According to their Japanese Q&A page, their membrane lasts as long or longer than hydrogen fuel cells. They then go on to say that typical household(?) upright hydrogen fuel cell stacks last around 40,000 hours.
For the case of 2ch, there are a plethora of specialized software(Japanese link) used to browse the message boards. For the occasional readers and posters on PCs, the web page interface is enough, but the more prevalent posters use the specialized software. For example, on Windows/Linux, I would use moz2ch, a plugin for Firefox/Mozilla that reads 2ch. On OSX, I would use CocoMonar, etc. These software packages have kind of obviated the massive need for a Web 2.0 style interface for 2ch.
When I bought my 12" Powerbook G4 a while back, it came with various random stuff, like trial versions of Office, loaded onto it. Granted, it's not as bad as some of the Windows preloads get, but it's still far from a "clean" system. On the other hand, the Mac Pro I just bought came very clean, and I only reinstalled Leopard because I wanted it on a different hard drive.
Not enough return on investment... (at least in the near future)
Time doesn't slow down but our perception does?
on
Can Time Slow Down?
·
· Score: 5, Interesting
I think our perception of the passage of time in the past is purely based on our memories. Thus, certain things that are very memorable will probably mess around with our perception of the flow of time during that moment. For example, if you remember nothing after passing out from drinking and wake up the next day, you probably wouldn't feel like you actually spent all that time lying on the floor.
2. No US media company would ever have bought hard-to-categorize shows like Death Note, Nana or Prince of Tennis before the fansub community proved that there was a market for such shows among western viewers. Fansubs are basically free market research for the distributors.
Don't be silly and use examples like Death Note. Death Note was insanely popular in Japan and has been out in translated manga form in the US for a while now and NANA has also been available in manga form in the US as well. If you're going to use examples, at least mention Azumanga Daioh, which is more likely to have been licensed because of fansub popularity. Keep in mind that research from fansubs determine what people want to watch, and not necessarily what people want to pay money for.
3. The big money in US anime distribution comes from dubbing shows with English-speaking actors and putting it on cable TV. When a show is released to DVD as a subtitle-only set (such as season 2 of SuperGALS!, or the "Uncut" editions of Seasons 1 & 2 of Sailor Moon,) sales have been lackluster at best.
I think you might be confusing cause and effect here. Subtitle-only DVD releases are generally only done when they believe there would not be a good enough return on the dubbing. Thus, usually it's the relatively niche shows that get that treatment, instead of the lack of dubbing causing less people to buy it.
Microsoft owes everything it has to Unix, since C was created for Unix, and Windows couldn't have been written with C...
Ah, so that explains why Vista runs so slowly... Instead of C, they decided to write Vista using C# and then had it run in interpreted mode!
Interesting... Coincidentally, Planar demoed a type of monitor like this in the Emerging Technologies room at SIGGRAPH 2007. It sounds like this Sharp monitor uses embedded sensors, but the Planar one just uses one of the transistors in the TFT matrix as an phototransistor. I tried one out and it was pretty neat. One of the drawbacks right now is that shadows causes problems when trying to determine if a user has pressed a certain part of the screen. Furthermore, this sensor requires some sort of illumination, such as a lamp or something shining on the display. They talked about perhaps integrating some sort of microlens system to focus on a particular plane, such as right above the screen, to highlight objects in that plane.
One neat thing they did was place a document on top of the monitor and turned off the surrounding lights. If the monitor displayed a pure white image, the resulting sensor readout gave you a (low-quality) scan of the document.
Inpainting works fine for the examples they gave on the GREYCstoration page, but it totally fails at large, contiguous areas. GREYCstoration's target problem is to restore thin or small areas whereas this paper seeks to restore larger areas. If you tried to inpaint, say a large block of a photograph that is just ocean, it will fail miserably.
