want more information? Increase that area.
It's actually the size of the aperture, not the image plane, that imposes diffraction limits. On the image plane, the limit is imposed by the size of the beam waist you can get given the focal length of the lens and its aperture (which give the angle at which the beam converges, which gives the beam waist size for any given wavelength).
In the best possible case, you get a pixel size comparable to a wavelength of light (say one micron, for visible light). This requires a lens (or mirror!) with a diameter comparable to its focal length.
Typical pixel dimension is something like 5 microns. This is imposed by fabrication constraints (which change as process technologies get better) and by the fact that you have to have enough light falling on the pixel to produce a useful and low-noise signal (S:N improves as the square root of photon count, as long as circuit noise is low; more photons = less noise). There is a limit to how much light you can concentrate into a small region of a detector before damage occurs.
(For more information than you ever wanted to know about at least one type of image sensor, see our research group's web page.)
This problem can be solved easily if each ISP in the world is required to have a unique ID, and that no one can use SMTP without a unique ID issued by a trusted org.
They have unique IDs - IP addresses. You just have to trust that all servers relaying the message aren't tampering with the message. This is a problem with any ID scheme, short of strong cryptographic signatures (which have their own problems).
The best you get in practice is the ability to backtrack to the last known-trustworthy relay point. We can do this now without any change to the email system being needed.
ISP would then make sure that their customers have unique digital IDs (the customers themselves should not need to do anything).
The problem is that the ISP is _not_ _trustworthy_. This is a very large part of the current spamming problem: ISPs that don't terminate spamming and spam-related accounts.
Any proposed solution has to work despite significant numbers of untrustworthy nodes at all levels of the routing network.
You can suggest terminating troublesome ISPs, but we have the ability to do that already - the various blackhole lists are the limit of what we've had the will to implement so far. I don't see how a different infrastructure would change this.
With SMTP, it's easy to forge identities.
That's something that could at least theoretically be solved cleanly with a new system.
Only if all mail servers are trustworthy. The problem is that they aren't, and won't be for the forseeable future.
And the crux of the matter;
SMTP has no good mechanismism for dealing with "strangers". you either accept the email, or reject it.
How does changing mail transport help with this? I can already bin mail into "known-good" and "potential spam" folders, but I still have to either read or not read whatever's in the "potential spam" folder. Kind of hard to avoid that decision.
A new system could provide for things like pennyblack, hashcash, challenge/response, or sender risks.
Any system that tries to force the sender to invest something valuable into the mail transaction requires a unified system of email administration across the planet. We don't have this, and aren't likely to any time soon. Most implementations I've heard about also require that mail servers be trustworthy, which we don't have, and won't have any time soon.
You also have the problem of any resource investment high enough to be a deterrent being high enough that users will switch to an alternate transport system, or just keep SMTP, rather than pay it. Zombies will also send mail just fine on the user's tab.
So, I have doubts about these proposals being usefully workable.
If each email is signed to be from a certified source (e.g. accountability is added) then total strangers can still email you.
Part of my point is that the signing sources won't be trustworthy, making signing valueless.
You can reject signatories that you don't trust, but you can reject mail servers you don't trust right now. In both cases, you end up throwing out legitimate mail with the spam.
A large part of the problem now is that there are ISPs who through apathy or intent harbour spammers. If we can't apply enough pressure to yank the accounts or de-peer these ISPs under the current system, what leads you to believe that yanking the accounts under a signing system or blacklisting these ISPs as signatories will be any easier?
That is _much_ better than today, were said spammer can blast millions of messages each day out - as fast as the network pipe allows and nothing ever happens to him.
I fail to see how signing changes this. We already have a record of where mail comes from - the header, as far along the chain as we trust the servers listed. Signing fails for the same reason headers fail - as soon as there's a step in the chain that we don't trust, we're sunk.
IMHO completely dropping email as we have it now is the only way against spam.
The problems is that any system with the features we demand of email has the faults of email.
The crux of it is - do you want someone you haven't heard of before to be able to email you?
If the answer is "yes", then you get spam.
If the answer is "no", you get something fundamentally different from email. You can also already implement this, by using a whitelist for both email addresses and originating mail servers (to filter forged friends' addresses).
