In a research family tree point of view, he's behind this too. The MIT algorithm is a modification of RRT that instead of using one-connectedness uses k-connectness like PRM. RRT is from LaValle and Kuffner, who were Latoumbe's students at Stanford. PRM is from Lydia Kavraki, who was also Latoumbe's student at Stanford.
The PR2 video was kind or irksome, because they chose the very first version of RRT (circa 2000) to compare against their new algorithm, rather than a number of variants that became available within a few years (For example ERRT from 2002, or multi-bride RRT from 2006, which are my small contributions to the space). In particular D Ferguson's DRRT from 2006 which would replan iteratively with a cost constraint would have been a natural baseline to use.
If they had used any of those later variants, in particular any of the ones that implemented path smoothing (considered an absolute necessity since the beginning), you probably could not tell the difference in the results. IOW it was a strawman just like comparing MyNewSort to bubblesort instead of quicksort.
It is still a contribution though. I saw a very similar algorithm proposed by some Taiwanese researchers (forget the paper) that was also a graph-based extension of RRT, but Karaman and Frazzoli have gone through and done the proofs to verify reasonable properties for their system (ideas are easier than analysis). However rather than the article's statement:
"By combining two innovative algorithms developed at MIT..." it might be more accurate to say:
"Building on Stanford's RRT and PRM, MIT researchers have combined them in a new algorithm yielding the best properties of both. It can generate efficient paths without excessive memory use or post-processing."
I'm fine with Google developing their own language, but I prefer Facebook's style. Reaching the limitations of php, they built their own compiler (even though there already was a decent compiler), where as google has the scrap it all and start from scratch.
Now you'll probably say "but that's not a tablet!". I would reply that we're talking about design here, and confusingly similar looks, not operation. In fact the behavior of the device is explicitly called out as not being part of the design patent Apple filed [1].
Furthermore if how you use something matters, rather than trade dress, we can just call the Galaxy Tab a "digital picture frame and music player", and close the case in Samsung's favor.
[1] Patent title: "Electronic Device" Primary Claim: "We claim the ornamental design for an electronic device, substantially as shown and described." Label for only picture showing someone using the device: "FIG. 9 is an exemplary diagram of the use of the electronic device thereof the broken lines being shown for illustrative puposes only and form no part of the claimed design. " IOW, there is nothing whatsoever in this patent that is inconsistent with a digital picture frame and music player, and they are not claiming anything in the patent that is specific to a tablet. http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.htm&r=1&f=G&l=50&s1=D504,889.PN.&OS=PN/D504,889&RS=PN/D504,889
Of course according to you Google wouldn't do this since it is their core business, right?
DNT is a nuclear weapon for honest companies, and a joke to be ignored for dishonest ones (it's like the famous Evil Bit). What is really needed is European-style privacy regulations (i.e. with actual teeth for bad actors), and a fine-grained permissions model on Websites (like iOS or Android has for apps).
Form follows function, which is why every TV and computing device is destined to look the same once a level of miniaturization is reached. That's why you see similar tablets in a 1970s TV show ("The Tomorrow People") and a movie from the 1960s ("2001 a Space Odyssey").
Apple certainly is a style trendsetter, but really it is more about bringing things to market that are the closest to what the visionaries have already described. There's a rather clear evolution from other mp3 players through to the ipod, iphone, and then ipad. And yes, along the way things came from non-Apple sources too.. the next iphone will have screen dimensions suspiciously like an HTC EVO, and a notification bar straight from Android. That's what happens in competition.
I'm sorry this conflicts with your worldview that all these nice Apple products were invented in a vacuum.
When I read the summary, I thought exactly that -- I wonder if this is another planar walker?
Planar walkers were already doing flips 20 years ago, and walking quite convincingly and naturally 12 years ago. Many of those researchers have moved to 3D now, with some of the most impressive recent work showing up at Boston Dynamics.
There still is good research that can be done on semi-passive force control and efficiency, and a lot of that can be tested on planar walkers. However, for the love of all that is honest, at least mention in the summary that all the things you are claiming are done in a 3 d.o.f. workspace rather than the 6 d.o.f. you have in the real world. A bar also offers a metric load of damping, which is the thing keeping the full-3D walkers from really working naturally, and thus that kind of simplification should be mentioned.