When they gave this presentation at work, one of the things they wanted to stress was that before, scientists were coming up with clever abstractions and trying to develop an artifical intelligence to solve many problems. Nowadays, if there is enough data, using the data in search/matching will rival the power of the "AI." For example, let's suppose our problem is OCR of people's handwriting. Without enough data, we might want to construct a "smart" approach whereby we analyze a few people's handwriting and try to derive rules to detect characters. However, if you have a gigantic dictionary/library of handwriting samples, you might do better to try to match your input to a sample from the library rather than the spend time developing rules.
Apart from the registration of images to each other, they are very different systems. Photosynth is used to organize a collection of similar images into a cluster and register them with each other in order extract an intuitive 3D browsing method of photographs of the same thing. It requires a tremendous amount of pre-processing to use Photosynth. For this hole-filling algorithm, although it is searching for matches as well, it is doing so in a less strict fashion; it merely tries to find "similar" images instead of images of the same thing.
For example, if I use a Photosynth collection of the Notre Dame, and I wanted to pan past the right of an image, Photosynth will find another image of the Notre Dame from the viewpoint I want, even if the lighting is different (day vs. night). If I used this with this hole completion algorithm described in the article, it'll find something that is similar to the right side of my image, but it might turn out to be, say, the Pantheon under similar lighting conditions.
Actually, over at the Center for Computer Research in Music and Acoustics, the majority of the machines are running Linux and at least at the workstations I've used, they were using external M-Audio firewire boxes, with good results.
What I find interesting is not the fact that he took audio/video pieces and spliced together a song. What I find interesting is the way he presents the process of making a song and the fact that what was commonly done for music (snippet sampling) is applied in unison to video. That being said, the technical details aren't really that interesting.
PS: And yes, I do compose/produce music, reverse engineer and splice music videos together in my spare time.
Other than the cadence section, I didn't really find any direct comparisons of de-interlacing quality from the review. One feature that's particularly interesting on high-end stand-alone DVD players is DCDi, which is some nice looking adaptive de-interlacing method. It looks like nVidia's PureVideo has some smart de-interlacing method in its decoder, but how does that compare to DCDi?
Your argument, reworded: Although it's always nice to have more tools for content and information dissemination...these so called webpages will probably be made by a bunch of idiots, and therefore will be worth less than the time it takes to read / download them...
Maybe you should realize that the point of visual novels isn't the engine itself, but rather the content. Besides, any type of entertainment software is copyable, but would you want to play 20 copies of the same game?
I had written about the whole Leaf debacle in an article submission previously to Slashdot but it was rejected. Anyway, a few weeks after their declaration, an anonymous person uploaded the entire source code to an upload service, thus giving it out to anyone who wanted it. Keep in mind, though, that the guts of Leaf's games are the actual content, like graphics, music, art and story, instead of the engine. In fact, I'm pretty sure some capable programmer could come up with Leaf's engine in maybe a month.
You will find that while a majority of these games are crappy porn-laden games, the ones that sell record quantities are the ones that keep the porn part to a minimum and focus on the story, character development, art and music. In fact, once these get popular enough, clean versions of the games are made (for the PS2, for example) and sell quite well compared to the "dirty" versions. There are also many of these games that start out without the adult content and sell well because of the story and character interaction. One particular example would be CLANNAD.
While one may debate whether these are actually "games," the top selling visual novels are definitely high quality entertainment products comparable to many well-written dead-tree novels. In fact, one thing interactive fiction has over regular fiction is that it allows the reader to explore a story, layer by layer, based on the choices the reader makes. The visual novel also uses music and graphics to carry the point across even more. Ever17, one of the top selling (fully clean) visual novel style games in Japan used this technique and you can in fact purchase an English version from http://www.hirameki-int.com/.
So yes, it is a stereotype, just as anime used to be stereotyped as "cartoon porn" even when the majority of anime was not pornographic at all. Visual novels, as it is referred to in English, is simply a type of medium upon which to deliver content. It is up to the creator to decide what type of content to put on it. In fact, one of the top selling visual novel series in Japan isn't even made by a company. It's made by an amateur production group, and the games create a running mystery-suspense story.