Authenticating users and rubber-stamping their mail at mailservers doesn't help, because there are always untrustworthy mailservers run by ISPs who don't know enough or don't care enough to fix them. This is half of the source of the _current_ spamming problem. So, any decentralized email-like system is vulnerable to having spamming users and compromised mail servers exist. Compromised mail servers bring back forging, and you're pretty much back to square one. It gets a little harder to convincingly forge a sender address from a different mail server, but you can _already_ filter for that by using a server whitelist or using a DNS lookup (forward or reverse) for server lines in inbound mail.
Having a centralized mail server makes it harder to insert bogus traffic, but creates a huge bandwidth bottleneck, and concentrates power over mail in a way that's unlikely to be acceptable.
In just about any scheme, you can also get compromised user machines spewing mail from their own accounts with legitimate sign-in to any type of mail system at all.
In summary, the spam problem isn't going away under any system that serves the same purpose as email. You can also modify a standard email system to get most of the benefits of the different types of system that _would_ be more spam-resistant. So, there doesn't seem to be much point in proposing a system-wide overhaul.
I can't wait until plants can do the job of making oxygen instead of relying on machines.
Why? Machines do it more efficiently and with far less mass overhead.
None.
If it hadn't been for music & speech/debate, the whole fucking twelve years would have been a near total loss.
So you neither read nor use simple arithmetic? It's interesting that you're able to post here, then.
Most kids don't like learning these subjects, despite needing them fairly shortly afterwards.
You can't force learning. At best, you can present information that is quickly discarded. Why is, say, history required? I've *never* needed to have history. I don't know anyone that ever needed to know what year or on what boats Colombus sailed. Yet I imagine that this is on a test at least once for every American student.
You don't remember what year he sailed, but you remember that he _did_ sail. You also picked up quite a bit of context about the history of the US that you almost certainly use without even thinking about it, every time you have to identify what the source of an issue you read about in the paper is.
Without a knowledge of history, however superficial, there is no concept of where your culture came from and _why_.
You don't want to ask that of me. All the ones I didn't care about are the ones I don't use. What good is, say, English Literature (required at my school) for a network administrator?
It lets you know that there _is_ more to be learned about the topic, and gives you a few directions to go if you decide to look for it. You may not follow up on these leads, but others will.
Topics like history and literature are there to establish a minimal overview of topics, so that very fundamental concepts are known, and people know where to learn more if they want to or need to. It's very sobering to sit down and think about what kind of person would result if they had _no_ knowledge of just about any specific discipline students are exposed to in high school.
Sorry. You are wrong. You are viewing a child that has already been tainted by home life and more importantly the school system.
I keep seeing this statement repeated, but it never sounds plausible.
You are making the implicit assumption that kids _are_ interested in _everything_. Work with a few kids. Preferably, work with a _lot_ of kids. Kids are, by and large, self-centered little munchkins who optimize their lives for personal pleasure (arguably the same is true for most adults). To most kids, math means nothing, and is stuff their teacher has to make them study. Ditto literature in high school. Ditto history. The ones who aren't interested in math for its own sake won't care about applying math to personal finances until years after they'd have to have learned it, and the ones who aren't interested in literature or history for its own sake are unlikely to understand why having it as cultural context is important for understanding the world at large.
Kids left to their own devices play. Occasionally, they'll study a handful of subjects that interest them until they get bored. Very occasionally, they won't get bored, and from there you get the computer hobbyists and artists and literature hobbyists and so forth. These are _rare_. Most kids just want to have mostly-non-educational fun.
Visit a laboratory school at any university and see whats up. Children WANT to know.
Bollocks. The kids in those schools are there (and stay there) because they _wanted_ to learn. You have a very selective pre-filter. It's the _rest_ that are the problem.
You have to package it into something they will enjoy.
Package trig in a way that all high-school students will enjoy and be motivated to learn more about. Go ahead.
I keep seeing things like _this_ proposed to. The fact remains that you can't _make_ the students care. The best you can do is encourage the ones who do, and force the rest to study enough to pass their tests. Of the latter, some may retain something, and most will at least remember the subject exists.