Normal users often search for vague terms, and don't understand things like operators, so the default has to be to support that majority of users. Power users can learn to use quoting, plus, and other operators, so it is up to you to learn those features and employ them.
Last I heard google has not commented on the accident.
It was in the first article on any news site:
A Google spokesperson gave us this quote about the accident: "Safety is our top priority. One of our goals is to prevent fender-benders like this one, which occurred while a person was manually driving the car."
This is slashdot, you're supposed to make overly broad generalizations from minimal data. Why relate your story when you can use a position you once had to make a statement about the entire industry?:)
It's probably too late to get modded up so you can see this, but:
The technical term you are looking for is "coverage path planning", and there are well known algorithms to solve it efficiently such as Choset's Boustrophedon. I don't think any of them are optimal for a nontrivial shape, but they will probably beat most human heuristics.
In addition to lawns, this is important for machining (material removal), de-mining (completeness is essential, but overlap is expensive), and large floor cleaning.
I don't really have a strong opinion on the subject, but what you claim and what I've seen while traveling abroad don't seem to agree. I did some digging about excessive fees and found the following:
There are many ways of calculating prevailing-party attorney fees. Most courts recognize that actual costs may be disproportionate and inequitable. Thus, many jurisdictions rely on other calculations. Many courts or laws invoke a lodestar' calculation: reasonably expected billable hours multiplied by a reasonable hourly rate, sometimes multiplied by a factor reflecting the risk or complexity of the case.
An old but fairly well written article on the subject: http://reason.com/archives/1995/06/01/civil-suits/1 (I realize many will not trust this source, but the article pretty much just states what the situation is in European countries.)
A common fear about loser-pays is that the side who loses a routine dispute will get handed a bill for 10,000 hours from Cravath, Swaine & Moore. But European courts are well aware of the danger that successful litigants will overinvest in their cases and gold-plate their fee requests. They carefully control the process to prevent that danger, giving the losing side a full chance to dispute a fee award, requiring that work be reasonable and necessary, providing that elite lawyer rates not be paid if a Main Street lawyer could have done the job, and so forth.
Although fee awards are usually set high enough to apply serious incentive pressure, it is notable that no country appears to let winners recoup all the money they have spent on a suit. Some leave a portion of lawyers' hours unreimbursed, while others apply hourly rates that fall below prevailing levels. Some shift expert witness fees, but others do not.
Here's an interesting nugget from later in the article. Unsurprisingly, lawyers are the most outspoken critics of loser-pays, yet their clients don't seem to agree:
The American public itself seems not to be as terrified of full loser-pays as its lawyers. Even when U.S. News and World Report asked the question in its more menacing form--"If you sue someone and lose the case, should you pay his costs?"--a far-from-shabby 44 percent of those polled agreed that they should. And with sides reversed, nearly everyone saw the idea's fairness. "If someone sues you and you win the case, should he pay your legal costs?" Eighty-five percent said yes to that one.
Can you stop posting repeatedly about how 600GB is big and hard to miss? On your home machine, yes you'd notice it. However, compared to the size of four years worth of high-res panoramas taken every few meters on a significant fraction of the world's developed roads? In that context, 600GB is quite literally nothing. When a car dumps 1TB of photos, you're not necessarily going to notice that an adjoining tar file of text logs is a couple MB too large. When you store that on 10000 machines, you're not going to notice that each one is using 60MB of hard drive space more than you expected.
Are you running apache anywhere? Can you recite the exact settings and the log retention time? Might there be extra switches you left on and forgot about?
You should have tried 20% TaskForce projects, they work great. You bank up a quarter's worth of 20% time (1-2 weeks) then use it all at once to work in small groups at a shared location. My new team started up this spring, and was originally based on a 20% TaskForce from last fall.
So no, 20% time is not dead, but you do need to get off your ass to use it. Just like vacation time, nobody can force you to use it, and if you don't make use of it you lose it.