I actually work with it in the lab. It was quite fun trying to read the tiny little terminal windows on the screen. Eventually, I just kicked the font size up and it looked beautifully crisp. Unfortunately, for some reason, MATLAB had defaulted to the super large font size, and made it temporarily impossible to use now that I switched back to a smaller dpi display.
Slashdot Mirror
According to their Japanese Q&A page, their membrane lasts as long or longer than hydrogen fuel cells. They then go on to say that typical household(?) upright hydrogen fuel cell stacks last around 40,000 hours.
For the case of 2ch, there are a plethora of specialized software(Japanese link) used to browse the message boards. For the occasional readers and posters on PCs, the web page interface is enough, but the more prevalent posters use the specialized software. For example, on Windows/Linux, I would use moz2ch, a plugin for Firefox/Mozilla that reads 2ch. On OSX, I would use CocoMonar, etc. These software packages have kind of obviated the massive need for a Web 2.0 style interface for 2ch.
When I bought my 12" Powerbook G4 a while back, it came with various random stuff, like trial versions of Office, loaded onto it. Granted, it's not as bad as some of the Windows preloads get, but it's still far from a "clean" system. On the other hand, the Mac Pro I just bought came very clean, and I only reinstalled Leopard because I wanted it on a different hard drive.
Yeah, it was such a Paine without you!
Not enough return on investment... (at least in the near future)
I think our perception of the passage of time in the past is purely based on our memories. Thus, certain things that are very memorable will probably mess around with our perception of the flow of time during that moment. For example, if you remember nothing after passing out from drinking and wake up the next day, you probably wouldn't feel like you actually spent all that time lying on the floor.
Microsoft owes everything it has to Unix, since C was created for Unix, and Windows couldn't have been written with C...
Ah, so that explains why Vista runs so slowly... Instead of C, they decided to write Vista using C# and then had it run in interpreted mode!
Interesting... Coincidentally, Planar demoed a type of monitor like this in the Emerging Technologies room at SIGGRAPH 2007. It sounds like this Sharp monitor uses embedded sensors, but the Planar one just uses one of the transistors in the TFT matrix as an phototransistor. I tried one out and it was pretty neat. One of the drawbacks right now is that shadows causes problems when trying to determine if a user has pressed a certain part of the screen. Furthermore, this sensor requires some sort of illumination, such as a lamp or something shining on the display. They talked about perhaps integrating some sort of microlens system to focus on a particular plane, such as right above the screen, to highlight objects in that plane.
One neat thing they did was place a document on top of the monitor and turned off the surrounding lights. If the monitor displayed a pure white image, the resulting sensor readout gave you a (low-quality) scan of the document.
Kind of reminds me of http://bash.org/?104052
Yes, you are correct. I took a look at the original article which actually had a scale bar for the image and it amounts to roughly 200 micron. The linked article is kind of scant on details, so here's a more detailed writeup: http://www.sciencedaily.com/releases/2007/08/07081 2173253.htm
Inpainting works fine for the examples they gave on the GREYCstoration page, but it totally fails at large, contiguous areas. GREYCstoration's target problem is to restore thin or small areas whereas this paper seeks to restore larger areas. If you tried to inpaint, say a large block of a photograph that is just ocean, it will fail miserably.
When they gave this presentation at work, one of the things they wanted to stress was that before, scientists were coming up with clever abstractions and trying to develop an artifical intelligence to solve many problems. Nowadays, if there is enough data, using the data in search/matching will rival the power of the "AI." For example, let's suppose our problem is OCR of people's handwriting. Without enough data, we might want to construct a "smart" approach whereby we analyze a few people's handwriting and try to derive rules to detect characters. However, if you have a gigantic dictionary/library of handwriting samples, you might do better to try to match your input to a sample from the library rather than the spend time developing rules.
Apart from the registration of images to each other, they are very different systems. Photosynth is used to organize a collection of similar images into a cluster and register them with each other in order extract an intuitive 3D browsing method of photographs of the same thing. It requires a tremendous amount of pre-processing to use Photosynth. For this hole-filling algorithm, although it is searching for matches as well, it is doing so in a less strict fashion; it merely tries to find "similar" images instead of images of the same thing.
