It's different because it degrades gracefully for text-only browsers and screen readers. As I recall, the JavaScript estimates the size of the headline based on your font settings and renders the headline accordingly, so people browsing with their font size at extra large aren't affected.
The only time it doesn't work for triple-sized fonts is if you adjust the font size while viewing the page. The JavaScript runs on page load, so you'd have to refresh to get the increased sizes to be reflected by the Flash headline.
Yes, screen readers can handle it. The text remains in the HTML, and the JavaScript just creates Flash headlines containing it on the fly for user agents supporting it. One of the most important advantages of sIFR over PNG/GIF headlines is that the original text remains with no trickery, so it's 100% accessible for people with disabilities and for text-only web browsers.
As for your custom font settings, yes, I suppose it does, but so do PNG/GIF headline images. If you disable JavaScript or Flash, you should see your preferred font.
You don't have to have it enabled. It degrades gracefully -- if JavaScript is disabled or if the user doesn't have flash, regular text is displayed. That's the big advantage, and that's why it works well enough to have "arrived." Users who don't want JavaScript enabled don't lose any content, but users who do (a vast majority) see the page presented as the designer intended.
sIFR has come far enough to be used on major websites. Aside from seeing it in places like one of the website of the person who perfected it, it's also appearing on mainstream websites -- take ABC News. Their headlines are rendered using sIFR for browsers that support it.
I'm not a hardcore gamer, but I've had good luck with a GeForce4 Ti that I bought last year running under Gentoo. A lot of new games are coming available for Linux (Doom 3 comes to mind -- I don't have it, but I play Quake 3 Arena and find the Linux Q3A experience to be more pleasant than the Windows experience), and for those that don't, WINE and Cedega can sometimes pick up the slack.
I have both WINE and Cedega installed on my Gentoo system (and, FWIW, you can obtain Cedega from CVS without paying for it so long as you don't need the Point2Play feature, which I don't use despite having a licensed version). Cedega tackles some games relatively well. Grand Theft Auto III, for instance, runs nicely for me, and Starsiege: Tribes is essentially flawless.
I don't have a Windows machine to use for gaming and I find that the Linux gaming options available are most sufficient for me. The nvidia drivers aren't perfect (they are the only reason I ever see crashes), but they work and perform decently, and games under Cedega run well enough that I can enjoy them. Native games like Q3A, UT2k3 and UT2k4 run flawlessly.
The subject of photoblogging reminds me quite a bit of "lomography," which has apparently taken off among certain circles. It gets its name from a Russian camera called the Lomo which is a consumer "point and shoot" camera with some unusual properties as far as quality goes.
As I understand it, the hardware acceleration in Aqua is a "recent" development. Initial releases of OS X did not include hardware acceleration (using OpenGL and video card memory instead of CPU operations and system memory) for compositing, and the newer versions of the OS that do include that capability can default back to a software rendering mode.
If you install OS X 10.3 (Panther) on an older system without a competent video card, for instance, the OS simply disables Quartz Extreme (the hardware compositing feature) without any fuss. (There's also no apparent way to turn it on manually -- as with some of the other features, the OS simply thinks it knows what is best for you and does what it thinks is right, which isn't always a bad thing.)
And a few days after the auction is won, who wants to bet that 1300 hard-to-find ROMs for obscure game systems are going to appear on popular emulation sites?
Computer Surplus Outlet always looked good to me. I've never purchased from them, so I can't vouch for them as far as service and reliability goes, but they have newer and older hardware for pretty nice prices at times.
I'm sorry to hear that some of the article's technical details were edited out. I can understand why the editors would do so, though. Ars obviously isn't only targeting Slashdot readers, hehe.
Would you be willing (or able) to post your original version of the article somewhere for interested parties to read?
I read this piece yesterday, and while it did "dumb down" the basics (as the first poster noted), I thought it did a very good job of putting it all into a nutshell that those of us not as familiar with Big-O and schedulers in general might easily understand. For Linux.Ars' format, I thought it was of appropriate length, and had enough detail to "belong." I'm sure there are more detailed writeups on the O(1) scheduler in place in 2.6. Does anyone have any links?
I most definitely concede that supercomputers have most definitely consumed a far greater amount of power traditionally than modern enthusiast machines. I'm more curious about the heat output (I know the figures cited in the original article referred to power consumption, but heat was mentioned). My post was a bit vague. A clarification might help.
I wonder how long ago that type of heat output would have been considered extreme in a supercomputer, or perhaps at least in one node making up a supercomputer.
Of course when we compare heat output of a traditional supercomputer to a top-of-the-line processor from an enthusiast machine (say a Pentium 4 EE or an Athlon 64 FX), a modern CPU might offer better performance on the order of hundreds of times (if not thousands) for the same amount of heat produced.