If you've worked with little kids, one of the first things you notice is that almost every single one of them really, really wants to learn.
But somehow, during about K-4th grade, most of the kids in the US educational system seem to have that crushed out of them.
I work with kids anywhere from 4 years old to 15 on a regular basis. Kids are curious - yes. Kids want to learn what's important for them to learn - no. They want to learn about what they think is cool.
Think back to your high school days. How many of the courses did you take that you actually cared about? Given the option, would you have been in that school, or been outside playing or at home playing computer games? How many of these courses that you didn't care about then, are you glad you took now?
The whole premise behind the school system is that there are things kids Need To Know, and they're going to learn them whether they care about them at the time or not. Every time I hear someone suggest that kids should only learn what they're interested in I shake my head. It's only _after_ you need it that you realize what you needed to know, and very few kids have "planning for the future" as a priority at all.
In summary, your observations are adequately explained by kids not being interested in complex subjects they don't care about, not by their desire to learn being "crushed" by some oppressive authority.
Anywhere in this city, as far as I can tell. You see cheques too, but cash is popular for two reasons:
It's pretty common for friends (at least in my age bracket) to split rent on an apartment (fact of life; rent is high, students and former students paying off student loans are poor). Scraping together rent in the form of cash is a lot easier than dealing with cheques in situations like this.
If you have steady income and enough of a cash cushion in your bank account to rent your own place and pay by cheque, good for you. There are a very large number of people who struggle to make ends meet who don't.
No matter what the something-for-nothing kiddies here will say, every time a song is illegally copied, the downloader is able to hear a work they would have otherwise paid for. Barring a random play on radio, you personally have to pay to listen to most copyrighted works.
The CDs I didn't buy as a result of hearing on the radio or in a movie theatre, I bought because I heard them played by friends and liked them.
Very, very rarely do I buy a CD without knowing that I like what's on it beforehand. The money spent on a lemon could be so much better spent elsewhere - it's not worth the risk. This, combined with the ease of listening beforehand from legitimate sources (do none of you have friends with CD collections?) makes me skeptical of your argument.
As the sonoluminescence experiements and the fusors show, extreme temperatures are NOT a prereq for fusion, just for ultra-high energy fusion
:). Both of these are conventional high-temperature plasma fusion (just with unusual confinement schemes). Neither come anywhere close to breakeven (or significant power production period), instead being used as convenient laboratory sources for x-rays and neutrons.
Bad examples, I'm afraid
P&F and other cold fusion advocates claimed that they _did_ produce substantial amounts of fusion energy without a plausible mechanism (density of deuterium in the electrodes was still far too low for tunnelling to produce a measurable rate of fusion).
How many $100 or even $50 bills do you see on a daily basis coming out of people's wallets? (or even flashed from within one?)
:).
I see them on a monthly basis - that's how most people I know pay rent.
I didn't think banks gave _out_ the 1000-dollar bills. Live and learn
Canada has become so hooked on Interac that cash is almost a nono..
Not in Toronto, that I've seen. Cash is just rarely used for _big_ things, as carrying that much around is a big hit if it gets lost or stolen (whereas I wouldn't be too upset if I was mugged for $20-$30). Cash for buying lunch, and credit or debit for buying anything more substantial.
Your mileage may vary.
and tell me again how 1 in 290 is rare.
0.34_%_ - not even close to being "pennies on the dollar". Seems rare enough to me.
They'd cry "But where are all the neutrons", or "You'd be dead by now if that much excess heat were actually being produced." What most of these so-called entrenched scientists failed to realize was, this was something entirely new.
What you appear to be overlooking is the fact that they _should_ be dead by now if nuclear processes were generating the measured heat. I have a far easier time believing that electrochemical energy storage was happening with an input that wasn't being adequately measured, or that some other mode of chemical energy production was occurring, than believing that a century of nuclear physics has to be outright repealed (as opposed to just extended). Fusion produces excited nuclei. Excited nuclei shed energy in ways that are a) very measurable and b) unhealthy for bystanders. The bystanders are alive, and the neutrons and gamma rays are missing. That rules out fusion.