Of course, 20% time to work on something completely random with no possible purpose for the company doesn't exist, but it never existed anyway. That's what you call "projects in your spare time" and it works much the way it does in other semi-flexible companies. I'm glad the company lets me use equipment and resources (within reason) for random projects like that, but expecting them to pay me for that time is unreasonable.
Slashdot Mirror
It was the "Helium Privatization Act of 1996", which would have been passed by the Republican congress at the time and signed by Clinton:
http://en.wikipedia.org/wiki/National_Helium_Reserve
It usually only takes a moment to verify something you "just heard", and it really helps to avoid the echo box effect.
50% of all Bitcoins in existence are owned by 0.1% of users.
I don't have any bitcoin, but in the face of such computational greed, I will be heard.
WE ARE THE 99.9999%.
oops...
s/multi-bride RRT/multi-bridge RRT/
Latoumbe at Stanford was behind a lot of that.
In a research family tree point of view, he's behind this too. The MIT algorithm is a modification of RRT that instead of using one-connectedness uses k-connectness like PRM. RRT is from LaValle and Kuffner, who were Latoumbe's students at Stanford. PRM is from Lydia Kavraki, who was also Latoumbe's student at Stanford.
The PR2 video was kind or irksome, because they chose the very first version of RRT (circa 2000) to compare against their new algorithm, rather than a number of variants that became available within a few years (For example ERRT from 2002, or multi-bride RRT from 2006, which are my small contributions to the space). In particular D Ferguson's DRRT from 2006 which would replan iteratively with a cost constraint would have been a natural baseline to use.
If they had used any of those later variants, in particular any of the ones that implemented path smoothing (considered an absolute necessity since the beginning), you probably could not tell the difference in the results. IOW it was a strawman just like comparing MyNewSort to bubblesort instead of quicksort.
It is still a contribution though. I saw a very similar algorithm proposed by some Taiwanese researchers (forget the paper) that was also a graph-based extension of RRT, but Karaman and Frazzoli have gone through and done the proofs to verify reasonable properties for their system (ideas are easier than analysis). However rather than the article's statement:
"By combining two innovative algorithms developed at MIT..."
it might be more accurate to say:
"Building on Stanford's RRT and PRM, MIT researchers have combined them in a new algorithm yielding the best properties of both. It can generate efficient paths without excessive memory use or post-processing."
Anyone can breed a rat. But only the developers can create or leave a bug in their own software.
I love this quote. I think it gets better without context.
"...the patent is assigned to Stanford University and not to Google."
http://en.wikipedia.org/wiki/PageRank
So you found a device that match the first claim of the Apple patent.
Thanks. In fact it matches the only claim, because it is a design patent and there is only one claim.
Great. Now you'll need to find a device that match ALL claims of Apple's patent.
Ok. All (one) claim(s) are matched by my example. You've also unambiguously revealed that you didn't bother to read the patent.
That's the Galaxy S and Tab.
...and the Samsung digital picture frame, most modern TVs, and a lot of other things.
Acid is quite finicky but that is indeed quite strange. I'll attempt to report an issue.
I'm fine with Google developing their own language, but I prefer Facebook's style. Reaching the limitations of php, they built their own compiler (even though there already was a decent compiler), where as google has the scrap it all and start from scratch.
Um, V8?
Digital picture frame from 2006:
http://www.engadget.com/2006/03/09/samsung-digital-picture-frame-stores-pics-movies-music/
It looks exactly like an iPad, and exactly like the "electronic device" in Apple's design patent.
Now you'll probably say "but that's not a tablet!". I would reply that we're talking about design here, and confusingly similar looks, not operation. In fact the behavior of the device is explicitly called out as not being part of the design patent Apple filed [1].
Furthermore if how you use something matters, rather than trade dress, we can just call the Galaxy Tab a "digital picture frame and music player", and close the case in Samsung's favor.
[1] Patent title: "Electronic Device"
Primary Claim: "We claim the ornamental design for an electronic device, substantially as shown and described."