Most definitely nothing to sneeze at. I have to ask, though: how long ago was it insane for a supercomputer to put out as much heat as the average enthusiast PC puts out today?
How, though, do they intend to verify that those applying for these cards really have the "credentials" being given to them? Background searches on that kind of a scale would be an extremely intensive undertaking for any organization. Furthermore, there is no way this could be done for the $30 or $50 mentioned in the article.
They could, I suppose, require the applicant to submit proof that they meet the requirements for obtaining one of these cards, but then that raises a new problem: falsified records/information.
"He said that the system was probably unworkable." I'd say so.
The oldest hardware I am currently using actively is my floppy drive, pulled from a 33MHz 486 machine before it bit the dust. I also have a Creative Labs RivaTNT (not TNT2, just TNT) card in heavy use on a secondary box.
Out of hardware used slightly less frequently, the oldest hardware I run would have to be my IBM PS/2 Model 55 (Intel 386 at 25MHz). There is still something pleasing about it, even sixteen years after it was manufactured. It runs DOS because I'm being nostalgic and preserving the filesystem in the condition I left it in when I graduated to 486-land.;)
Although I am not sure about the other schools in my school district, my particular high school does use Pinnacle. I find it to be a very useful tool, personally, because I am able to verify that my teachers have my grades entered correctly and so on and so forth.
Until recently, there was a very large hole in its security, at least from inside our school's intranet. Usernames for our Pinnacle system are the student ID numbers, and passwords correspond to abbreviated birthdays. A Word document containing all of this information (for teacher use) was available on a server on the school's intranet, visible to anyone logged in (presumably district-wide). The server was recently password-protected, but a careless teacher could leave this page open, allowing an observant student to obtain the necessary information to examine a classmate's grades. I am also aware of one student-created mirror of these intranet pages, although the student responsible is not distributing the information.
I think it's almost definitely insecure. The Pinnacle web frontend ties into the grades database inside the intranet. As best I can tell, at my school, both the backend database and the website live on the same server. Very insecure, in my eyes.
The system has its upsides and downsides... But I don't believe my district has safely and securely implemented its Pinnacle installation, so that does definitely raise concerns for me.
Mr. Shatner,
When looking back on your career thus far, is there anything you've done which you feel to have been an important contribution either to society or to humanity? Why do you feel that contribution is important?
Slashdot Mirror
It's different because it degrades gracefully for text-only browsers and screen readers. As I recall, the JavaScript estimates the size of the headline based on your font settings and renders the headline accordingly, so people browsing with their font size at extra large aren't affected.
The only time it doesn't work for triple-sized fonts is if you adjust the font size while viewing the page. The JavaScript runs on page load, so you'd have to refresh to get the increased sizes to be reflected by the Flash headline.
No, because the text remains in the HTML file.
Yes, screen readers can handle it. The text remains in the HTML, and the JavaScript just creates Flash headlines containing it on the fly for user agents supporting it. One of the most important advantages of sIFR over PNG/GIF headlines is that the original text remains with no trickery, so it's 100% accessible for people with disabilities and for text-only web browsers.
As for your custom font settings, yes, I suppose it does, but so do PNG/GIF headline images. If you disable JavaScript or Flash, you should see your preferred font.
You don't have to have it enabled. It degrades gracefully -- if JavaScript is disabled or if the user doesn't have flash, regular text is displayed. That's the big advantage, and that's why it works well enough to have "arrived." Users who don't want JavaScript enabled don't lose any content, but users who do (a vast majority) see the page presented as the designer intended.
sIFR has come far enough to be used on major websites. Aside from seeing it in places like one of the website of the person who perfected it, it's also appearing on mainstream websites -- take ABC News. Their headlines are rendered using sIFR for browsers that support it.
I'm not a hardcore gamer, but I've had good luck with a GeForce4 Ti that I bought last year running under Gentoo. A lot of new games are coming available for Linux (Doom 3 comes to mind -- I don't have it, but I play Quake 3 Arena and find the Linux Q3A experience to be more pleasant than the Windows experience), and for those that don't, WINE and Cedega can sometimes pick up the slack.
I have both WINE and Cedega installed on my Gentoo system (and, FWIW, you can obtain Cedega from CVS without paying for it so long as you don't need the Point2Play feature, which I don't use despite having a licensed version). Cedega tackles some games relatively well. Grand Theft Auto III, for instance, runs nicely for me, and Starsiege: Tribes is essentially flawless.
I don't have a Windows machine to use for gaming and I find that the Linux gaming options available are most sufficient for me. The nvidia drivers aren't perfect (they are the only reason I ever see crashes), but they work and perform decently, and games under Cedega run well enough that I can enjoy them. Native games like Q3A, UT2k3 and UT2k4 run flawlessly.
That's odd. It's working just fine for me in Firefox under Linux.