For example, if I use a Photosynth collection of the Notre Dame, and I wanted to pan past the right of an image, Photosynth will find another image of the Notre Dame from the viewpoint I want, even if the lighting is different (day vs. night). If I used this with this hole completion algorithm described in the article, it'll find something that is similar to the right side of my image, but it might turn out to be, say, the Pantheon under similar lighting conditions.
How about: Pidgin Is Definitely Gaim In Nature
?
Actually, over at the Center for Computer Research in Music and Acoustics, the majority of the machines are running Linux and at least at the workstations I've used, they were using external M-Audio firewire boxes, with good results.
What I find interesting is not the fact that he took audio/video pieces and spliced together a song. What I find interesting is the way he presents the process of making a song and the fact that what was commonly done for music (snippet sampling) is applied in unison to video. That being said, the technical details aren't really that interesting.
PS: And yes, I do compose/produce music, reverse engineer and splice music videos together in my spare time.
Other than the cadence section, I didn't really find any direct comparisons of de-interlacing quality from the review. One feature that's particularly interesting on high-end stand-alone DVD players is DCDi, which is some nice looking adaptive de-interlacing method. It looks like nVidia's PureVideo has some smart de-interlacing method in its decoder, but how does that compare to DCDi?
Your argument, reworded:
Although it's always nice to have more tools for content and information dissemination...these so called webpages will probably be made by a bunch of idiots, and therefore will be worth less than the time it takes to read / download them...
Maybe you should realize that the point of visual novels isn't the engine itself, but rather the content. Besides, any type of entertainment software is copyable, but would you want to play 20 copies of the same game?
I had written about the whole Leaf debacle in an article submission previously to Slashdot but it was rejected. Anyway, a few weeks after their declaration, an anonymous person uploaded the entire source code to an upload service, thus giving it out to anyone who wanted it. Keep in mind, though, that the guts of Leaf's games are the actual content, like graphics, music, art and story, instead of the engine. In fact, I'm pretty sure some capable programmer could come up with Leaf's engine in maybe a month.
You will find that while a majority of these games are crappy porn-laden games, the ones that sell record quantities are the ones that keep the porn part to a minimum and focus on the story, character development, art and music. In fact, once these get popular enough, clean versions of the games are made (for the PS2, for example) and sell quite well compared to the "dirty" versions. There are also many of these games that start out without the adult content and sell well because of the story and character interaction. One particular example would be CLANNAD.
While one may debate whether these are actually "games," the top selling visual novels are definitely high quality entertainment products comparable to many well-written dead-tree novels. In fact, one thing interactive fiction has over regular fiction is that it allows the reader to explore a story, layer by layer, based on the choices the reader makes. The visual novel also uses music and graphics to carry the point across even more. Ever17, one of the top selling (fully clean) visual novel style games in Japan used this technique and you can in fact purchase an English version from http://www.hirameki-int.com/.
So yes, it is a stereotype, just as anime used to be stereotyped as "cartoon porn" even when the majority of anime was not pornographic at all. Visual novels, as it is referred to in English, is simply a type of medium upon which to deliver content. It is up to the creator to decide what type of content to put on it. In fact, one of the top selling visual novel series in Japan isn't even made by a company. It's made by an amateur production group, and the games create a running mystery-suspense story.
I actually work with it in the lab. It was quite fun trying to read the tiny little terminal windows on the screen. Eventually, I just kicked the font size up and it looked beautifully crisp. Unfortunately, for some reason, MATLAB had defaulted to the super large font size, and made it temporarily impossible to use now that I switched back to a smaller dpi display.
Apparently, interactive fiction is a lot more popular in Japan than over here: http://en.wikipedia.org/wiki/Visual_novel
Dunno about using force pull on the laptop, but the laptop can sure force pull its own power cable.
Having that many extra pixels allows one to capture higher resolution light fields and thus allow for after-the-fact refocusing:
http://graphics.stanford.edu/papers/lfcamera/