Label for only picture showing someone using the device: "FIG. 9 is an exemplary diagram of the use of the electronic device thereof the broken lines being shown for illustrative puposes only and form no part of the claimed design. "
IOW, there is nothing whatsoever in this patent that is inconsistent with a digital picture frame and music player, and they are not claiming anything in the patent that is specific to a tablet.
http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.htm&r=1&f=G&l=50&s1=D504,889.PN.&OS=PN/D504,889&RS=PN/D504,889
Yes because going to:
http://www.google.com/privacy/ads/
and clicking on a button labelled "Opt out" is too damn hard.
Of course according to you Google wouldn't do this since it is their core business, right?
DNT is a nuclear weapon for honest companies, and a joke to be ignored for dishonest ones (it's like the famous Evil Bit). What is really needed is European-style privacy regulations (i.e. with actual teeth for bad actors), and a fine-grained permissions model on Websites (like iOS or Android has for apps).
So, if visual design is what matters, and not the software (which is different since the iOS is so superior as you would remind us), what do you think about this Samsung digital picture frame from 2006?:
http://www.engadget.com/2006/03/09/samsung-digital-picture-frame-stores-pics-movies-music/
Form follows function, which is why every TV and computing device is destined to look the same once a level of miniaturization is reached. That's why you see similar tablets in a 1970s TV show ("The Tomorrow People") and a movie from the 1960s ("2001 a Space Odyssey").
Apple certainly is a style trendsetter, but really it is more about bringing things to market that are the closest to what the visionaries have already described. There's a rather clear evolution from other mp3 players through to the ipod, iphone, and then ipad. And yes, along the way things came from non-Apple sources too.. the next iphone will have screen dimensions suspiciously like an HTC EVO, and a notification bar straight from Android. That's what happens in competition.
I'm sorry this conflicts with your worldview that all these nice Apple products were invented in a vacuum.
I want to obliterate certain sites from search results
You can :)
google.com | gear icon | search settings | Manage blocked sites
How will you feel about it if the Roddenberry estate and Stanley Kubrick sue Apple (Look at the pad Bowman uses in 2001).
Here's a good link: :)
http://www.androidpads.com/2010/05/19/back-to-the-future-the-newspad-of-2001-a-space-odyssey/
In comparison current tablets are still too thick and screens too small
Homes are the only thing that are still designed solely for bipeds. The public world is already mostly ADA compliant:
http://en.wikipedia.org/wiki/Americans_with_Disabilities_Act_of_1990#Title_III_-_Public_Accommodations_.28and_Commercial_Facilities.29
(other countries have similar laws, some even more comprehensive)
I'd happily modify my house somewhat to accommodate robots (door handles, rails for climbing).
IMHO walking is not what is keeping robots out of our homes and our lives. It may be what is keeping them off of the battlefield however.
When I read the summary, I thought exactly that -- I wonder if this is another planar walker?
Planar walkers were already doing flips 20 years ago, and walking quite convincingly and naturally 12 years ago. Many of those researchers have moved to 3D now, with some of the most impressive recent work showing up at Boston Dynamics.
There still is good research that can be done on semi-passive force control and efficiency, and a lot of that can be tested on planar walkers. However, for the love of all that is honest, at least mention in the summary that all the things you are claiming are done in a 3 d.o.f. workspace rather than the 6 d.o.f. you have in the real world. A bar also offers a metric load of damping, which is the thing keeping the full-3D walkers from really working naturally, and thus that kind of simplification should be mentioned.
NP. The torrent of articles decided to copy the "ZOMG Google car crashes!" part, but left out the additional information from primary sources.
There's no way to have it search what you told it to search any more.
http://www.google.com/#q=%2B%22yes+there+is%22
Normal users often search for vague terms, and don't understand things like operators, so the default has to be to support that majority of users. Power users can learn to use quoting, plus, and other operators, so it is up to you to learn those features and employ them.
Last I heard google has not commented on the accident.