The subject of photoblogging reminds me quite a bit of "lomography," which has apparently taken off among certain circles. It gets its name from a Russian camera called the Lomo which is a consumer "point and shoot" camera with some unusual properties as far as quality goes.
Anything and Everything is a good place to start checking out lomography.
As I understand it, the hardware acceleration in Aqua is a "recent" development. Initial releases of OS X did not include hardware acceleration (using OpenGL and video card memory instead of CPU operations and system memory) for compositing, and the newer versions of the OS that do include that capability can default back to a software rendering mode. If you install OS X 10.3 (Panther) on an older system without a competent video card, for instance, the OS simply disables Quartz Extreme (the hardware compositing feature) without any fuss. (There's also no apparent way to turn it on manually -- as with some of the other features, the OS simply thinks it knows what is best for you and does what it thinks is right, which isn't always a bad thing.)
And a few days after the auction is won, who wants to bet that 1300 hard-to-find ROMs for obscure game systems are going to appear on popular emulation sites?
Computer Surplus Outlet always looked good to me. I've never purchased from them, so I can't vouch for them as far as service and reliability goes, but they have newer and older hardware for pretty nice prices at times.
I'm sorry to hear that some of the article's technical details were edited out. I can understand why the editors would do so, though. Ars obviously isn't only targeting Slashdot readers, hehe.
Would you be willing (or able) to post your original version of the article somewhere for interested parties to read?
I read this piece yesterday, and while it did "dumb down" the basics (as the first poster noted), I thought it did a very good job of putting it all into a nutshell that those of us not as familiar with Big-O and schedulers in general might easily understand. For Linux.Ars' format, I thought it was of appropriate length, and had enough detail to "belong." I'm sure there are more detailed writeups on the O(1) scheduler in place in 2.6. Does anyone have any links?
I most definitely concede that supercomputers have most definitely consumed a far greater amount of power traditionally than modern enthusiast machines. I'm more curious about the heat output (I know the figures cited in the original article referred to power consumption, but heat was mentioned). My post was a bit vague. A clarification might help. I wonder how long ago that type of heat output would have been considered extreme in a supercomputer, or perhaps at least in one node making up a supercomputer. Of course when we compare heat output of a traditional supercomputer to a top-of-the-line processor from an enthusiast machine (say a Pentium 4 EE or an Athlon 64 FX), a modern CPU might offer better performance on the order of hundreds of times (if not thousands) for the same amount of heat produced.
Most definitely nothing to sneeze at. I have to ask, though: how long ago was it insane for a supercomputer to put out as much heat as the average enthusiast PC puts out today?
How, though, do they intend to verify that those applying for these cards really have the "credentials" being given to them? Background searches on that kind of a scale would be an extremely intensive undertaking for any organization. Furthermore, there is no way this could be done for the $30 or $50 mentioned in the article. They could, I suppose, require the applicant to submit proof that they meet the requirements for obtaining one of these cards, but then that raises a new problem: falsified records/information. "He said that the system was probably unworkable." I'd say so.
The oldest hardware I am currently using actively is my floppy drive, pulled from a 33MHz 486 machine before it bit the dust. I also have a Creative Labs RivaTNT (not TNT2, just TNT) card in heavy use on a secondary box. Out of hardware used slightly less frequently, the oldest hardware I run would have to be my IBM PS/2 Model 55 (Intel 386 at 25MHz). There is still something pleasing about it, even sixteen years after it was manufactured. It runs DOS because I'm being nostalgic and preserving the filesystem in the condition I left it in when I graduated to 486-land. ;)
Yes indeed it is a first post. Congratulations. ;)
You beat me to it. ;)
Although I am not sure about the other schools in my school district, my particular high school does use Pinnacle. I find it to be a very useful tool, personally, because I am able to verify that my teachers have my grades entered correctly and so on and so forth.
Until recently, there was a very large hole in its security, at least from inside our school's intranet. Usernames for our Pinnacle system are the student ID numbers, and passwords correspond to abbreviated birthdays. A Word document containing all of this information (for teacher use) was available on a server on the school's intranet, visible to anyone logged in (presumably district-wide). The server was recently password-protected, but a careless teacher could leave this page open, allowing an observant student to obtain the necessary information to examine a classmate's grades. I am also aware of one student-created mirror of these intranet pages, although the student responsible is not distributing the information.
I think it's almost definitely insecure. The Pinnacle web frontend ties into the grades database inside the intranet. As best I can tell, at my school, both the backend database and the website live on the same server. Very insecure, in my eyes.
The system has its upsides and downsides... But I don't believe my district has safely and securely implemented its Pinnacle installation, so that does definitely raise concerns for me.
Mr. Shatner, When looking back on your career thus far, is there anything you've done which you feel to have been an important contribution either to society or to humanity? Why do you feel that contribution is important?