It was in the first article on any news site:
A Google spokesperson gave us this quote about the accident: "Safety is our top priority. One of our goals is to prevent fender-benders like this one, which occurred while a person was manually driving the car."
http://www.businessinsider.com/googles-self-driving-cars-get-in-their-first-accident-2011-8
This is slashdot, you're supposed to make overly broad generalizations from minimal data. Why relate your story when you can use a position you once had to make a statement about the entire industry? :)
It's probably too late to get modded up so you can see this, but:
The technical term you are looking for is "coverage path planning", and there are well known algorithms to solve it efficiently such as Choset's Boustrophedon. I don't think any of them are optimal for a nontrivial shape, but they will probably beat most human heuristics.
In addition to lawns, this is important for machining (material removal), de-mining (completeness is essential, but overlap is expensive), and large floor cleaning.
http://scholar.google.com/scholar?q=coverage+path+planning
I don't really have a strong opinion on the subject, but what you claim and what I've seen while traveling abroad don't seem to agree. I did some digging about excessive fees and found the following:
From: http://en.wikipedia.org/wiki/Attorney's_fee#Which_party_pays
There are many ways of calculating prevailing-party attorney fees. Most courts recognize that actual costs may be disproportionate and inequitable. Thus, many jurisdictions rely on other calculations. Many courts or laws invoke a lodestar' calculation: reasonably expected billable hours multiplied by a reasonable hourly rate, sometimes multiplied by a factor reflecting the risk or complexity of the case.
An old but fairly well written article on the subject: http://reason.com/archives/1995/06/01/civil-suits/1
(I realize many will not trust this source, but the article pretty much just states what the situation is in European countries.)
A common fear about loser-pays is that the side who loses a routine dispute will get handed a bill for 10,000 hours from Cravath, Swaine & Moore. But European courts are well aware of the danger that successful litigants will overinvest in their cases and gold-plate their fee requests. They carefully control the process to prevent that danger, giving the losing side a full chance to dispute a fee award, requiring that work be reasonable and necessary, providing that elite lawyer rates not be paid if a Main Street lawyer could have done the job, and so forth.
Although fee awards are usually set high enough to apply serious incentive pressure, it is notable that no country appears to let winners recoup all the money they have spent on a suit. Some leave a portion of lawyers' hours unreimbursed, while others apply hourly rates that fall below prevailing levels. Some shift expert witness fees, but others do not.
Here's an interesting nugget from later in the article. Unsurprisingly, lawyers are the most outspoken critics of loser-pays, yet their clients don't seem to agree:
The American public itself seems not to be as terrified of full loser-pays as its lawyers. Even when U.S. News and World Report asked the question in its more menacing form--"If you sue someone and lose the case, should you pay his costs?"--a far-from-shabby 44 percent of those polled agreed that they should. And with sides reversed, nearly everyone saw the idea's fairness. "If someone sues you and you win the case, should he pay your legal costs?" Eighty-five percent said yes to that one.
so unless you plan on following a bunch of spam bots on Google+
A *lot* of people follow Robert Scoble you insensitive clod!
Can you stop posting repeatedly about how 600GB is big and hard to miss? On your home machine, yes you'd notice it. However, compared to the size of four years worth of high-res panoramas taken every few meters on a significant fraction of the world's developed roads? In that context, 600GB is quite literally nothing. When a car dumps 1TB of photos, you're not necessarily going to notice that an adjoining tar file of text logs is a couple MB too large. When you store that on 10000 machines, you're not going to notice that each one is using 60MB of hard drive space more than you expected.
Are you running apache anywhere? Can you recite the exact settings and the log retention time? Might there be extra switches you left on and forgot about?
You should have tried 20% TaskForce projects, they work great. You bank up a quarter's worth of 20% time (1-2 weeks) then use it all at once to work in small groups at a shared location. My new team started up this spring, and was originally based on a 20% TaskForce from last fall.
So no, 20% time is not dead, but you do need to get off your ass to use it. Just like vacation time, nobody can force you to use it, and if you don't make use of it you lose it.
Of course, 20% time to work on something completely random with no possible purpose for the company doesn't exist, but it never existed anyway. That's what you call "projects in your spare time" and it works much the way it does in other semi-flexible companies. I'm glad the company lets me use equipment and resources (within reason) for random projects like that, but expecting them to pay me for that time is unreasonable.